Download - Lead Corrosion Control Chemistry
1
Lead Corrosion Control Chemistry
Simoni TriantafyllidouUS EPArsquos Office of Research and Development Cincinnati
EPA ORD-Region 4 2020 Small Drinking Water Systems Meeting10142020
2
Who we are -Acknowledgments
Jennifer Tully
SimoniTriantafyllidou
Mike DeSantis
Christy Muhlen
DarrenLytle
Steve Harmon
Casey Formal
Mike Schock
Dan Williams
Regan Murray
Jonathan Burkhardt
Evelyne Doreacute
Mariah Caballero
33
What we do
ORD provides the data tools and information that form the sound scientific foundation the Agency relies on to fulfill its mission to protect the environment and safeguard public healthORD at a Glance httpswwwepagovaboutepaabout-office-research-and-development-ord Center for Environmental
Solutions amp Emergency Response (CESER) in CincinnatiWe conduct applied stakeholder-driven research and provide responsive technical support to help solve the Nationrsquos environmental challenges
CESER at a Glance httpswwwepagovaboutepaabout-center-environmental-solutions-and-emergency-response-ceser
Pb sources
BUILDING
bull Lead Service Lines (LSLs)bull Lead Goosenecksbull Leaded Solder
bull Leaded Brass (valves fittings faucets fountains)bull Galvanized Iron Pipe downstream of leaded plumbing
Triantafyllidou amp Edwards 2012
Pb sources
Full Lead Service Line
Clark et al 2013
Brass unionvalve vs plastic
Partial Lead Service Line
Lead gooseneckTriantafyllidou et al 2020
What is in the distribution system Materials inventory is critical to understand where and what lead sources still exist
Pb sources
Leaded Solder
2 Pb
lt05Pb
3 Pb
Leaded Brass FaucetSelover 2005
Gal
vani
zed
Diverse legacy leaded materials may undergo different corrosion reactions and impact water quality differently
Corrosion is oxidation-reduction
Pb harr Pb2+ + 2e-
Pb harr Pb4+ + 4e-
OxidationLead metal losing electronsat anode
OCl- + H+ + 2e- harr Cl- + OH-ReductionOxidant gainingelectronsat cathode
Oxidant is dissolved oxygen free chlorine chloramine chlorine dioxide etc
2OCl- + 2H+ + 4e-harr 2Cl- + 2OH-
There are millions of anodecathode sites across interior fresh lead pipe surface
Corrosion and scale formation
8
Corrosion
Pb(s)
OCl- OH-
2 e-
Pb+2 (or Pb+4)
2 e-
Scale Formation
water
Pb pipe
Pb pipe
water
Pb amp other solids
Pb2+
Pb(+
2) c
arbo
nate
sPb
(+2)
hyd
roxy
carb
onat
esPb
(+2)
orth
opho
spha
te
harr Pb4+
Pb(+
4) o
xide
harr
Idealized scenario of scale solids Scale is way more complex (heterogeneous several layers amorphous compounds) and it controls lead release
Tria
ntaf
yllid
ou e
t al
202
0
Corrosion and metal release
bull Pure corrosion (ie electrochemical oxidation-reduction reactions) becomes less important when the surface becomes covered with scale
bull Metal solubility or destabilization of the scale at the water contact will be more important (if the scale is porous corrosion may still occur at the base of the scale)
bull Metal solubility varies by factor of 5 to 10 or more across drinking water systems
CCT includes both pure corrosion and control of metal release from the pipe scale
Corrosion typesbull Uniform corrosion
- Materials degradation- Metal release (lead copper etc)
bull Non-uniform corrosion- Pinhole leaks (copper)- Dezincification (brass)- Tuberculation (iron galvanized brass)- Galvanic Corrosion
bull Galvanic corrosion- Soldered joints- Brass devices- Partial LSLs- Any coupling of different metals
Different leaded materials may undergo different corrosion mechanisms depending on composition configuration geometry
water quality and other factors
Pb pipe uniform
Pb pipe galvanic
DeSantis et al 2018
Important factors affecting corrosion and metal release
bull pH and AlkalinityDissolved Inorganic Carbon
bull Type and Concentration of Oxidants (chlorine species dissolved oxygen etc)OxidationReduction Potential
bull Corrosion Inhibitors
bull Chloride to Sulfate Mass Ratio
bull Manganese
pH is master variable
12
-1
0
1
2
0 2 4 6 8 10 12 14
E (V
OLT
S)
pH
IMMUNE
Simplified Pourbaix diagram (EH-pH Diagram)
Desirable pH-EH combinations allow passivation (ie formation ofprotective pipe scales)
pH is master variable
12
PbO2 (plattnerite)
Pb ++
deg s)(3O 2 -2C H) 2)b O 3P O
00
( 2 -- -2) C 43 ( )
CO
b HP O( ( DIC = 18 mg CL
3 b
Pb P Pb = 0010 mgL
Pb metal
-10
13
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
ndash8
ndash6
ndash4
ndash2
2
4
6
8
10
pH
Eh (v
olts
)
Dissolved Inorganic Carbon (DIC) and Total ALKalinity (TALK)
DIC = [CO32-] + [H2CO3
] + [HCO3- ]]
TALK = 2 [CO32-] + [HCO3
-] + [OH-] - [H+]
mg CL Dissolved Inorganic Carbon (DIC)0 10 20 30 40 50
mg
CaCO
3L
Tota
l Alka
linity
0255075
100125150175
200225250
pH 60pH 70pH 80pH 90pH 100
bull To understand corrosion it isimportant to keep up with thecarbonate system
bull DIC and TALK have linearrelationship but are not thesame thing
bullbullbullbullbull
Oxidants in drinking water
bull Disinfection (Free chlorine Chloramine Chlorine dioxide)Pre-(O3 H2O2 ClO2 KMnO4) Oxidative metal removal (eg As Fe Mn) Ammonia removalAeration (corrosion control VOCRn2S removal) Taste and odor control
bull Dissolved oxygen
Oxidants in drinking water
Oxidant Dosage (mgL)0 2 4 6 8 10
E H (V
olts
vs S
HE)
03
04
05
06
07
08
09
10
11
MCA Electrode 1 MCA Electrode 2 KMnO4 Electrode 1 KMnO4 Electrode 2 ClO2 Electrode 1 ClO2 Electrode 2 DO Electrode 1 DO Electrode 2 Cl2 Electrode 1Cl2 Electrode 2
ClO2 HOCldeg
KMnO4NH2Cl
DO
Different oxidants have different oxidizing power
Oxidation-Reduction Potential of several disinfectants in experimentJames et al 2004 (pH 7 10 mg CL 25degC)
17
ldquoClassicrdquo divalent Pb+2 solubility
pH6 7 8 9 10 11
mg
PbL
001
01
1
10
100 1 mg CL 5 mg CL 10 mg CL 20 mg CL 35 mg CL 50 mg CL 75 mg CL100 mg CL
DIC
Optimum pHDIC rangefor LSLs
High pH is needed to minimize Pb solubility
Lead ldquocorrosion Inhibitorsrdquo
bull Phosphoric acidbull Alkalai-metal orthophosphatebull Zinc orthophosphatebull Blended orthopolyphosphatesbull Polyphosphatesbull Sodium Silicate
Polyphosphates are not corrosion inhibitors but rather sequestering agentsEfficiency depends on sufficient doseconcentration for thebackground pH and other reactive water quality constituentswhat if the dose is too low to actually work 18
Illustrative reactions of P in pipebull OrthophosphatePrecipitation (consumption of orthophosphate)3Pb+2 + 2PO4
-3 --gt Pb3(PO4)2 (solid)decreased Pb solubility
bull PolyphosphatesReversion partially (formation of orthophosphate)P3O10
-5 ---gt P2O7-3 + PO4
-3
polymer --gt monomersComplexation (soluble complexes)3Pb+2 + P3O10
-5 --gt PbP3O10-2
higher Pb solubility
Slid
e cr
edit
Mar
c Ed
war
ds
Conceptual possible reactions Actual reactants and products may bedifferent Polyphosphates cannot be easily quantified
20
Ortho-P Treatment for Pb+2
Effectiveness Depends on Dose DIC pH and ldquoCleanlinessrdquo of Pipe Surface
mg PO4L00 10 20 30 40 50
mg
PbL
000
005
010
015
020
025
030
035
040
48 mg CL
48 mg CL
pH = 70pH = 75pH = 80pH = 85
bull pH less critical at low DICbull pH less critical at high PO4bull Point of diminishing returns higher with high DIC
bull Faster Pb reduction at high PO4
Typical UK Dosages 4-6 mgL
Most PWSs with LSLs do not have corrosion control treatment that minimizes Pb release and exposure
Effect of pH and ortho-P on Pb release
21
DIC = 10 mg CL 1 mg PO4L
60 65 70 75 80 85 90 95
gL
b m
P
0001
001
01
1
10
US Action Level
Dissolved Pb with ortho-PTotal Pb with ortho-PDissolved Pb with CO3Total Pb with CO3
At low DIC orthophosphate improves lead release regardless of pH
pHSchock M R DeSantis M K Metz D H Welch M M Hyland R N Nadagouda M N Revisiting the pH Effect on the Orthophosphate Control of Plumbosolvency Proc AWWA Annual Conference and Exposition Atlanta GA 2008
Ortho-P at pH 90 (DIC 6 mgL)
22
Pb(μgL)
bull Low Pb at 1 mgL relative to higher doses needs to be further verified across water qualities
bull Must do dose optimization study for your own water quality especially at high pH
bull Ortho-P may precipitate with Ca
Miller S A Investigation of Lead Solubility and Orthophosphate Addition in High pH Low DIC Water Master of Science Department of Biomedical Chemical and Environmental Engineering College of Engineering and Applied Science University of Cincinnati Cincinnati OH 2014
Ortho-Ppoint of diminishing returns
bull Orthophosphate addition to where large increments result in small reductions in lead release
bull Key to cost-effective lead release control and exposure reduction
bull Becomes the rdquomaintenancerdquo dosage unless the scale ldquoagesrdquo
bull Background constituents such as aluminum hardness ions iron manganese and others interfere with optimum reduction of lead release
bull Varies with the background water chemistry from system to system
23
Sodium silicate
bull No systematic studies to look at pH carbonate silicate background chemistry relationships
bull Unclear mechanism pH increase Buffer intensity increase Protective film (thin amp amorphous)
bull Can sequester ironmanganese
bull Canrsquot be evaluated with fresh surfaces
24pH5 6 7 8 9 10 11
β ty
nsi
nte
Iuf
fer
B
00000
00005
00010TIC=48 mg CL20 mgL SiO250 mgL PO4β H2OTotal β
Silicate may contribute to Buffer Intensity
Carbonate orthophosphate silicate I=001 25ordmC
Chemical changes may reduce Pb+4 to Pb+2
10
8
6
4
)ts 2ol
h (v
E
ndash2
ndash4
ndash6
ndash8
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
pH
12
PbO2 (plattnerite)
Pb++ )s
deg
(2
CO
3
(OH
) -2 2) 3
Pb
-2
0 (CO
-- 2
0
) 43(C
O
Pb
3 (OH
)
DIC = 18 mg CL
Pb Pb Pb = 0010 mgL
Pb metal
-10
Drop in ORP from treatment change or DS
oxidant demand
Drop in pH at surface from treatment change
chemical reactions nitrification etc
C D
ngto
nhisa
W
Newark NJ
Disinfectant demand in DS must be controlled amp enough free chlorine consistently maintained throughout LSL area to keep protective Pb+4 corrosion scales in place
Switching disinfectants may reduce Pb+4 to Pb+2
Low
Switch from chlorine to chloramine disinfectant dissolved lead from pipe scales during the Washington DC ldquolead-in-water crisisrdquo in 2001-2004
Example profile of Pb(+4) (ie PbO2) scale house
27
20
L)
Cl1 Cincinnati OH
18
microg 16
n ( 14
o 12
tia 10
trn 8
e 6
nco 4
C 2
b P 0-1 0 1 2 3 4 5 6 7 8 9 10 11
Cumulative Water Volume (L)
CI-1 10 h 1209CI-1 10 h 0410
Plumbing sequence corresponding to water volumeFaucet LSLCu pipe Brass
Water MainMCu pipe 2
015
a
let
dou
illyaf
ant
irT
bull Pb release to water from PbO2 coated LSLs was found to be lowbull Be aware are there are no known cases of exhumed PbO2 coated pipes being able to
re-stabilize in pipe rigs
Chloride and sulfateCSMR a factor controlling galvanic corrosion (Oliphant 1983 Gregory 1986)- Chloride detrimental- Sulfate beneficial if sufficient to overcome chloride thus- CSMR ratio was created
[Cl- ] 12 mgL Cl-Chloride to Sulfate Mass Ratio (CSMR) = =[SO -2 2 = 06
4 ] 20 mgL SO-4
Confirmed empirically in case studies of water utilities (Dodrill and Edwards 1995 Edwards and Triantafyllidou 2007) - High CSMR gt 05-06 more corrosive- Low CSMR lt 05-06 less corrosive
28
Verified in laboratory experiments with fresh galvanic Pb junctions (eg Wang et al 2013 Triantafyllidou et al 2011 Nguyen et al 2010 2011)
Tria
ntaf
yllid
ou e
t al
201
1
Chloride and sulfate
29
CSMR considered contributing factor to the Flint water Pb crisis (Pieper et al 2017)
30
Chloride and sulfate
bull Absolute concentrations of chloride and sulfate important chloride considered particularly detrimental to galvanic Pb corrosion
bull Increasing chloride and sulfate concentrations at a constant CSMR of 1 increased total lead concentrations in galvanic corrosion experiments (Ng amp Lin 2016)
bull High CSMR ratio proven empirically important in aggravating galvanic lead corrosion (verified by experiments and modeling)
bull Water source or treatment changes that affect the CSMR must be evaluated prior to change
bull Absolute concentrations also important particularly for chloride
31
Manganese
Two distinct pipe scale categories in 5 LSLs (Lead 1-5 Pipes) excavated from same city
31
Schock et al 2014 JAWWAbull The outer pipe scale is in direct contact with the flowing water affecting
water lead contamination the mostbull Plattnerite an insoluble Pb(+4) scale that protects water from lead
contamination was inhibited by Mn (and Fe) presence
Manganese
In this case study (Schock et al JAWWA 2014)bull Where Mn was present in more abundance
- Pb scales had less PbO2- Pb associated with Mn sorbing onto Mn+Fe-rich surface deposit
bull Erratic Pb persisted after full LSLR in areas rich with Mn deposits (and to lesser extent Fe)
bull Galvanized interior plumbing also accumulated dissolved Pb while LSL present
32
33
Manganese
Other studiesbull Dong et al 2000 Mn acted as vector for particulate Pb movementbull Trueman et al 2019 Increasing Mn increased Pb in experimentbull McNeill and Edwards 2004 Mn sequestration by polyphosphate destabilized
Pb scales and caused Pb release
bull Pan et al 2019 Possible benefit by promoting formation of Pb(+4)
Reference Analysis Mn Impact on PbSchock et al 2014 Excavated LSLs
amp historical water Pb data in a US city
bull Promoted amorphous Pb scalebull Inhibited insoluble Pb(+4) scalebull Erratic Pb persisted after full LSLR in areas rich
with Mn deposits
Tully et al 2019 Excavated LSLs amp water Pb data in several Midwest case studies
bull Mn was a component of amorphous pipe scales in several systems
bull Amorphous scales not amenable to determining their adherence to pipe walls their responses to changes in various water chemistry variables or quantitatively calculating their lead release
Itrsquos not just manganese
34
Broad variety of coatings on LSLs analyzed at EPA
Al Mn Fe P and Ca-rich coatings may interfere withorthophosphate pH adjustment or other CCT
35
More water parameters affect metal release
bull pH amp AlkalinityDICbull Concentration and type of
oxidants (disinfectantsdissolved oxygen)
bull ORPcorrosion potentialbull Corrosion Inhibitorsbull Chloridebull Sulfatebull Manganese
bull Temperaturebull Sorptive surfaces downstream of
LSLs (ie galvanized interior pipe)bull Iron (deposition and corrosion)bull Calciumbull Aluminumbull NOM (type amount)bull Mixing from water treatment plants
or sourcesbull Ammoniabull Hydrogen Sulfidebull Microbial activity (nitrification amp
other)
Corrosion Control Treatment is intertwined with all treatments affecting water chemistry
Water Quality Parameters(WQP)
bull pH alkalinity calcium orthophosphate or silica othersbull WQP measurements are important for
bull Process control (measurement frequency needs to be HIGH)bull Keeping treatment from being turned offbull Keeping inhibitor dosages amp pH control from being cut to save $$$bull Keeping required repairs to treatment on tight schedulebull Preventing frequent variations and alternations of water sources
bull Effectiveness of Pb treatment only determinable from tap sampling by combination of different assessment approaches for different lead plumbing sources to which people can be exposed
36
bull OWQP measurements important for process control look at yourdata to understand what they mean and avoid problems
bull OWQP do not predict lead release There are no substitutes fordirectly monitoring lead release
37
M+
M+
M+
M+
Corrosion barrier film hypothesis
Inert barrier film (eg CaCO3)
Pipe Wall
Barrier film would allow tobull Protect ALL pipe materialsbull Monitor corrosion based on one chemical reaction
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
2
Who we are -Acknowledgments
Jennifer Tully
SimoniTriantafyllidou
Mike DeSantis
Christy Muhlen
DarrenLytle
Steve Harmon
Casey Formal
Mike Schock
Dan Williams
Regan Murray
Jonathan Burkhardt
Evelyne Doreacute
Mariah Caballero
33
What we do
ORD provides the data tools and information that form the sound scientific foundation the Agency relies on to fulfill its mission to protect the environment and safeguard public healthORD at a Glance httpswwwepagovaboutepaabout-office-research-and-development-ord Center for Environmental
Solutions amp Emergency Response (CESER) in CincinnatiWe conduct applied stakeholder-driven research and provide responsive technical support to help solve the Nationrsquos environmental challenges
CESER at a Glance httpswwwepagovaboutepaabout-center-environmental-solutions-and-emergency-response-ceser
Pb sources
BUILDING
bull Lead Service Lines (LSLs)bull Lead Goosenecksbull Leaded Solder
bull Leaded Brass (valves fittings faucets fountains)bull Galvanized Iron Pipe downstream of leaded plumbing
Triantafyllidou amp Edwards 2012
Pb sources
Full Lead Service Line
Clark et al 2013
Brass unionvalve vs plastic
Partial Lead Service Line
Lead gooseneckTriantafyllidou et al 2020
What is in the distribution system Materials inventory is critical to understand where and what lead sources still exist
Pb sources
Leaded Solder
2 Pb
lt05Pb
3 Pb
Leaded Brass FaucetSelover 2005
Gal
vani
zed
Diverse legacy leaded materials may undergo different corrosion reactions and impact water quality differently
Corrosion is oxidation-reduction
Pb harr Pb2+ + 2e-
Pb harr Pb4+ + 4e-
OxidationLead metal losing electronsat anode
OCl- + H+ + 2e- harr Cl- + OH-ReductionOxidant gainingelectronsat cathode
Oxidant is dissolved oxygen free chlorine chloramine chlorine dioxide etc
2OCl- + 2H+ + 4e-harr 2Cl- + 2OH-
There are millions of anodecathode sites across interior fresh lead pipe surface
Corrosion and scale formation
8
Corrosion
Pb(s)
OCl- OH-
2 e-
Pb+2 (or Pb+4)
2 e-
Scale Formation
water
Pb pipe
Pb pipe
water
Pb amp other solids
Pb2+
Pb(+
2) c
arbo
nate
sPb
(+2)
hyd
roxy
carb
onat
esPb
(+2)
orth
opho
spha
te
harr Pb4+
Pb(+
4) o
xide
harr
Idealized scenario of scale solids Scale is way more complex (heterogeneous several layers amorphous compounds) and it controls lead release
Tria
ntaf
yllid
ou e
t al
202
0
Corrosion and metal release
bull Pure corrosion (ie electrochemical oxidation-reduction reactions) becomes less important when the surface becomes covered with scale
bull Metal solubility or destabilization of the scale at the water contact will be more important (if the scale is porous corrosion may still occur at the base of the scale)
bull Metal solubility varies by factor of 5 to 10 or more across drinking water systems
CCT includes both pure corrosion and control of metal release from the pipe scale
Corrosion typesbull Uniform corrosion
- Materials degradation- Metal release (lead copper etc)
bull Non-uniform corrosion- Pinhole leaks (copper)- Dezincification (brass)- Tuberculation (iron galvanized brass)- Galvanic Corrosion
bull Galvanic corrosion- Soldered joints- Brass devices- Partial LSLs- Any coupling of different metals
Different leaded materials may undergo different corrosion mechanisms depending on composition configuration geometry
water quality and other factors
Pb pipe uniform
Pb pipe galvanic
DeSantis et al 2018
Important factors affecting corrosion and metal release
bull pH and AlkalinityDissolved Inorganic Carbon
bull Type and Concentration of Oxidants (chlorine species dissolved oxygen etc)OxidationReduction Potential
bull Corrosion Inhibitors
bull Chloride to Sulfate Mass Ratio
bull Manganese
pH is master variable
12
-1
0
1
2
0 2 4 6 8 10 12 14
E (V
OLT
S)
pH
IMMUNE
Simplified Pourbaix diagram (EH-pH Diagram)
Desirable pH-EH combinations allow passivation (ie formation ofprotective pipe scales)
pH is master variable
12
PbO2 (plattnerite)
Pb ++
deg s)(3O 2 -2C H) 2)b O 3P O
00
( 2 -- -2) C 43 ( )
CO
b HP O( ( DIC = 18 mg CL
3 b
Pb P Pb = 0010 mgL
Pb metal
-10
13
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
ndash8
ndash6
ndash4
ndash2
2
4
6
8
10
pH
Eh (v
olts
)
Dissolved Inorganic Carbon (DIC) and Total ALKalinity (TALK)
DIC = [CO32-] + [H2CO3
] + [HCO3- ]]
TALK = 2 [CO32-] + [HCO3
-] + [OH-] - [H+]
mg CL Dissolved Inorganic Carbon (DIC)0 10 20 30 40 50
mg
CaCO
3L
Tota
l Alka
linity
0255075
100125150175
200225250
pH 60pH 70pH 80pH 90pH 100
bull To understand corrosion it isimportant to keep up with thecarbonate system
bull DIC and TALK have linearrelationship but are not thesame thing
bullbullbullbullbull
Oxidants in drinking water
bull Disinfection (Free chlorine Chloramine Chlorine dioxide)Pre-(O3 H2O2 ClO2 KMnO4) Oxidative metal removal (eg As Fe Mn) Ammonia removalAeration (corrosion control VOCRn2S removal) Taste and odor control
bull Dissolved oxygen
Oxidants in drinking water
Oxidant Dosage (mgL)0 2 4 6 8 10
E H (V
olts
vs S
HE)
03
04
05
06
07
08
09
10
11
MCA Electrode 1 MCA Electrode 2 KMnO4 Electrode 1 KMnO4 Electrode 2 ClO2 Electrode 1 ClO2 Electrode 2 DO Electrode 1 DO Electrode 2 Cl2 Electrode 1Cl2 Electrode 2
ClO2 HOCldeg
KMnO4NH2Cl
DO
Different oxidants have different oxidizing power
Oxidation-Reduction Potential of several disinfectants in experimentJames et al 2004 (pH 7 10 mg CL 25degC)
17
ldquoClassicrdquo divalent Pb+2 solubility
pH6 7 8 9 10 11
mg
PbL
001
01
1
10
100 1 mg CL 5 mg CL 10 mg CL 20 mg CL 35 mg CL 50 mg CL 75 mg CL100 mg CL
DIC
Optimum pHDIC rangefor LSLs
High pH is needed to minimize Pb solubility
Lead ldquocorrosion Inhibitorsrdquo
bull Phosphoric acidbull Alkalai-metal orthophosphatebull Zinc orthophosphatebull Blended orthopolyphosphatesbull Polyphosphatesbull Sodium Silicate
Polyphosphates are not corrosion inhibitors but rather sequestering agentsEfficiency depends on sufficient doseconcentration for thebackground pH and other reactive water quality constituentswhat if the dose is too low to actually work 18
Illustrative reactions of P in pipebull OrthophosphatePrecipitation (consumption of orthophosphate)3Pb+2 + 2PO4
-3 --gt Pb3(PO4)2 (solid)decreased Pb solubility
bull PolyphosphatesReversion partially (formation of orthophosphate)P3O10
-5 ---gt P2O7-3 + PO4
-3
polymer --gt monomersComplexation (soluble complexes)3Pb+2 + P3O10
-5 --gt PbP3O10-2
higher Pb solubility
Slid
e cr
edit
Mar
c Ed
war
ds
Conceptual possible reactions Actual reactants and products may bedifferent Polyphosphates cannot be easily quantified
20
Ortho-P Treatment for Pb+2
Effectiveness Depends on Dose DIC pH and ldquoCleanlinessrdquo of Pipe Surface
mg PO4L00 10 20 30 40 50
mg
PbL
000
005
010
015
020
025
030
035
040
48 mg CL
48 mg CL
pH = 70pH = 75pH = 80pH = 85
bull pH less critical at low DICbull pH less critical at high PO4bull Point of diminishing returns higher with high DIC
bull Faster Pb reduction at high PO4
Typical UK Dosages 4-6 mgL
Most PWSs with LSLs do not have corrosion control treatment that minimizes Pb release and exposure
Effect of pH and ortho-P on Pb release
21
DIC = 10 mg CL 1 mg PO4L
60 65 70 75 80 85 90 95
gL
b m
P
0001
001
01
1
10
US Action Level
Dissolved Pb with ortho-PTotal Pb with ortho-PDissolved Pb with CO3Total Pb with CO3
At low DIC orthophosphate improves lead release regardless of pH
pHSchock M R DeSantis M K Metz D H Welch M M Hyland R N Nadagouda M N Revisiting the pH Effect on the Orthophosphate Control of Plumbosolvency Proc AWWA Annual Conference and Exposition Atlanta GA 2008
Ortho-P at pH 90 (DIC 6 mgL)
22
Pb(μgL)
bull Low Pb at 1 mgL relative to higher doses needs to be further verified across water qualities
bull Must do dose optimization study for your own water quality especially at high pH
bull Ortho-P may precipitate with Ca
Miller S A Investigation of Lead Solubility and Orthophosphate Addition in High pH Low DIC Water Master of Science Department of Biomedical Chemical and Environmental Engineering College of Engineering and Applied Science University of Cincinnati Cincinnati OH 2014
Ortho-Ppoint of diminishing returns
bull Orthophosphate addition to where large increments result in small reductions in lead release
bull Key to cost-effective lead release control and exposure reduction
bull Becomes the rdquomaintenancerdquo dosage unless the scale ldquoagesrdquo
bull Background constituents such as aluminum hardness ions iron manganese and others interfere with optimum reduction of lead release
bull Varies with the background water chemistry from system to system
23
Sodium silicate
bull No systematic studies to look at pH carbonate silicate background chemistry relationships
bull Unclear mechanism pH increase Buffer intensity increase Protective film (thin amp amorphous)
bull Can sequester ironmanganese
bull Canrsquot be evaluated with fresh surfaces
24pH5 6 7 8 9 10 11
β ty
nsi
nte
Iuf
fer
B
00000
00005
00010TIC=48 mg CL20 mgL SiO250 mgL PO4β H2OTotal β
Silicate may contribute to Buffer Intensity
Carbonate orthophosphate silicate I=001 25ordmC
Chemical changes may reduce Pb+4 to Pb+2
10
8
6
4
)ts 2ol
h (v
E
ndash2
ndash4
ndash6
ndash8
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
pH
12
PbO2 (plattnerite)
Pb++ )s
deg
(2
CO
3
(OH
) -2 2) 3
Pb
-2
0 (CO
-- 2
0
) 43(C
O
Pb
3 (OH
)
DIC = 18 mg CL
Pb Pb Pb = 0010 mgL
Pb metal
-10
Drop in ORP from treatment change or DS
oxidant demand
Drop in pH at surface from treatment change
chemical reactions nitrification etc
C D
ngto
nhisa
W
Newark NJ
Disinfectant demand in DS must be controlled amp enough free chlorine consistently maintained throughout LSL area to keep protective Pb+4 corrosion scales in place
Switching disinfectants may reduce Pb+4 to Pb+2
Low
Switch from chlorine to chloramine disinfectant dissolved lead from pipe scales during the Washington DC ldquolead-in-water crisisrdquo in 2001-2004
Example profile of Pb(+4) (ie PbO2) scale house
27
20
L)
Cl1 Cincinnati OH
18
microg 16
n ( 14
o 12
tia 10
trn 8
e 6
nco 4
C 2
b P 0-1 0 1 2 3 4 5 6 7 8 9 10 11
Cumulative Water Volume (L)
CI-1 10 h 1209CI-1 10 h 0410
Plumbing sequence corresponding to water volumeFaucet LSLCu pipe Brass
Water MainMCu pipe 2
015
a
let
dou
illyaf
ant
irT
bull Pb release to water from PbO2 coated LSLs was found to be lowbull Be aware are there are no known cases of exhumed PbO2 coated pipes being able to
re-stabilize in pipe rigs
Chloride and sulfateCSMR a factor controlling galvanic corrosion (Oliphant 1983 Gregory 1986)- Chloride detrimental- Sulfate beneficial if sufficient to overcome chloride thus- CSMR ratio was created
[Cl- ] 12 mgL Cl-Chloride to Sulfate Mass Ratio (CSMR) = =[SO -2 2 = 06
4 ] 20 mgL SO-4
Confirmed empirically in case studies of water utilities (Dodrill and Edwards 1995 Edwards and Triantafyllidou 2007) - High CSMR gt 05-06 more corrosive- Low CSMR lt 05-06 less corrosive
28
Verified in laboratory experiments with fresh galvanic Pb junctions (eg Wang et al 2013 Triantafyllidou et al 2011 Nguyen et al 2010 2011)
Tria
ntaf
yllid
ou e
t al
201
1
Chloride and sulfate
29
CSMR considered contributing factor to the Flint water Pb crisis (Pieper et al 2017)
30
Chloride and sulfate
bull Absolute concentrations of chloride and sulfate important chloride considered particularly detrimental to galvanic Pb corrosion
bull Increasing chloride and sulfate concentrations at a constant CSMR of 1 increased total lead concentrations in galvanic corrosion experiments (Ng amp Lin 2016)
bull High CSMR ratio proven empirically important in aggravating galvanic lead corrosion (verified by experiments and modeling)
bull Water source or treatment changes that affect the CSMR must be evaluated prior to change
bull Absolute concentrations also important particularly for chloride
31
Manganese
Two distinct pipe scale categories in 5 LSLs (Lead 1-5 Pipes) excavated from same city
31
Schock et al 2014 JAWWAbull The outer pipe scale is in direct contact with the flowing water affecting
water lead contamination the mostbull Plattnerite an insoluble Pb(+4) scale that protects water from lead
contamination was inhibited by Mn (and Fe) presence
Manganese
In this case study (Schock et al JAWWA 2014)bull Where Mn was present in more abundance
- Pb scales had less PbO2- Pb associated with Mn sorbing onto Mn+Fe-rich surface deposit
bull Erratic Pb persisted after full LSLR in areas rich with Mn deposits (and to lesser extent Fe)
bull Galvanized interior plumbing also accumulated dissolved Pb while LSL present
32
33
Manganese
Other studiesbull Dong et al 2000 Mn acted as vector for particulate Pb movementbull Trueman et al 2019 Increasing Mn increased Pb in experimentbull McNeill and Edwards 2004 Mn sequestration by polyphosphate destabilized
Pb scales and caused Pb release
bull Pan et al 2019 Possible benefit by promoting formation of Pb(+4)
Reference Analysis Mn Impact on PbSchock et al 2014 Excavated LSLs
amp historical water Pb data in a US city
bull Promoted amorphous Pb scalebull Inhibited insoluble Pb(+4) scalebull Erratic Pb persisted after full LSLR in areas rich
with Mn deposits
Tully et al 2019 Excavated LSLs amp water Pb data in several Midwest case studies
bull Mn was a component of amorphous pipe scales in several systems
bull Amorphous scales not amenable to determining their adherence to pipe walls their responses to changes in various water chemistry variables or quantitatively calculating their lead release
Itrsquos not just manganese
34
Broad variety of coatings on LSLs analyzed at EPA
Al Mn Fe P and Ca-rich coatings may interfere withorthophosphate pH adjustment or other CCT
35
More water parameters affect metal release
bull pH amp AlkalinityDICbull Concentration and type of
oxidants (disinfectantsdissolved oxygen)
bull ORPcorrosion potentialbull Corrosion Inhibitorsbull Chloridebull Sulfatebull Manganese
bull Temperaturebull Sorptive surfaces downstream of
LSLs (ie galvanized interior pipe)bull Iron (deposition and corrosion)bull Calciumbull Aluminumbull NOM (type amount)bull Mixing from water treatment plants
or sourcesbull Ammoniabull Hydrogen Sulfidebull Microbial activity (nitrification amp
other)
Corrosion Control Treatment is intertwined with all treatments affecting water chemistry
Water Quality Parameters(WQP)
bull pH alkalinity calcium orthophosphate or silica othersbull WQP measurements are important for
bull Process control (measurement frequency needs to be HIGH)bull Keeping treatment from being turned offbull Keeping inhibitor dosages amp pH control from being cut to save $$$bull Keeping required repairs to treatment on tight schedulebull Preventing frequent variations and alternations of water sources
bull Effectiveness of Pb treatment only determinable from tap sampling by combination of different assessment approaches for different lead plumbing sources to which people can be exposed
36
bull OWQP measurements important for process control look at yourdata to understand what they mean and avoid problems
bull OWQP do not predict lead release There are no substitutes fordirectly monitoring lead release
37
M+
M+
M+
M+
Corrosion barrier film hypothesis
Inert barrier film (eg CaCO3)
Pipe Wall
Barrier film would allow tobull Protect ALL pipe materialsbull Monitor corrosion based on one chemical reaction
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
33
What we do
ORD provides the data tools and information that form the sound scientific foundation the Agency relies on to fulfill its mission to protect the environment and safeguard public healthORD at a Glance httpswwwepagovaboutepaabout-office-research-and-development-ord Center for Environmental
Solutions amp Emergency Response (CESER) in CincinnatiWe conduct applied stakeholder-driven research and provide responsive technical support to help solve the Nationrsquos environmental challenges
CESER at a Glance httpswwwepagovaboutepaabout-center-environmental-solutions-and-emergency-response-ceser
Pb sources
BUILDING
bull Lead Service Lines (LSLs)bull Lead Goosenecksbull Leaded Solder
bull Leaded Brass (valves fittings faucets fountains)bull Galvanized Iron Pipe downstream of leaded plumbing
Triantafyllidou amp Edwards 2012
Pb sources
Full Lead Service Line
Clark et al 2013
Brass unionvalve vs plastic
Partial Lead Service Line
Lead gooseneckTriantafyllidou et al 2020
What is in the distribution system Materials inventory is critical to understand where and what lead sources still exist
Pb sources
Leaded Solder
2 Pb
lt05Pb
3 Pb
Leaded Brass FaucetSelover 2005
Gal
vani
zed
Diverse legacy leaded materials may undergo different corrosion reactions and impact water quality differently
Corrosion is oxidation-reduction
Pb harr Pb2+ + 2e-
Pb harr Pb4+ + 4e-
OxidationLead metal losing electronsat anode
OCl- + H+ + 2e- harr Cl- + OH-ReductionOxidant gainingelectronsat cathode
Oxidant is dissolved oxygen free chlorine chloramine chlorine dioxide etc
2OCl- + 2H+ + 4e-harr 2Cl- + 2OH-
There are millions of anodecathode sites across interior fresh lead pipe surface
Corrosion and scale formation
8
Corrosion
Pb(s)
OCl- OH-
2 e-
Pb+2 (or Pb+4)
2 e-
Scale Formation
water
Pb pipe
Pb pipe
water
Pb amp other solids
Pb2+
Pb(+
2) c
arbo
nate
sPb
(+2)
hyd
roxy
carb
onat
esPb
(+2)
orth
opho
spha
te
harr Pb4+
Pb(+
4) o
xide
harr
Idealized scenario of scale solids Scale is way more complex (heterogeneous several layers amorphous compounds) and it controls lead release
Tria
ntaf
yllid
ou e
t al
202
0
Corrosion and metal release
bull Pure corrosion (ie electrochemical oxidation-reduction reactions) becomes less important when the surface becomes covered with scale
bull Metal solubility or destabilization of the scale at the water contact will be more important (if the scale is porous corrosion may still occur at the base of the scale)
bull Metal solubility varies by factor of 5 to 10 or more across drinking water systems
CCT includes both pure corrosion and control of metal release from the pipe scale
Corrosion typesbull Uniform corrosion
- Materials degradation- Metal release (lead copper etc)
bull Non-uniform corrosion- Pinhole leaks (copper)- Dezincification (brass)- Tuberculation (iron galvanized brass)- Galvanic Corrosion
bull Galvanic corrosion- Soldered joints- Brass devices- Partial LSLs- Any coupling of different metals
Different leaded materials may undergo different corrosion mechanisms depending on composition configuration geometry
water quality and other factors
Pb pipe uniform
Pb pipe galvanic
DeSantis et al 2018
Important factors affecting corrosion and metal release
bull pH and AlkalinityDissolved Inorganic Carbon
bull Type and Concentration of Oxidants (chlorine species dissolved oxygen etc)OxidationReduction Potential
bull Corrosion Inhibitors
bull Chloride to Sulfate Mass Ratio
bull Manganese
pH is master variable
12
-1
0
1
2
0 2 4 6 8 10 12 14
E (V
OLT
S)
pH
IMMUNE
Simplified Pourbaix diagram (EH-pH Diagram)
Desirable pH-EH combinations allow passivation (ie formation ofprotective pipe scales)
pH is master variable
12
PbO2 (plattnerite)
Pb ++
deg s)(3O 2 -2C H) 2)b O 3P O
00
( 2 -- -2) C 43 ( )
CO
b HP O( ( DIC = 18 mg CL
3 b
Pb P Pb = 0010 mgL
Pb metal
-10
13
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
ndash8
ndash6
ndash4
ndash2
2
4
6
8
10
pH
Eh (v
olts
)
Dissolved Inorganic Carbon (DIC) and Total ALKalinity (TALK)
DIC = [CO32-] + [H2CO3
] + [HCO3- ]]
TALK = 2 [CO32-] + [HCO3
-] + [OH-] - [H+]
mg CL Dissolved Inorganic Carbon (DIC)0 10 20 30 40 50
mg
CaCO
3L
Tota
l Alka
linity
0255075
100125150175
200225250
pH 60pH 70pH 80pH 90pH 100
bull To understand corrosion it isimportant to keep up with thecarbonate system
bull DIC and TALK have linearrelationship but are not thesame thing
bullbullbullbullbull
Oxidants in drinking water
bull Disinfection (Free chlorine Chloramine Chlorine dioxide)Pre-(O3 H2O2 ClO2 KMnO4) Oxidative metal removal (eg As Fe Mn) Ammonia removalAeration (corrosion control VOCRn2S removal) Taste and odor control
bull Dissolved oxygen
Oxidants in drinking water
Oxidant Dosage (mgL)0 2 4 6 8 10
E H (V
olts
vs S
HE)
03
04
05
06
07
08
09
10
11
MCA Electrode 1 MCA Electrode 2 KMnO4 Electrode 1 KMnO4 Electrode 2 ClO2 Electrode 1 ClO2 Electrode 2 DO Electrode 1 DO Electrode 2 Cl2 Electrode 1Cl2 Electrode 2
ClO2 HOCldeg
KMnO4NH2Cl
DO
Different oxidants have different oxidizing power
Oxidation-Reduction Potential of several disinfectants in experimentJames et al 2004 (pH 7 10 mg CL 25degC)
17
ldquoClassicrdquo divalent Pb+2 solubility
pH6 7 8 9 10 11
mg
PbL
001
01
1
10
100 1 mg CL 5 mg CL 10 mg CL 20 mg CL 35 mg CL 50 mg CL 75 mg CL100 mg CL
DIC
Optimum pHDIC rangefor LSLs
High pH is needed to minimize Pb solubility
Lead ldquocorrosion Inhibitorsrdquo
bull Phosphoric acidbull Alkalai-metal orthophosphatebull Zinc orthophosphatebull Blended orthopolyphosphatesbull Polyphosphatesbull Sodium Silicate
Polyphosphates are not corrosion inhibitors but rather sequestering agentsEfficiency depends on sufficient doseconcentration for thebackground pH and other reactive water quality constituentswhat if the dose is too low to actually work 18
Illustrative reactions of P in pipebull OrthophosphatePrecipitation (consumption of orthophosphate)3Pb+2 + 2PO4
-3 --gt Pb3(PO4)2 (solid)decreased Pb solubility
bull PolyphosphatesReversion partially (formation of orthophosphate)P3O10
-5 ---gt P2O7-3 + PO4
-3
polymer --gt monomersComplexation (soluble complexes)3Pb+2 + P3O10
-5 --gt PbP3O10-2
higher Pb solubility
Slid
e cr
edit
Mar
c Ed
war
ds
Conceptual possible reactions Actual reactants and products may bedifferent Polyphosphates cannot be easily quantified
20
Ortho-P Treatment for Pb+2
Effectiveness Depends on Dose DIC pH and ldquoCleanlinessrdquo of Pipe Surface
mg PO4L00 10 20 30 40 50
mg
PbL
000
005
010
015
020
025
030
035
040
48 mg CL
48 mg CL
pH = 70pH = 75pH = 80pH = 85
bull pH less critical at low DICbull pH less critical at high PO4bull Point of diminishing returns higher with high DIC
bull Faster Pb reduction at high PO4
Typical UK Dosages 4-6 mgL
Most PWSs with LSLs do not have corrosion control treatment that minimizes Pb release and exposure
Effect of pH and ortho-P on Pb release
21
DIC = 10 mg CL 1 mg PO4L
60 65 70 75 80 85 90 95
gL
b m
P
0001
001
01
1
10
US Action Level
Dissolved Pb with ortho-PTotal Pb with ortho-PDissolved Pb with CO3Total Pb with CO3
At low DIC orthophosphate improves lead release regardless of pH
pHSchock M R DeSantis M K Metz D H Welch M M Hyland R N Nadagouda M N Revisiting the pH Effect on the Orthophosphate Control of Plumbosolvency Proc AWWA Annual Conference and Exposition Atlanta GA 2008
Ortho-P at pH 90 (DIC 6 mgL)
22
Pb(μgL)
bull Low Pb at 1 mgL relative to higher doses needs to be further verified across water qualities
bull Must do dose optimization study for your own water quality especially at high pH
bull Ortho-P may precipitate with Ca
Miller S A Investigation of Lead Solubility and Orthophosphate Addition in High pH Low DIC Water Master of Science Department of Biomedical Chemical and Environmental Engineering College of Engineering and Applied Science University of Cincinnati Cincinnati OH 2014
Ortho-Ppoint of diminishing returns
bull Orthophosphate addition to where large increments result in small reductions in lead release
bull Key to cost-effective lead release control and exposure reduction
bull Becomes the rdquomaintenancerdquo dosage unless the scale ldquoagesrdquo
bull Background constituents such as aluminum hardness ions iron manganese and others interfere with optimum reduction of lead release
bull Varies with the background water chemistry from system to system
23
Sodium silicate
bull No systematic studies to look at pH carbonate silicate background chemistry relationships
bull Unclear mechanism pH increase Buffer intensity increase Protective film (thin amp amorphous)
bull Can sequester ironmanganese
bull Canrsquot be evaluated with fresh surfaces
24pH5 6 7 8 9 10 11
β ty
nsi
nte
Iuf
fer
B
00000
00005
00010TIC=48 mg CL20 mgL SiO250 mgL PO4β H2OTotal β
Silicate may contribute to Buffer Intensity
Carbonate orthophosphate silicate I=001 25ordmC
Chemical changes may reduce Pb+4 to Pb+2
10
8
6
4
)ts 2ol
h (v
E
ndash2
ndash4
ndash6
ndash8
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
pH
12
PbO2 (plattnerite)
Pb++ )s
deg
(2
CO
3
(OH
) -2 2) 3
Pb
-2
0 (CO
-- 2
0
) 43(C
O
Pb
3 (OH
)
DIC = 18 mg CL
Pb Pb Pb = 0010 mgL
Pb metal
-10
Drop in ORP from treatment change or DS
oxidant demand
Drop in pH at surface from treatment change
chemical reactions nitrification etc
C D
ngto
nhisa
W
Newark NJ
Disinfectant demand in DS must be controlled amp enough free chlorine consistently maintained throughout LSL area to keep protective Pb+4 corrosion scales in place
Switching disinfectants may reduce Pb+4 to Pb+2
Low
Switch from chlorine to chloramine disinfectant dissolved lead from pipe scales during the Washington DC ldquolead-in-water crisisrdquo in 2001-2004
Example profile of Pb(+4) (ie PbO2) scale house
27
20
L)
Cl1 Cincinnati OH
18
microg 16
n ( 14
o 12
tia 10
trn 8
e 6
nco 4
C 2
b P 0-1 0 1 2 3 4 5 6 7 8 9 10 11
Cumulative Water Volume (L)
CI-1 10 h 1209CI-1 10 h 0410
Plumbing sequence corresponding to water volumeFaucet LSLCu pipe Brass
Water MainMCu pipe 2
015
a
let
dou
illyaf
ant
irT
bull Pb release to water from PbO2 coated LSLs was found to be lowbull Be aware are there are no known cases of exhumed PbO2 coated pipes being able to
re-stabilize in pipe rigs
Chloride and sulfateCSMR a factor controlling galvanic corrosion (Oliphant 1983 Gregory 1986)- Chloride detrimental- Sulfate beneficial if sufficient to overcome chloride thus- CSMR ratio was created
[Cl- ] 12 mgL Cl-Chloride to Sulfate Mass Ratio (CSMR) = =[SO -2 2 = 06
4 ] 20 mgL SO-4
Confirmed empirically in case studies of water utilities (Dodrill and Edwards 1995 Edwards and Triantafyllidou 2007) - High CSMR gt 05-06 more corrosive- Low CSMR lt 05-06 less corrosive
28
Verified in laboratory experiments with fresh galvanic Pb junctions (eg Wang et al 2013 Triantafyllidou et al 2011 Nguyen et al 2010 2011)
Tria
ntaf
yllid
ou e
t al
201
1
Chloride and sulfate
29
CSMR considered contributing factor to the Flint water Pb crisis (Pieper et al 2017)
30
Chloride and sulfate
bull Absolute concentrations of chloride and sulfate important chloride considered particularly detrimental to galvanic Pb corrosion
bull Increasing chloride and sulfate concentrations at a constant CSMR of 1 increased total lead concentrations in galvanic corrosion experiments (Ng amp Lin 2016)
bull High CSMR ratio proven empirically important in aggravating galvanic lead corrosion (verified by experiments and modeling)
bull Water source or treatment changes that affect the CSMR must be evaluated prior to change
bull Absolute concentrations also important particularly for chloride
31
Manganese
Two distinct pipe scale categories in 5 LSLs (Lead 1-5 Pipes) excavated from same city
31
Schock et al 2014 JAWWAbull The outer pipe scale is in direct contact with the flowing water affecting
water lead contamination the mostbull Plattnerite an insoluble Pb(+4) scale that protects water from lead
contamination was inhibited by Mn (and Fe) presence
Manganese
In this case study (Schock et al JAWWA 2014)bull Where Mn was present in more abundance
- Pb scales had less PbO2- Pb associated with Mn sorbing onto Mn+Fe-rich surface deposit
bull Erratic Pb persisted after full LSLR in areas rich with Mn deposits (and to lesser extent Fe)
bull Galvanized interior plumbing also accumulated dissolved Pb while LSL present
32
33
Manganese
Other studiesbull Dong et al 2000 Mn acted as vector for particulate Pb movementbull Trueman et al 2019 Increasing Mn increased Pb in experimentbull McNeill and Edwards 2004 Mn sequestration by polyphosphate destabilized
Pb scales and caused Pb release
bull Pan et al 2019 Possible benefit by promoting formation of Pb(+4)
Reference Analysis Mn Impact on PbSchock et al 2014 Excavated LSLs
amp historical water Pb data in a US city
bull Promoted amorphous Pb scalebull Inhibited insoluble Pb(+4) scalebull Erratic Pb persisted after full LSLR in areas rich
with Mn deposits
Tully et al 2019 Excavated LSLs amp water Pb data in several Midwest case studies
bull Mn was a component of amorphous pipe scales in several systems
bull Amorphous scales not amenable to determining their adherence to pipe walls their responses to changes in various water chemistry variables or quantitatively calculating their lead release
Itrsquos not just manganese
34
Broad variety of coatings on LSLs analyzed at EPA
Al Mn Fe P and Ca-rich coatings may interfere withorthophosphate pH adjustment or other CCT
35
More water parameters affect metal release
bull pH amp AlkalinityDICbull Concentration and type of
oxidants (disinfectantsdissolved oxygen)
bull ORPcorrosion potentialbull Corrosion Inhibitorsbull Chloridebull Sulfatebull Manganese
bull Temperaturebull Sorptive surfaces downstream of
LSLs (ie galvanized interior pipe)bull Iron (deposition and corrosion)bull Calciumbull Aluminumbull NOM (type amount)bull Mixing from water treatment plants
or sourcesbull Ammoniabull Hydrogen Sulfidebull Microbial activity (nitrification amp
other)
Corrosion Control Treatment is intertwined with all treatments affecting water chemistry
Water Quality Parameters(WQP)
bull pH alkalinity calcium orthophosphate or silica othersbull WQP measurements are important for
bull Process control (measurement frequency needs to be HIGH)bull Keeping treatment from being turned offbull Keeping inhibitor dosages amp pH control from being cut to save $$$bull Keeping required repairs to treatment on tight schedulebull Preventing frequent variations and alternations of water sources
bull Effectiveness of Pb treatment only determinable from tap sampling by combination of different assessment approaches for different lead plumbing sources to which people can be exposed
36
bull OWQP measurements important for process control look at yourdata to understand what they mean and avoid problems
bull OWQP do not predict lead release There are no substitutes fordirectly monitoring lead release
37
M+
M+
M+
M+
Corrosion barrier film hypothesis
Inert barrier film (eg CaCO3)
Pipe Wall
Barrier film would allow tobull Protect ALL pipe materialsbull Monitor corrosion based on one chemical reaction
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
Pb sources
BUILDING
bull Lead Service Lines (LSLs)bull Lead Goosenecksbull Leaded Solder
bull Leaded Brass (valves fittings faucets fountains)bull Galvanized Iron Pipe downstream of leaded plumbing
Triantafyllidou amp Edwards 2012
Pb sources
Full Lead Service Line
Clark et al 2013
Brass unionvalve vs plastic
Partial Lead Service Line
Lead gooseneckTriantafyllidou et al 2020
What is in the distribution system Materials inventory is critical to understand where and what lead sources still exist
Pb sources
Leaded Solder
2 Pb
lt05Pb
3 Pb
Leaded Brass FaucetSelover 2005
Gal
vani
zed
Diverse legacy leaded materials may undergo different corrosion reactions and impact water quality differently
Corrosion is oxidation-reduction
Pb harr Pb2+ + 2e-
Pb harr Pb4+ + 4e-
OxidationLead metal losing electronsat anode
OCl- + H+ + 2e- harr Cl- + OH-ReductionOxidant gainingelectronsat cathode
Oxidant is dissolved oxygen free chlorine chloramine chlorine dioxide etc
2OCl- + 2H+ + 4e-harr 2Cl- + 2OH-
There are millions of anodecathode sites across interior fresh lead pipe surface
Corrosion and scale formation
8
Corrosion
Pb(s)
OCl- OH-
2 e-
Pb+2 (or Pb+4)
2 e-
Scale Formation
water
Pb pipe
Pb pipe
water
Pb amp other solids
Pb2+
Pb(+
2) c
arbo
nate
sPb
(+2)
hyd
roxy
carb
onat
esPb
(+2)
orth
opho
spha
te
harr Pb4+
Pb(+
4) o
xide
harr
Idealized scenario of scale solids Scale is way more complex (heterogeneous several layers amorphous compounds) and it controls lead release
Tria
ntaf
yllid
ou e
t al
202
0
Corrosion and metal release
bull Pure corrosion (ie electrochemical oxidation-reduction reactions) becomes less important when the surface becomes covered with scale
bull Metal solubility or destabilization of the scale at the water contact will be more important (if the scale is porous corrosion may still occur at the base of the scale)
bull Metal solubility varies by factor of 5 to 10 or more across drinking water systems
CCT includes both pure corrosion and control of metal release from the pipe scale
Corrosion typesbull Uniform corrosion
- Materials degradation- Metal release (lead copper etc)
bull Non-uniform corrosion- Pinhole leaks (copper)- Dezincification (brass)- Tuberculation (iron galvanized brass)- Galvanic Corrosion
bull Galvanic corrosion- Soldered joints- Brass devices- Partial LSLs- Any coupling of different metals
Different leaded materials may undergo different corrosion mechanisms depending on composition configuration geometry
water quality and other factors
Pb pipe uniform
Pb pipe galvanic
DeSantis et al 2018
Important factors affecting corrosion and metal release
bull pH and AlkalinityDissolved Inorganic Carbon
bull Type and Concentration of Oxidants (chlorine species dissolved oxygen etc)OxidationReduction Potential
bull Corrosion Inhibitors
bull Chloride to Sulfate Mass Ratio
bull Manganese
pH is master variable
12
-1
0
1
2
0 2 4 6 8 10 12 14
E (V
OLT
S)
pH
IMMUNE
Simplified Pourbaix diagram (EH-pH Diagram)
Desirable pH-EH combinations allow passivation (ie formation ofprotective pipe scales)
pH is master variable
12
PbO2 (plattnerite)
Pb ++
deg s)(3O 2 -2C H) 2)b O 3P O
00
( 2 -- -2) C 43 ( )
CO
b HP O( ( DIC = 18 mg CL
3 b
Pb P Pb = 0010 mgL
Pb metal
-10
13
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
ndash8
ndash6
ndash4
ndash2
2
4
6
8
10
pH
Eh (v
olts
)
Dissolved Inorganic Carbon (DIC) and Total ALKalinity (TALK)
DIC = [CO32-] + [H2CO3
] + [HCO3- ]]
TALK = 2 [CO32-] + [HCO3
-] + [OH-] - [H+]
mg CL Dissolved Inorganic Carbon (DIC)0 10 20 30 40 50
mg
CaCO
3L
Tota
l Alka
linity
0255075
100125150175
200225250
pH 60pH 70pH 80pH 90pH 100
bull To understand corrosion it isimportant to keep up with thecarbonate system
bull DIC and TALK have linearrelationship but are not thesame thing
bullbullbullbullbull
Oxidants in drinking water
bull Disinfection (Free chlorine Chloramine Chlorine dioxide)Pre-(O3 H2O2 ClO2 KMnO4) Oxidative metal removal (eg As Fe Mn) Ammonia removalAeration (corrosion control VOCRn2S removal) Taste and odor control
bull Dissolved oxygen
Oxidants in drinking water
Oxidant Dosage (mgL)0 2 4 6 8 10
E H (V
olts
vs S
HE)
03
04
05
06
07
08
09
10
11
MCA Electrode 1 MCA Electrode 2 KMnO4 Electrode 1 KMnO4 Electrode 2 ClO2 Electrode 1 ClO2 Electrode 2 DO Electrode 1 DO Electrode 2 Cl2 Electrode 1Cl2 Electrode 2
ClO2 HOCldeg
KMnO4NH2Cl
DO
Different oxidants have different oxidizing power
Oxidation-Reduction Potential of several disinfectants in experimentJames et al 2004 (pH 7 10 mg CL 25degC)
17
ldquoClassicrdquo divalent Pb+2 solubility
pH6 7 8 9 10 11
mg
PbL
001
01
1
10
100 1 mg CL 5 mg CL 10 mg CL 20 mg CL 35 mg CL 50 mg CL 75 mg CL100 mg CL
DIC
Optimum pHDIC rangefor LSLs
High pH is needed to minimize Pb solubility
Lead ldquocorrosion Inhibitorsrdquo
bull Phosphoric acidbull Alkalai-metal orthophosphatebull Zinc orthophosphatebull Blended orthopolyphosphatesbull Polyphosphatesbull Sodium Silicate
Polyphosphates are not corrosion inhibitors but rather sequestering agentsEfficiency depends on sufficient doseconcentration for thebackground pH and other reactive water quality constituentswhat if the dose is too low to actually work 18
Illustrative reactions of P in pipebull OrthophosphatePrecipitation (consumption of orthophosphate)3Pb+2 + 2PO4
-3 --gt Pb3(PO4)2 (solid)decreased Pb solubility
bull PolyphosphatesReversion partially (formation of orthophosphate)P3O10
-5 ---gt P2O7-3 + PO4
-3
polymer --gt monomersComplexation (soluble complexes)3Pb+2 + P3O10
-5 --gt PbP3O10-2
higher Pb solubility
Slid
e cr
edit
Mar
c Ed
war
ds
Conceptual possible reactions Actual reactants and products may bedifferent Polyphosphates cannot be easily quantified
20
Ortho-P Treatment for Pb+2
Effectiveness Depends on Dose DIC pH and ldquoCleanlinessrdquo of Pipe Surface
mg PO4L00 10 20 30 40 50
mg
PbL
000
005
010
015
020
025
030
035
040
48 mg CL
48 mg CL
pH = 70pH = 75pH = 80pH = 85
bull pH less critical at low DICbull pH less critical at high PO4bull Point of diminishing returns higher with high DIC
bull Faster Pb reduction at high PO4
Typical UK Dosages 4-6 mgL
Most PWSs with LSLs do not have corrosion control treatment that minimizes Pb release and exposure
Effect of pH and ortho-P on Pb release
21
DIC = 10 mg CL 1 mg PO4L
60 65 70 75 80 85 90 95
gL
b m
P
0001
001
01
1
10
US Action Level
Dissolved Pb with ortho-PTotal Pb with ortho-PDissolved Pb with CO3Total Pb with CO3
At low DIC orthophosphate improves lead release regardless of pH
pHSchock M R DeSantis M K Metz D H Welch M M Hyland R N Nadagouda M N Revisiting the pH Effect on the Orthophosphate Control of Plumbosolvency Proc AWWA Annual Conference and Exposition Atlanta GA 2008
Ortho-P at pH 90 (DIC 6 mgL)
22
Pb(μgL)
bull Low Pb at 1 mgL relative to higher doses needs to be further verified across water qualities
bull Must do dose optimization study for your own water quality especially at high pH
bull Ortho-P may precipitate with Ca
Miller S A Investigation of Lead Solubility and Orthophosphate Addition in High pH Low DIC Water Master of Science Department of Biomedical Chemical and Environmental Engineering College of Engineering and Applied Science University of Cincinnati Cincinnati OH 2014
Ortho-Ppoint of diminishing returns
bull Orthophosphate addition to where large increments result in small reductions in lead release
bull Key to cost-effective lead release control and exposure reduction
bull Becomes the rdquomaintenancerdquo dosage unless the scale ldquoagesrdquo
bull Background constituents such as aluminum hardness ions iron manganese and others interfere with optimum reduction of lead release
bull Varies with the background water chemistry from system to system
23
Sodium silicate
bull No systematic studies to look at pH carbonate silicate background chemistry relationships
bull Unclear mechanism pH increase Buffer intensity increase Protective film (thin amp amorphous)
bull Can sequester ironmanganese
bull Canrsquot be evaluated with fresh surfaces
24pH5 6 7 8 9 10 11
β ty
nsi
nte
Iuf
fer
B
00000
00005
00010TIC=48 mg CL20 mgL SiO250 mgL PO4β H2OTotal β
Silicate may contribute to Buffer Intensity
Carbonate orthophosphate silicate I=001 25ordmC
Chemical changes may reduce Pb+4 to Pb+2
10
8
6
4
)ts 2ol
h (v
E
ndash2
ndash4
ndash6
ndash8
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
pH
12
PbO2 (plattnerite)
Pb++ )s
deg
(2
CO
3
(OH
) -2 2) 3
Pb
-2
0 (CO
-- 2
0
) 43(C
O
Pb
3 (OH
)
DIC = 18 mg CL
Pb Pb Pb = 0010 mgL
Pb metal
-10
Drop in ORP from treatment change or DS
oxidant demand
Drop in pH at surface from treatment change
chemical reactions nitrification etc
C D
ngto
nhisa
W
Newark NJ
Disinfectant demand in DS must be controlled amp enough free chlorine consistently maintained throughout LSL area to keep protective Pb+4 corrosion scales in place
Switching disinfectants may reduce Pb+4 to Pb+2
Low
Switch from chlorine to chloramine disinfectant dissolved lead from pipe scales during the Washington DC ldquolead-in-water crisisrdquo in 2001-2004
Example profile of Pb(+4) (ie PbO2) scale house
27
20
L)
Cl1 Cincinnati OH
18
microg 16
n ( 14
o 12
tia 10
trn 8
e 6
nco 4
C 2
b P 0-1 0 1 2 3 4 5 6 7 8 9 10 11
Cumulative Water Volume (L)
CI-1 10 h 1209CI-1 10 h 0410
Plumbing sequence corresponding to water volumeFaucet LSLCu pipe Brass
Water MainMCu pipe 2
015
a
let
dou
illyaf
ant
irT
bull Pb release to water from PbO2 coated LSLs was found to be lowbull Be aware are there are no known cases of exhumed PbO2 coated pipes being able to
re-stabilize in pipe rigs
Chloride and sulfateCSMR a factor controlling galvanic corrosion (Oliphant 1983 Gregory 1986)- Chloride detrimental- Sulfate beneficial if sufficient to overcome chloride thus- CSMR ratio was created
[Cl- ] 12 mgL Cl-Chloride to Sulfate Mass Ratio (CSMR) = =[SO -2 2 = 06
4 ] 20 mgL SO-4
Confirmed empirically in case studies of water utilities (Dodrill and Edwards 1995 Edwards and Triantafyllidou 2007) - High CSMR gt 05-06 more corrosive- Low CSMR lt 05-06 less corrosive
28
Verified in laboratory experiments with fresh galvanic Pb junctions (eg Wang et al 2013 Triantafyllidou et al 2011 Nguyen et al 2010 2011)
Tria
ntaf
yllid
ou e
t al
201
1
Chloride and sulfate
29
CSMR considered contributing factor to the Flint water Pb crisis (Pieper et al 2017)
30
Chloride and sulfate
bull Absolute concentrations of chloride and sulfate important chloride considered particularly detrimental to galvanic Pb corrosion
bull Increasing chloride and sulfate concentrations at a constant CSMR of 1 increased total lead concentrations in galvanic corrosion experiments (Ng amp Lin 2016)
bull High CSMR ratio proven empirically important in aggravating galvanic lead corrosion (verified by experiments and modeling)
bull Water source or treatment changes that affect the CSMR must be evaluated prior to change
bull Absolute concentrations also important particularly for chloride
31
Manganese
Two distinct pipe scale categories in 5 LSLs (Lead 1-5 Pipes) excavated from same city
31
Schock et al 2014 JAWWAbull The outer pipe scale is in direct contact with the flowing water affecting
water lead contamination the mostbull Plattnerite an insoluble Pb(+4) scale that protects water from lead
contamination was inhibited by Mn (and Fe) presence
Manganese
In this case study (Schock et al JAWWA 2014)bull Where Mn was present in more abundance
- Pb scales had less PbO2- Pb associated with Mn sorbing onto Mn+Fe-rich surface deposit
bull Erratic Pb persisted after full LSLR in areas rich with Mn deposits (and to lesser extent Fe)
bull Galvanized interior plumbing also accumulated dissolved Pb while LSL present
32
33
Manganese
Other studiesbull Dong et al 2000 Mn acted as vector for particulate Pb movementbull Trueman et al 2019 Increasing Mn increased Pb in experimentbull McNeill and Edwards 2004 Mn sequestration by polyphosphate destabilized
Pb scales and caused Pb release
bull Pan et al 2019 Possible benefit by promoting formation of Pb(+4)
Reference Analysis Mn Impact on PbSchock et al 2014 Excavated LSLs
amp historical water Pb data in a US city
bull Promoted amorphous Pb scalebull Inhibited insoluble Pb(+4) scalebull Erratic Pb persisted after full LSLR in areas rich
with Mn deposits
Tully et al 2019 Excavated LSLs amp water Pb data in several Midwest case studies
bull Mn was a component of amorphous pipe scales in several systems
bull Amorphous scales not amenable to determining their adherence to pipe walls their responses to changes in various water chemistry variables or quantitatively calculating their lead release
Itrsquos not just manganese
34
Broad variety of coatings on LSLs analyzed at EPA
Al Mn Fe P and Ca-rich coatings may interfere withorthophosphate pH adjustment or other CCT
35
More water parameters affect metal release
bull pH amp AlkalinityDICbull Concentration and type of
oxidants (disinfectantsdissolved oxygen)
bull ORPcorrosion potentialbull Corrosion Inhibitorsbull Chloridebull Sulfatebull Manganese
bull Temperaturebull Sorptive surfaces downstream of
LSLs (ie galvanized interior pipe)bull Iron (deposition and corrosion)bull Calciumbull Aluminumbull NOM (type amount)bull Mixing from water treatment plants
or sourcesbull Ammoniabull Hydrogen Sulfidebull Microbial activity (nitrification amp
other)
Corrosion Control Treatment is intertwined with all treatments affecting water chemistry
Water Quality Parameters(WQP)
bull pH alkalinity calcium orthophosphate or silica othersbull WQP measurements are important for
bull Process control (measurement frequency needs to be HIGH)bull Keeping treatment from being turned offbull Keeping inhibitor dosages amp pH control from being cut to save $$$bull Keeping required repairs to treatment on tight schedulebull Preventing frequent variations and alternations of water sources
bull Effectiveness of Pb treatment only determinable from tap sampling by combination of different assessment approaches for different lead plumbing sources to which people can be exposed
36
bull OWQP measurements important for process control look at yourdata to understand what they mean and avoid problems
bull OWQP do not predict lead release There are no substitutes fordirectly monitoring lead release
37
M+
M+
M+
M+
Corrosion barrier film hypothesis
Inert barrier film (eg CaCO3)
Pipe Wall
Barrier film would allow tobull Protect ALL pipe materialsbull Monitor corrosion based on one chemical reaction
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
Pb sources
Full Lead Service Line
Clark et al 2013
Brass unionvalve vs plastic
Partial Lead Service Line
Lead gooseneckTriantafyllidou et al 2020
What is in the distribution system Materials inventory is critical to understand where and what lead sources still exist
Pb sources
Leaded Solder
2 Pb
lt05Pb
3 Pb
Leaded Brass FaucetSelover 2005
Gal
vani
zed
Diverse legacy leaded materials may undergo different corrosion reactions and impact water quality differently
Corrosion is oxidation-reduction
Pb harr Pb2+ + 2e-
Pb harr Pb4+ + 4e-
OxidationLead metal losing electronsat anode
OCl- + H+ + 2e- harr Cl- + OH-ReductionOxidant gainingelectronsat cathode
Oxidant is dissolved oxygen free chlorine chloramine chlorine dioxide etc
2OCl- + 2H+ + 4e-harr 2Cl- + 2OH-
There are millions of anodecathode sites across interior fresh lead pipe surface
Corrosion and scale formation
8
Corrosion
Pb(s)
OCl- OH-
2 e-
Pb+2 (or Pb+4)
2 e-
Scale Formation
water
Pb pipe
Pb pipe
water
Pb amp other solids
Pb2+
Pb(+
2) c
arbo
nate
sPb
(+2)
hyd
roxy
carb
onat
esPb
(+2)
orth
opho
spha
te
harr Pb4+
Pb(+
4) o
xide
harr
Idealized scenario of scale solids Scale is way more complex (heterogeneous several layers amorphous compounds) and it controls lead release
Tria
ntaf
yllid
ou e
t al
202
0
Corrosion and metal release
bull Pure corrosion (ie electrochemical oxidation-reduction reactions) becomes less important when the surface becomes covered with scale
bull Metal solubility or destabilization of the scale at the water contact will be more important (if the scale is porous corrosion may still occur at the base of the scale)
bull Metal solubility varies by factor of 5 to 10 or more across drinking water systems
CCT includes both pure corrosion and control of metal release from the pipe scale
Corrosion typesbull Uniform corrosion
- Materials degradation- Metal release (lead copper etc)
bull Non-uniform corrosion- Pinhole leaks (copper)- Dezincification (brass)- Tuberculation (iron galvanized brass)- Galvanic Corrosion
bull Galvanic corrosion- Soldered joints- Brass devices- Partial LSLs- Any coupling of different metals
Different leaded materials may undergo different corrosion mechanisms depending on composition configuration geometry
water quality and other factors
Pb pipe uniform
Pb pipe galvanic
DeSantis et al 2018
Important factors affecting corrosion and metal release
bull pH and AlkalinityDissolved Inorganic Carbon
bull Type and Concentration of Oxidants (chlorine species dissolved oxygen etc)OxidationReduction Potential
bull Corrosion Inhibitors
bull Chloride to Sulfate Mass Ratio
bull Manganese
pH is master variable
12
-1
0
1
2
0 2 4 6 8 10 12 14
E (V
OLT
S)
pH
IMMUNE
Simplified Pourbaix diagram (EH-pH Diagram)
Desirable pH-EH combinations allow passivation (ie formation ofprotective pipe scales)
pH is master variable
12
PbO2 (plattnerite)
Pb ++
deg s)(3O 2 -2C H) 2)b O 3P O
00
( 2 -- -2) C 43 ( )
CO
b HP O( ( DIC = 18 mg CL
3 b
Pb P Pb = 0010 mgL
Pb metal
-10
13
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
ndash8
ndash6
ndash4
ndash2
2
4
6
8
10
pH
Eh (v
olts
)
Dissolved Inorganic Carbon (DIC) and Total ALKalinity (TALK)
DIC = [CO32-] + [H2CO3
] + [HCO3- ]]
TALK = 2 [CO32-] + [HCO3
-] + [OH-] - [H+]
mg CL Dissolved Inorganic Carbon (DIC)0 10 20 30 40 50
mg
CaCO
3L
Tota
l Alka
linity
0255075
100125150175
200225250
pH 60pH 70pH 80pH 90pH 100
bull To understand corrosion it isimportant to keep up with thecarbonate system
bull DIC and TALK have linearrelationship but are not thesame thing
bullbullbullbullbull
Oxidants in drinking water
bull Disinfection (Free chlorine Chloramine Chlorine dioxide)Pre-(O3 H2O2 ClO2 KMnO4) Oxidative metal removal (eg As Fe Mn) Ammonia removalAeration (corrosion control VOCRn2S removal) Taste and odor control
bull Dissolved oxygen
Oxidants in drinking water
Oxidant Dosage (mgL)0 2 4 6 8 10
E H (V
olts
vs S
HE)
03
04
05
06
07
08
09
10
11
MCA Electrode 1 MCA Electrode 2 KMnO4 Electrode 1 KMnO4 Electrode 2 ClO2 Electrode 1 ClO2 Electrode 2 DO Electrode 1 DO Electrode 2 Cl2 Electrode 1Cl2 Electrode 2
ClO2 HOCldeg
KMnO4NH2Cl
DO
Different oxidants have different oxidizing power
Oxidation-Reduction Potential of several disinfectants in experimentJames et al 2004 (pH 7 10 mg CL 25degC)
17
ldquoClassicrdquo divalent Pb+2 solubility
pH6 7 8 9 10 11
mg
PbL
001
01
1
10
100 1 mg CL 5 mg CL 10 mg CL 20 mg CL 35 mg CL 50 mg CL 75 mg CL100 mg CL
DIC
Optimum pHDIC rangefor LSLs
High pH is needed to minimize Pb solubility
Lead ldquocorrosion Inhibitorsrdquo
bull Phosphoric acidbull Alkalai-metal orthophosphatebull Zinc orthophosphatebull Blended orthopolyphosphatesbull Polyphosphatesbull Sodium Silicate
Polyphosphates are not corrosion inhibitors but rather sequestering agentsEfficiency depends on sufficient doseconcentration for thebackground pH and other reactive water quality constituentswhat if the dose is too low to actually work 18
Illustrative reactions of P in pipebull OrthophosphatePrecipitation (consumption of orthophosphate)3Pb+2 + 2PO4
-3 --gt Pb3(PO4)2 (solid)decreased Pb solubility
bull PolyphosphatesReversion partially (formation of orthophosphate)P3O10
-5 ---gt P2O7-3 + PO4
-3
polymer --gt monomersComplexation (soluble complexes)3Pb+2 + P3O10
-5 --gt PbP3O10-2
higher Pb solubility
Slid
e cr
edit
Mar
c Ed
war
ds
Conceptual possible reactions Actual reactants and products may bedifferent Polyphosphates cannot be easily quantified
20
Ortho-P Treatment for Pb+2
Effectiveness Depends on Dose DIC pH and ldquoCleanlinessrdquo of Pipe Surface
mg PO4L00 10 20 30 40 50
mg
PbL
000
005
010
015
020
025
030
035
040
48 mg CL
48 mg CL
pH = 70pH = 75pH = 80pH = 85
bull pH less critical at low DICbull pH less critical at high PO4bull Point of diminishing returns higher with high DIC
bull Faster Pb reduction at high PO4
Typical UK Dosages 4-6 mgL
Most PWSs with LSLs do not have corrosion control treatment that minimizes Pb release and exposure
Effect of pH and ortho-P on Pb release
21
DIC = 10 mg CL 1 mg PO4L
60 65 70 75 80 85 90 95
gL
b m
P
0001
001
01
1
10
US Action Level
Dissolved Pb with ortho-PTotal Pb with ortho-PDissolved Pb with CO3Total Pb with CO3
At low DIC orthophosphate improves lead release regardless of pH
pHSchock M R DeSantis M K Metz D H Welch M M Hyland R N Nadagouda M N Revisiting the pH Effect on the Orthophosphate Control of Plumbosolvency Proc AWWA Annual Conference and Exposition Atlanta GA 2008
Ortho-P at pH 90 (DIC 6 mgL)
22
Pb(μgL)
bull Low Pb at 1 mgL relative to higher doses needs to be further verified across water qualities
bull Must do dose optimization study for your own water quality especially at high pH
bull Ortho-P may precipitate with Ca
Miller S A Investigation of Lead Solubility and Orthophosphate Addition in High pH Low DIC Water Master of Science Department of Biomedical Chemical and Environmental Engineering College of Engineering and Applied Science University of Cincinnati Cincinnati OH 2014
Ortho-Ppoint of diminishing returns
bull Orthophosphate addition to where large increments result in small reductions in lead release
bull Key to cost-effective lead release control and exposure reduction
bull Becomes the rdquomaintenancerdquo dosage unless the scale ldquoagesrdquo
bull Background constituents such as aluminum hardness ions iron manganese and others interfere with optimum reduction of lead release
bull Varies with the background water chemistry from system to system
23
Sodium silicate
bull No systematic studies to look at pH carbonate silicate background chemistry relationships
bull Unclear mechanism pH increase Buffer intensity increase Protective film (thin amp amorphous)
bull Can sequester ironmanganese
bull Canrsquot be evaluated with fresh surfaces
24pH5 6 7 8 9 10 11
β ty
nsi
nte
Iuf
fer
B
00000
00005
00010TIC=48 mg CL20 mgL SiO250 mgL PO4β H2OTotal β
Silicate may contribute to Buffer Intensity
Carbonate orthophosphate silicate I=001 25ordmC
Chemical changes may reduce Pb+4 to Pb+2
10
8
6
4
)ts 2ol
h (v
E
ndash2
ndash4
ndash6
ndash8
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
pH
12
PbO2 (plattnerite)
Pb++ )s
deg
(2
CO
3
(OH
) -2 2) 3
Pb
-2
0 (CO
-- 2
0
) 43(C
O
Pb
3 (OH
)
DIC = 18 mg CL
Pb Pb Pb = 0010 mgL
Pb metal
-10
Drop in ORP from treatment change or DS
oxidant demand
Drop in pH at surface from treatment change
chemical reactions nitrification etc
C D
ngto
nhisa
W
Newark NJ
Disinfectant demand in DS must be controlled amp enough free chlorine consistently maintained throughout LSL area to keep protective Pb+4 corrosion scales in place
Switching disinfectants may reduce Pb+4 to Pb+2
Low
Switch from chlorine to chloramine disinfectant dissolved lead from pipe scales during the Washington DC ldquolead-in-water crisisrdquo in 2001-2004
Example profile of Pb(+4) (ie PbO2) scale house
27
20
L)
Cl1 Cincinnati OH
18
microg 16
n ( 14
o 12
tia 10
trn 8
e 6
nco 4
C 2
b P 0-1 0 1 2 3 4 5 6 7 8 9 10 11
Cumulative Water Volume (L)
CI-1 10 h 1209CI-1 10 h 0410
Plumbing sequence corresponding to water volumeFaucet LSLCu pipe Brass
Water MainMCu pipe 2
015
a
let
dou
illyaf
ant
irT
bull Pb release to water from PbO2 coated LSLs was found to be lowbull Be aware are there are no known cases of exhumed PbO2 coated pipes being able to
re-stabilize in pipe rigs
Chloride and sulfateCSMR a factor controlling galvanic corrosion (Oliphant 1983 Gregory 1986)- Chloride detrimental- Sulfate beneficial if sufficient to overcome chloride thus- CSMR ratio was created
[Cl- ] 12 mgL Cl-Chloride to Sulfate Mass Ratio (CSMR) = =[SO -2 2 = 06
4 ] 20 mgL SO-4
Confirmed empirically in case studies of water utilities (Dodrill and Edwards 1995 Edwards and Triantafyllidou 2007) - High CSMR gt 05-06 more corrosive- Low CSMR lt 05-06 less corrosive
28
Verified in laboratory experiments with fresh galvanic Pb junctions (eg Wang et al 2013 Triantafyllidou et al 2011 Nguyen et al 2010 2011)
Tria
ntaf
yllid
ou e
t al
201
1
Chloride and sulfate
29
CSMR considered contributing factor to the Flint water Pb crisis (Pieper et al 2017)
30
Chloride and sulfate
bull Absolute concentrations of chloride and sulfate important chloride considered particularly detrimental to galvanic Pb corrosion
bull Increasing chloride and sulfate concentrations at a constant CSMR of 1 increased total lead concentrations in galvanic corrosion experiments (Ng amp Lin 2016)
bull High CSMR ratio proven empirically important in aggravating galvanic lead corrosion (verified by experiments and modeling)
bull Water source or treatment changes that affect the CSMR must be evaluated prior to change
bull Absolute concentrations also important particularly for chloride
31
Manganese
Two distinct pipe scale categories in 5 LSLs (Lead 1-5 Pipes) excavated from same city
31
Schock et al 2014 JAWWAbull The outer pipe scale is in direct contact with the flowing water affecting
water lead contamination the mostbull Plattnerite an insoluble Pb(+4) scale that protects water from lead
contamination was inhibited by Mn (and Fe) presence
Manganese
In this case study (Schock et al JAWWA 2014)bull Where Mn was present in more abundance
- Pb scales had less PbO2- Pb associated with Mn sorbing onto Mn+Fe-rich surface deposit
bull Erratic Pb persisted after full LSLR in areas rich with Mn deposits (and to lesser extent Fe)
bull Galvanized interior plumbing also accumulated dissolved Pb while LSL present
32
33
Manganese
Other studiesbull Dong et al 2000 Mn acted as vector for particulate Pb movementbull Trueman et al 2019 Increasing Mn increased Pb in experimentbull McNeill and Edwards 2004 Mn sequestration by polyphosphate destabilized
Pb scales and caused Pb release
bull Pan et al 2019 Possible benefit by promoting formation of Pb(+4)
Reference Analysis Mn Impact on PbSchock et al 2014 Excavated LSLs
amp historical water Pb data in a US city
bull Promoted amorphous Pb scalebull Inhibited insoluble Pb(+4) scalebull Erratic Pb persisted after full LSLR in areas rich
with Mn deposits
Tully et al 2019 Excavated LSLs amp water Pb data in several Midwest case studies
bull Mn was a component of amorphous pipe scales in several systems
bull Amorphous scales not amenable to determining their adherence to pipe walls their responses to changes in various water chemistry variables or quantitatively calculating their lead release
Itrsquos not just manganese
34
Broad variety of coatings on LSLs analyzed at EPA
Al Mn Fe P and Ca-rich coatings may interfere withorthophosphate pH adjustment or other CCT
35
More water parameters affect metal release
bull pH amp AlkalinityDICbull Concentration and type of
oxidants (disinfectantsdissolved oxygen)
bull ORPcorrosion potentialbull Corrosion Inhibitorsbull Chloridebull Sulfatebull Manganese
bull Temperaturebull Sorptive surfaces downstream of
LSLs (ie galvanized interior pipe)bull Iron (deposition and corrosion)bull Calciumbull Aluminumbull NOM (type amount)bull Mixing from water treatment plants
or sourcesbull Ammoniabull Hydrogen Sulfidebull Microbial activity (nitrification amp
other)
Corrosion Control Treatment is intertwined with all treatments affecting water chemistry
Water Quality Parameters(WQP)
bull pH alkalinity calcium orthophosphate or silica othersbull WQP measurements are important for
bull Process control (measurement frequency needs to be HIGH)bull Keeping treatment from being turned offbull Keeping inhibitor dosages amp pH control from being cut to save $$$bull Keeping required repairs to treatment on tight schedulebull Preventing frequent variations and alternations of water sources
bull Effectiveness of Pb treatment only determinable from tap sampling by combination of different assessment approaches for different lead plumbing sources to which people can be exposed
36
bull OWQP measurements important for process control look at yourdata to understand what they mean and avoid problems
bull OWQP do not predict lead release There are no substitutes fordirectly monitoring lead release
37
M+
M+
M+
M+
Corrosion barrier film hypothesis
Inert barrier film (eg CaCO3)
Pipe Wall
Barrier film would allow tobull Protect ALL pipe materialsbull Monitor corrosion based on one chemical reaction
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
Pb sources
Leaded Solder
2 Pb
lt05Pb
3 Pb
Leaded Brass FaucetSelover 2005
Gal
vani
zed
Diverse legacy leaded materials may undergo different corrosion reactions and impact water quality differently
Corrosion is oxidation-reduction
Pb harr Pb2+ + 2e-
Pb harr Pb4+ + 4e-
OxidationLead metal losing electronsat anode
OCl- + H+ + 2e- harr Cl- + OH-ReductionOxidant gainingelectronsat cathode
Oxidant is dissolved oxygen free chlorine chloramine chlorine dioxide etc
2OCl- + 2H+ + 4e-harr 2Cl- + 2OH-
There are millions of anodecathode sites across interior fresh lead pipe surface
Corrosion and scale formation
8
Corrosion
Pb(s)
OCl- OH-
2 e-
Pb+2 (or Pb+4)
2 e-
Scale Formation
water
Pb pipe
Pb pipe
water
Pb amp other solids
Pb2+
Pb(+
2) c
arbo
nate
sPb
(+2)
hyd
roxy
carb
onat
esPb
(+2)
orth
opho
spha
te
harr Pb4+
Pb(+
4) o
xide
harr
Idealized scenario of scale solids Scale is way more complex (heterogeneous several layers amorphous compounds) and it controls lead release
Tria
ntaf
yllid
ou e
t al
202
0
Corrosion and metal release
bull Pure corrosion (ie electrochemical oxidation-reduction reactions) becomes less important when the surface becomes covered with scale
bull Metal solubility or destabilization of the scale at the water contact will be more important (if the scale is porous corrosion may still occur at the base of the scale)
bull Metal solubility varies by factor of 5 to 10 or more across drinking water systems
CCT includes both pure corrosion and control of metal release from the pipe scale
Corrosion typesbull Uniform corrosion
- Materials degradation- Metal release (lead copper etc)
bull Non-uniform corrosion- Pinhole leaks (copper)- Dezincification (brass)- Tuberculation (iron galvanized brass)- Galvanic Corrosion
bull Galvanic corrosion- Soldered joints- Brass devices- Partial LSLs- Any coupling of different metals
Different leaded materials may undergo different corrosion mechanisms depending on composition configuration geometry
water quality and other factors
Pb pipe uniform
Pb pipe galvanic
DeSantis et al 2018
Important factors affecting corrosion and metal release
bull pH and AlkalinityDissolved Inorganic Carbon
bull Type and Concentration of Oxidants (chlorine species dissolved oxygen etc)OxidationReduction Potential
bull Corrosion Inhibitors
bull Chloride to Sulfate Mass Ratio
bull Manganese
pH is master variable
12
-1
0
1
2
0 2 4 6 8 10 12 14
E (V
OLT
S)
pH
IMMUNE
Simplified Pourbaix diagram (EH-pH Diagram)
Desirable pH-EH combinations allow passivation (ie formation ofprotective pipe scales)
pH is master variable
12
PbO2 (plattnerite)
Pb ++
deg s)(3O 2 -2C H) 2)b O 3P O
00
( 2 -- -2) C 43 ( )
CO
b HP O( ( DIC = 18 mg CL
3 b
Pb P Pb = 0010 mgL
Pb metal
-10
13
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
ndash8
ndash6
ndash4
ndash2
2
4
6
8
10
pH
Eh (v
olts
)
Dissolved Inorganic Carbon (DIC) and Total ALKalinity (TALK)
DIC = [CO32-] + [H2CO3
] + [HCO3- ]]
TALK = 2 [CO32-] + [HCO3
-] + [OH-] - [H+]
mg CL Dissolved Inorganic Carbon (DIC)0 10 20 30 40 50
mg
CaCO
3L
Tota
l Alka
linity
0255075
100125150175
200225250
pH 60pH 70pH 80pH 90pH 100
bull To understand corrosion it isimportant to keep up with thecarbonate system
bull DIC and TALK have linearrelationship but are not thesame thing
bullbullbullbullbull
Oxidants in drinking water
bull Disinfection (Free chlorine Chloramine Chlorine dioxide)Pre-(O3 H2O2 ClO2 KMnO4) Oxidative metal removal (eg As Fe Mn) Ammonia removalAeration (corrosion control VOCRn2S removal) Taste and odor control
bull Dissolved oxygen
Oxidants in drinking water
Oxidant Dosage (mgL)0 2 4 6 8 10
E H (V
olts
vs S
HE)
03
04
05
06
07
08
09
10
11
MCA Electrode 1 MCA Electrode 2 KMnO4 Electrode 1 KMnO4 Electrode 2 ClO2 Electrode 1 ClO2 Electrode 2 DO Electrode 1 DO Electrode 2 Cl2 Electrode 1Cl2 Electrode 2
ClO2 HOCldeg
KMnO4NH2Cl
DO
Different oxidants have different oxidizing power
Oxidation-Reduction Potential of several disinfectants in experimentJames et al 2004 (pH 7 10 mg CL 25degC)
17
ldquoClassicrdquo divalent Pb+2 solubility
pH6 7 8 9 10 11
mg
PbL
001
01
1
10
100 1 mg CL 5 mg CL 10 mg CL 20 mg CL 35 mg CL 50 mg CL 75 mg CL100 mg CL
DIC
Optimum pHDIC rangefor LSLs
High pH is needed to minimize Pb solubility
Lead ldquocorrosion Inhibitorsrdquo
bull Phosphoric acidbull Alkalai-metal orthophosphatebull Zinc orthophosphatebull Blended orthopolyphosphatesbull Polyphosphatesbull Sodium Silicate
Polyphosphates are not corrosion inhibitors but rather sequestering agentsEfficiency depends on sufficient doseconcentration for thebackground pH and other reactive water quality constituentswhat if the dose is too low to actually work 18
Illustrative reactions of P in pipebull OrthophosphatePrecipitation (consumption of orthophosphate)3Pb+2 + 2PO4
-3 --gt Pb3(PO4)2 (solid)decreased Pb solubility
bull PolyphosphatesReversion partially (formation of orthophosphate)P3O10
-5 ---gt P2O7-3 + PO4
-3
polymer --gt monomersComplexation (soluble complexes)3Pb+2 + P3O10
-5 --gt PbP3O10-2
higher Pb solubility
Slid
e cr
edit
Mar
c Ed
war
ds
Conceptual possible reactions Actual reactants and products may bedifferent Polyphosphates cannot be easily quantified
20
Ortho-P Treatment for Pb+2
Effectiveness Depends on Dose DIC pH and ldquoCleanlinessrdquo of Pipe Surface
mg PO4L00 10 20 30 40 50
mg
PbL
000
005
010
015
020
025
030
035
040
48 mg CL
48 mg CL
pH = 70pH = 75pH = 80pH = 85
bull pH less critical at low DICbull pH less critical at high PO4bull Point of diminishing returns higher with high DIC
bull Faster Pb reduction at high PO4
Typical UK Dosages 4-6 mgL
Most PWSs with LSLs do not have corrosion control treatment that minimizes Pb release and exposure
Effect of pH and ortho-P on Pb release
21
DIC = 10 mg CL 1 mg PO4L
60 65 70 75 80 85 90 95
gL
b m
P
0001
001
01
1
10
US Action Level
Dissolved Pb with ortho-PTotal Pb with ortho-PDissolved Pb with CO3Total Pb with CO3
At low DIC orthophosphate improves lead release regardless of pH
pHSchock M R DeSantis M K Metz D H Welch M M Hyland R N Nadagouda M N Revisiting the pH Effect on the Orthophosphate Control of Plumbosolvency Proc AWWA Annual Conference and Exposition Atlanta GA 2008
Ortho-P at pH 90 (DIC 6 mgL)
22
Pb(μgL)
bull Low Pb at 1 mgL relative to higher doses needs to be further verified across water qualities
bull Must do dose optimization study for your own water quality especially at high pH
bull Ortho-P may precipitate with Ca
Miller S A Investigation of Lead Solubility and Orthophosphate Addition in High pH Low DIC Water Master of Science Department of Biomedical Chemical and Environmental Engineering College of Engineering and Applied Science University of Cincinnati Cincinnati OH 2014
Ortho-Ppoint of diminishing returns
bull Orthophosphate addition to where large increments result in small reductions in lead release
bull Key to cost-effective lead release control and exposure reduction
bull Becomes the rdquomaintenancerdquo dosage unless the scale ldquoagesrdquo
bull Background constituents such as aluminum hardness ions iron manganese and others interfere with optimum reduction of lead release
bull Varies with the background water chemistry from system to system
23
Sodium silicate
bull No systematic studies to look at pH carbonate silicate background chemistry relationships
bull Unclear mechanism pH increase Buffer intensity increase Protective film (thin amp amorphous)
bull Can sequester ironmanganese
bull Canrsquot be evaluated with fresh surfaces
24pH5 6 7 8 9 10 11
β ty
nsi
nte
Iuf
fer
B
00000
00005
00010TIC=48 mg CL20 mgL SiO250 mgL PO4β H2OTotal β
Silicate may contribute to Buffer Intensity
Carbonate orthophosphate silicate I=001 25ordmC
Chemical changes may reduce Pb+4 to Pb+2
10
8
6
4
)ts 2ol
h (v
E
ndash2
ndash4
ndash6
ndash8
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
pH
12
PbO2 (plattnerite)
Pb++ )s
deg
(2
CO
3
(OH
) -2 2) 3
Pb
-2
0 (CO
-- 2
0
) 43(C
O
Pb
3 (OH
)
DIC = 18 mg CL
Pb Pb Pb = 0010 mgL
Pb metal
-10
Drop in ORP from treatment change or DS
oxidant demand
Drop in pH at surface from treatment change
chemical reactions nitrification etc
C D
ngto
nhisa
W
Newark NJ
Disinfectant demand in DS must be controlled amp enough free chlorine consistently maintained throughout LSL area to keep protective Pb+4 corrosion scales in place
Switching disinfectants may reduce Pb+4 to Pb+2
Low
Switch from chlorine to chloramine disinfectant dissolved lead from pipe scales during the Washington DC ldquolead-in-water crisisrdquo in 2001-2004
Example profile of Pb(+4) (ie PbO2) scale house
27
20
L)
Cl1 Cincinnati OH
18
microg 16
n ( 14
o 12
tia 10
trn 8
e 6
nco 4
C 2
b P 0-1 0 1 2 3 4 5 6 7 8 9 10 11
Cumulative Water Volume (L)
CI-1 10 h 1209CI-1 10 h 0410
Plumbing sequence corresponding to water volumeFaucet LSLCu pipe Brass
Water MainMCu pipe 2
015
a
let
dou
illyaf
ant
irT
bull Pb release to water from PbO2 coated LSLs was found to be lowbull Be aware are there are no known cases of exhumed PbO2 coated pipes being able to
re-stabilize in pipe rigs
Chloride and sulfateCSMR a factor controlling galvanic corrosion (Oliphant 1983 Gregory 1986)- Chloride detrimental- Sulfate beneficial if sufficient to overcome chloride thus- CSMR ratio was created
[Cl- ] 12 mgL Cl-Chloride to Sulfate Mass Ratio (CSMR) = =[SO -2 2 = 06
4 ] 20 mgL SO-4
Confirmed empirically in case studies of water utilities (Dodrill and Edwards 1995 Edwards and Triantafyllidou 2007) - High CSMR gt 05-06 more corrosive- Low CSMR lt 05-06 less corrosive
28
Verified in laboratory experiments with fresh galvanic Pb junctions (eg Wang et al 2013 Triantafyllidou et al 2011 Nguyen et al 2010 2011)
Tria
ntaf
yllid
ou e
t al
201
1
Chloride and sulfate
29
CSMR considered contributing factor to the Flint water Pb crisis (Pieper et al 2017)
30
Chloride and sulfate
bull Absolute concentrations of chloride and sulfate important chloride considered particularly detrimental to galvanic Pb corrosion
bull Increasing chloride and sulfate concentrations at a constant CSMR of 1 increased total lead concentrations in galvanic corrosion experiments (Ng amp Lin 2016)
bull High CSMR ratio proven empirically important in aggravating galvanic lead corrosion (verified by experiments and modeling)
bull Water source or treatment changes that affect the CSMR must be evaluated prior to change
bull Absolute concentrations also important particularly for chloride
31
Manganese
Two distinct pipe scale categories in 5 LSLs (Lead 1-5 Pipes) excavated from same city
31
Schock et al 2014 JAWWAbull The outer pipe scale is in direct contact with the flowing water affecting
water lead contamination the mostbull Plattnerite an insoluble Pb(+4) scale that protects water from lead
contamination was inhibited by Mn (and Fe) presence
Manganese
In this case study (Schock et al JAWWA 2014)bull Where Mn was present in more abundance
- Pb scales had less PbO2- Pb associated with Mn sorbing onto Mn+Fe-rich surface deposit
bull Erratic Pb persisted after full LSLR in areas rich with Mn deposits (and to lesser extent Fe)
bull Galvanized interior plumbing also accumulated dissolved Pb while LSL present
32
33
Manganese
Other studiesbull Dong et al 2000 Mn acted as vector for particulate Pb movementbull Trueman et al 2019 Increasing Mn increased Pb in experimentbull McNeill and Edwards 2004 Mn sequestration by polyphosphate destabilized
Pb scales and caused Pb release
bull Pan et al 2019 Possible benefit by promoting formation of Pb(+4)
Reference Analysis Mn Impact on PbSchock et al 2014 Excavated LSLs
amp historical water Pb data in a US city
bull Promoted amorphous Pb scalebull Inhibited insoluble Pb(+4) scalebull Erratic Pb persisted after full LSLR in areas rich
with Mn deposits
Tully et al 2019 Excavated LSLs amp water Pb data in several Midwest case studies
bull Mn was a component of amorphous pipe scales in several systems
bull Amorphous scales not amenable to determining their adherence to pipe walls their responses to changes in various water chemistry variables or quantitatively calculating their lead release
Itrsquos not just manganese
34
Broad variety of coatings on LSLs analyzed at EPA
Al Mn Fe P and Ca-rich coatings may interfere withorthophosphate pH adjustment or other CCT
35
More water parameters affect metal release
bull pH amp AlkalinityDICbull Concentration and type of
oxidants (disinfectantsdissolved oxygen)
bull ORPcorrosion potentialbull Corrosion Inhibitorsbull Chloridebull Sulfatebull Manganese
bull Temperaturebull Sorptive surfaces downstream of
LSLs (ie galvanized interior pipe)bull Iron (deposition and corrosion)bull Calciumbull Aluminumbull NOM (type amount)bull Mixing from water treatment plants
or sourcesbull Ammoniabull Hydrogen Sulfidebull Microbial activity (nitrification amp
other)
Corrosion Control Treatment is intertwined with all treatments affecting water chemistry
Water Quality Parameters(WQP)
bull pH alkalinity calcium orthophosphate or silica othersbull WQP measurements are important for
bull Process control (measurement frequency needs to be HIGH)bull Keeping treatment from being turned offbull Keeping inhibitor dosages amp pH control from being cut to save $$$bull Keeping required repairs to treatment on tight schedulebull Preventing frequent variations and alternations of water sources
bull Effectiveness of Pb treatment only determinable from tap sampling by combination of different assessment approaches for different lead plumbing sources to which people can be exposed
36
bull OWQP measurements important for process control look at yourdata to understand what they mean and avoid problems
bull OWQP do not predict lead release There are no substitutes fordirectly monitoring lead release
37
M+
M+
M+
M+
Corrosion barrier film hypothesis
Inert barrier film (eg CaCO3)
Pipe Wall
Barrier film would allow tobull Protect ALL pipe materialsbull Monitor corrosion based on one chemical reaction
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
Corrosion is oxidation-reduction
Pb harr Pb2+ + 2e-
Pb harr Pb4+ + 4e-
OxidationLead metal losing electronsat anode
OCl- + H+ + 2e- harr Cl- + OH-ReductionOxidant gainingelectronsat cathode
Oxidant is dissolved oxygen free chlorine chloramine chlorine dioxide etc
2OCl- + 2H+ + 4e-harr 2Cl- + 2OH-
There are millions of anodecathode sites across interior fresh lead pipe surface
Corrosion and scale formation
8
Corrosion
Pb(s)
OCl- OH-
2 e-
Pb+2 (or Pb+4)
2 e-
Scale Formation
water
Pb pipe
Pb pipe
water
Pb amp other solids
Pb2+
Pb(+
2) c
arbo
nate
sPb
(+2)
hyd
roxy
carb
onat
esPb
(+2)
orth
opho
spha
te
harr Pb4+
Pb(+
4) o
xide
harr
Idealized scenario of scale solids Scale is way more complex (heterogeneous several layers amorphous compounds) and it controls lead release
Tria
ntaf
yllid
ou e
t al
202
0
Corrosion and metal release
bull Pure corrosion (ie electrochemical oxidation-reduction reactions) becomes less important when the surface becomes covered with scale
bull Metal solubility or destabilization of the scale at the water contact will be more important (if the scale is porous corrosion may still occur at the base of the scale)
bull Metal solubility varies by factor of 5 to 10 or more across drinking water systems
CCT includes both pure corrosion and control of metal release from the pipe scale
Corrosion typesbull Uniform corrosion
- Materials degradation- Metal release (lead copper etc)
bull Non-uniform corrosion- Pinhole leaks (copper)- Dezincification (brass)- Tuberculation (iron galvanized brass)- Galvanic Corrosion
bull Galvanic corrosion- Soldered joints- Brass devices- Partial LSLs- Any coupling of different metals
Different leaded materials may undergo different corrosion mechanisms depending on composition configuration geometry
water quality and other factors
Pb pipe uniform
Pb pipe galvanic
DeSantis et al 2018
Important factors affecting corrosion and metal release
bull pH and AlkalinityDissolved Inorganic Carbon
bull Type and Concentration of Oxidants (chlorine species dissolved oxygen etc)OxidationReduction Potential
bull Corrosion Inhibitors
bull Chloride to Sulfate Mass Ratio
bull Manganese
pH is master variable
12
-1
0
1
2
0 2 4 6 8 10 12 14
E (V
OLT
S)
pH
IMMUNE
Simplified Pourbaix diagram (EH-pH Diagram)
Desirable pH-EH combinations allow passivation (ie formation ofprotective pipe scales)
pH is master variable
12
PbO2 (plattnerite)
Pb ++
deg s)(3O 2 -2C H) 2)b O 3P O
00
( 2 -- -2) C 43 ( )
CO
b HP O( ( DIC = 18 mg CL
3 b
Pb P Pb = 0010 mgL
Pb metal
-10
13
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
ndash8
ndash6
ndash4
ndash2
2
4
6
8
10
pH
Eh (v
olts
)
Dissolved Inorganic Carbon (DIC) and Total ALKalinity (TALK)
DIC = [CO32-] + [H2CO3
] + [HCO3- ]]
TALK = 2 [CO32-] + [HCO3
-] + [OH-] - [H+]
mg CL Dissolved Inorganic Carbon (DIC)0 10 20 30 40 50
mg
CaCO
3L
Tota
l Alka
linity
0255075
100125150175
200225250
pH 60pH 70pH 80pH 90pH 100
bull To understand corrosion it isimportant to keep up with thecarbonate system
bull DIC and TALK have linearrelationship but are not thesame thing
bullbullbullbullbull
Oxidants in drinking water
bull Disinfection (Free chlorine Chloramine Chlorine dioxide)Pre-(O3 H2O2 ClO2 KMnO4) Oxidative metal removal (eg As Fe Mn) Ammonia removalAeration (corrosion control VOCRn2S removal) Taste and odor control
bull Dissolved oxygen
Oxidants in drinking water
Oxidant Dosage (mgL)0 2 4 6 8 10
E H (V
olts
vs S
HE)
03
04
05
06
07
08
09
10
11
MCA Electrode 1 MCA Electrode 2 KMnO4 Electrode 1 KMnO4 Electrode 2 ClO2 Electrode 1 ClO2 Electrode 2 DO Electrode 1 DO Electrode 2 Cl2 Electrode 1Cl2 Electrode 2
ClO2 HOCldeg
KMnO4NH2Cl
DO
Different oxidants have different oxidizing power
Oxidation-Reduction Potential of several disinfectants in experimentJames et al 2004 (pH 7 10 mg CL 25degC)
17
ldquoClassicrdquo divalent Pb+2 solubility
pH6 7 8 9 10 11
mg
PbL
001
01
1
10
100 1 mg CL 5 mg CL 10 mg CL 20 mg CL 35 mg CL 50 mg CL 75 mg CL100 mg CL
DIC
Optimum pHDIC rangefor LSLs
High pH is needed to minimize Pb solubility
Lead ldquocorrosion Inhibitorsrdquo
bull Phosphoric acidbull Alkalai-metal orthophosphatebull Zinc orthophosphatebull Blended orthopolyphosphatesbull Polyphosphatesbull Sodium Silicate
Polyphosphates are not corrosion inhibitors but rather sequestering agentsEfficiency depends on sufficient doseconcentration for thebackground pH and other reactive water quality constituentswhat if the dose is too low to actually work 18
Illustrative reactions of P in pipebull OrthophosphatePrecipitation (consumption of orthophosphate)3Pb+2 + 2PO4
-3 --gt Pb3(PO4)2 (solid)decreased Pb solubility
bull PolyphosphatesReversion partially (formation of orthophosphate)P3O10
-5 ---gt P2O7-3 + PO4
-3
polymer --gt monomersComplexation (soluble complexes)3Pb+2 + P3O10
-5 --gt PbP3O10-2
higher Pb solubility
Slid
e cr
edit
Mar
c Ed
war
ds
Conceptual possible reactions Actual reactants and products may bedifferent Polyphosphates cannot be easily quantified
20
Ortho-P Treatment for Pb+2
Effectiveness Depends on Dose DIC pH and ldquoCleanlinessrdquo of Pipe Surface
mg PO4L00 10 20 30 40 50
mg
PbL
000
005
010
015
020
025
030
035
040
48 mg CL
48 mg CL
pH = 70pH = 75pH = 80pH = 85
bull pH less critical at low DICbull pH less critical at high PO4bull Point of diminishing returns higher with high DIC
bull Faster Pb reduction at high PO4
Typical UK Dosages 4-6 mgL
Most PWSs with LSLs do not have corrosion control treatment that minimizes Pb release and exposure
Effect of pH and ortho-P on Pb release
21
DIC = 10 mg CL 1 mg PO4L
60 65 70 75 80 85 90 95
gL
b m
P
0001
001
01
1
10
US Action Level
Dissolved Pb with ortho-PTotal Pb with ortho-PDissolved Pb with CO3Total Pb with CO3
At low DIC orthophosphate improves lead release regardless of pH
pHSchock M R DeSantis M K Metz D H Welch M M Hyland R N Nadagouda M N Revisiting the pH Effect on the Orthophosphate Control of Plumbosolvency Proc AWWA Annual Conference and Exposition Atlanta GA 2008
Ortho-P at pH 90 (DIC 6 mgL)
22
Pb(μgL)
bull Low Pb at 1 mgL relative to higher doses needs to be further verified across water qualities
bull Must do dose optimization study for your own water quality especially at high pH
bull Ortho-P may precipitate with Ca
Miller S A Investigation of Lead Solubility and Orthophosphate Addition in High pH Low DIC Water Master of Science Department of Biomedical Chemical and Environmental Engineering College of Engineering and Applied Science University of Cincinnati Cincinnati OH 2014
Ortho-Ppoint of diminishing returns
bull Orthophosphate addition to where large increments result in small reductions in lead release
bull Key to cost-effective lead release control and exposure reduction
bull Becomes the rdquomaintenancerdquo dosage unless the scale ldquoagesrdquo
bull Background constituents such as aluminum hardness ions iron manganese and others interfere with optimum reduction of lead release
bull Varies with the background water chemistry from system to system
23
Sodium silicate
bull No systematic studies to look at pH carbonate silicate background chemistry relationships
bull Unclear mechanism pH increase Buffer intensity increase Protective film (thin amp amorphous)
bull Can sequester ironmanganese
bull Canrsquot be evaluated with fresh surfaces
24pH5 6 7 8 9 10 11
β ty
nsi
nte
Iuf
fer
B
00000
00005
00010TIC=48 mg CL20 mgL SiO250 mgL PO4β H2OTotal β
Silicate may contribute to Buffer Intensity
Carbonate orthophosphate silicate I=001 25ordmC
Chemical changes may reduce Pb+4 to Pb+2
10
8
6
4
)ts 2ol
h (v
E
ndash2
ndash4
ndash6
ndash8
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
pH
12
PbO2 (plattnerite)
Pb++ )s
deg
(2
CO
3
(OH
) -2 2) 3
Pb
-2
0 (CO
-- 2
0
) 43(C
O
Pb
3 (OH
)
DIC = 18 mg CL
Pb Pb Pb = 0010 mgL
Pb metal
-10
Drop in ORP from treatment change or DS
oxidant demand
Drop in pH at surface from treatment change
chemical reactions nitrification etc
C D
ngto
nhisa
W
Newark NJ
Disinfectant demand in DS must be controlled amp enough free chlorine consistently maintained throughout LSL area to keep protective Pb+4 corrosion scales in place
Switching disinfectants may reduce Pb+4 to Pb+2
Low
Switch from chlorine to chloramine disinfectant dissolved lead from pipe scales during the Washington DC ldquolead-in-water crisisrdquo in 2001-2004
Example profile of Pb(+4) (ie PbO2) scale house
27
20
L)
Cl1 Cincinnati OH
18
microg 16
n ( 14
o 12
tia 10
trn 8
e 6
nco 4
C 2
b P 0-1 0 1 2 3 4 5 6 7 8 9 10 11
Cumulative Water Volume (L)
CI-1 10 h 1209CI-1 10 h 0410
Plumbing sequence corresponding to water volumeFaucet LSLCu pipe Brass
Water MainMCu pipe 2
015
a
let
dou
illyaf
ant
irT
bull Pb release to water from PbO2 coated LSLs was found to be lowbull Be aware are there are no known cases of exhumed PbO2 coated pipes being able to
re-stabilize in pipe rigs
Chloride and sulfateCSMR a factor controlling galvanic corrosion (Oliphant 1983 Gregory 1986)- Chloride detrimental- Sulfate beneficial if sufficient to overcome chloride thus- CSMR ratio was created
[Cl- ] 12 mgL Cl-Chloride to Sulfate Mass Ratio (CSMR) = =[SO -2 2 = 06
4 ] 20 mgL SO-4
Confirmed empirically in case studies of water utilities (Dodrill and Edwards 1995 Edwards and Triantafyllidou 2007) - High CSMR gt 05-06 more corrosive- Low CSMR lt 05-06 less corrosive
28
Verified in laboratory experiments with fresh galvanic Pb junctions (eg Wang et al 2013 Triantafyllidou et al 2011 Nguyen et al 2010 2011)
Tria
ntaf
yllid
ou e
t al
201
1
Chloride and sulfate
29
CSMR considered contributing factor to the Flint water Pb crisis (Pieper et al 2017)
30
Chloride and sulfate
bull Absolute concentrations of chloride and sulfate important chloride considered particularly detrimental to galvanic Pb corrosion
bull Increasing chloride and sulfate concentrations at a constant CSMR of 1 increased total lead concentrations in galvanic corrosion experiments (Ng amp Lin 2016)
bull High CSMR ratio proven empirically important in aggravating galvanic lead corrosion (verified by experiments and modeling)
bull Water source or treatment changes that affect the CSMR must be evaluated prior to change
bull Absolute concentrations also important particularly for chloride
31
Manganese
Two distinct pipe scale categories in 5 LSLs (Lead 1-5 Pipes) excavated from same city
31
Schock et al 2014 JAWWAbull The outer pipe scale is in direct contact with the flowing water affecting
water lead contamination the mostbull Plattnerite an insoluble Pb(+4) scale that protects water from lead
contamination was inhibited by Mn (and Fe) presence
Manganese
In this case study (Schock et al JAWWA 2014)bull Where Mn was present in more abundance
- Pb scales had less PbO2- Pb associated with Mn sorbing onto Mn+Fe-rich surface deposit
bull Erratic Pb persisted after full LSLR in areas rich with Mn deposits (and to lesser extent Fe)
bull Galvanized interior plumbing also accumulated dissolved Pb while LSL present
32
33
Manganese
Other studiesbull Dong et al 2000 Mn acted as vector for particulate Pb movementbull Trueman et al 2019 Increasing Mn increased Pb in experimentbull McNeill and Edwards 2004 Mn sequestration by polyphosphate destabilized
Pb scales and caused Pb release
bull Pan et al 2019 Possible benefit by promoting formation of Pb(+4)
Reference Analysis Mn Impact on PbSchock et al 2014 Excavated LSLs
amp historical water Pb data in a US city
bull Promoted amorphous Pb scalebull Inhibited insoluble Pb(+4) scalebull Erratic Pb persisted after full LSLR in areas rich
with Mn deposits
Tully et al 2019 Excavated LSLs amp water Pb data in several Midwest case studies
bull Mn was a component of amorphous pipe scales in several systems
bull Amorphous scales not amenable to determining their adherence to pipe walls their responses to changes in various water chemistry variables or quantitatively calculating their lead release
Itrsquos not just manganese
34
Broad variety of coatings on LSLs analyzed at EPA
Al Mn Fe P and Ca-rich coatings may interfere withorthophosphate pH adjustment or other CCT
35
More water parameters affect metal release
bull pH amp AlkalinityDICbull Concentration and type of
oxidants (disinfectantsdissolved oxygen)
bull ORPcorrosion potentialbull Corrosion Inhibitorsbull Chloridebull Sulfatebull Manganese
bull Temperaturebull Sorptive surfaces downstream of
LSLs (ie galvanized interior pipe)bull Iron (deposition and corrosion)bull Calciumbull Aluminumbull NOM (type amount)bull Mixing from water treatment plants
or sourcesbull Ammoniabull Hydrogen Sulfidebull Microbial activity (nitrification amp
other)
Corrosion Control Treatment is intertwined with all treatments affecting water chemistry
Water Quality Parameters(WQP)
bull pH alkalinity calcium orthophosphate or silica othersbull WQP measurements are important for
bull Process control (measurement frequency needs to be HIGH)bull Keeping treatment from being turned offbull Keeping inhibitor dosages amp pH control from being cut to save $$$bull Keeping required repairs to treatment on tight schedulebull Preventing frequent variations and alternations of water sources
bull Effectiveness of Pb treatment only determinable from tap sampling by combination of different assessment approaches for different lead plumbing sources to which people can be exposed
36
bull OWQP measurements important for process control look at yourdata to understand what they mean and avoid problems
bull OWQP do not predict lead release There are no substitutes fordirectly monitoring lead release
37
M+
M+
M+
M+
Corrosion barrier film hypothesis
Inert barrier film (eg CaCO3)
Pipe Wall
Barrier film would allow tobull Protect ALL pipe materialsbull Monitor corrosion based on one chemical reaction
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
Corrosion and scale formation
8
Corrosion
Pb(s)
OCl- OH-
2 e-
Pb+2 (or Pb+4)
2 e-
Scale Formation
water
Pb pipe
Pb pipe
water
Pb amp other solids
Pb2+
Pb(+
2) c
arbo
nate
sPb
(+2)
hyd
roxy
carb
onat
esPb
(+2)
orth
opho
spha
te
harr Pb4+
Pb(+
4) o
xide
harr
Idealized scenario of scale solids Scale is way more complex (heterogeneous several layers amorphous compounds) and it controls lead release
Tria
ntaf
yllid
ou e
t al
202
0
Corrosion and metal release
bull Pure corrosion (ie electrochemical oxidation-reduction reactions) becomes less important when the surface becomes covered with scale
bull Metal solubility or destabilization of the scale at the water contact will be more important (if the scale is porous corrosion may still occur at the base of the scale)
bull Metal solubility varies by factor of 5 to 10 or more across drinking water systems
CCT includes both pure corrosion and control of metal release from the pipe scale
Corrosion typesbull Uniform corrosion
- Materials degradation- Metal release (lead copper etc)
bull Non-uniform corrosion- Pinhole leaks (copper)- Dezincification (brass)- Tuberculation (iron galvanized brass)- Galvanic Corrosion
bull Galvanic corrosion- Soldered joints- Brass devices- Partial LSLs- Any coupling of different metals
Different leaded materials may undergo different corrosion mechanisms depending on composition configuration geometry
water quality and other factors
Pb pipe uniform
Pb pipe galvanic
DeSantis et al 2018
Important factors affecting corrosion and metal release
bull pH and AlkalinityDissolved Inorganic Carbon
bull Type and Concentration of Oxidants (chlorine species dissolved oxygen etc)OxidationReduction Potential
bull Corrosion Inhibitors
bull Chloride to Sulfate Mass Ratio
bull Manganese
pH is master variable
12
-1
0
1
2
0 2 4 6 8 10 12 14
E (V
OLT
S)
pH
IMMUNE
Simplified Pourbaix diagram (EH-pH Diagram)
Desirable pH-EH combinations allow passivation (ie formation ofprotective pipe scales)
pH is master variable
12
PbO2 (plattnerite)
Pb ++
deg s)(3O 2 -2C H) 2)b O 3P O
00
( 2 -- -2) C 43 ( )
CO
b HP O( ( DIC = 18 mg CL
3 b
Pb P Pb = 0010 mgL
Pb metal
-10
13
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
ndash8
ndash6
ndash4
ndash2
2
4
6
8
10
pH
Eh (v
olts
)
Dissolved Inorganic Carbon (DIC) and Total ALKalinity (TALK)
DIC = [CO32-] + [H2CO3
] + [HCO3- ]]
TALK = 2 [CO32-] + [HCO3
-] + [OH-] - [H+]
mg CL Dissolved Inorganic Carbon (DIC)0 10 20 30 40 50
mg
CaCO
3L
Tota
l Alka
linity
0255075
100125150175
200225250
pH 60pH 70pH 80pH 90pH 100
bull To understand corrosion it isimportant to keep up with thecarbonate system
bull DIC and TALK have linearrelationship but are not thesame thing
bullbullbullbullbull
Oxidants in drinking water
bull Disinfection (Free chlorine Chloramine Chlorine dioxide)Pre-(O3 H2O2 ClO2 KMnO4) Oxidative metal removal (eg As Fe Mn) Ammonia removalAeration (corrosion control VOCRn2S removal) Taste and odor control
bull Dissolved oxygen
Oxidants in drinking water
Oxidant Dosage (mgL)0 2 4 6 8 10
E H (V
olts
vs S
HE)
03
04
05
06
07
08
09
10
11
MCA Electrode 1 MCA Electrode 2 KMnO4 Electrode 1 KMnO4 Electrode 2 ClO2 Electrode 1 ClO2 Electrode 2 DO Electrode 1 DO Electrode 2 Cl2 Electrode 1Cl2 Electrode 2
ClO2 HOCldeg
KMnO4NH2Cl
DO
Different oxidants have different oxidizing power
Oxidation-Reduction Potential of several disinfectants in experimentJames et al 2004 (pH 7 10 mg CL 25degC)
17
ldquoClassicrdquo divalent Pb+2 solubility
pH6 7 8 9 10 11
mg
PbL
001
01
1
10
100 1 mg CL 5 mg CL 10 mg CL 20 mg CL 35 mg CL 50 mg CL 75 mg CL100 mg CL
DIC
Optimum pHDIC rangefor LSLs
High pH is needed to minimize Pb solubility
Lead ldquocorrosion Inhibitorsrdquo
bull Phosphoric acidbull Alkalai-metal orthophosphatebull Zinc orthophosphatebull Blended orthopolyphosphatesbull Polyphosphatesbull Sodium Silicate
Polyphosphates are not corrosion inhibitors but rather sequestering agentsEfficiency depends on sufficient doseconcentration for thebackground pH and other reactive water quality constituentswhat if the dose is too low to actually work 18
Illustrative reactions of P in pipebull OrthophosphatePrecipitation (consumption of orthophosphate)3Pb+2 + 2PO4
-3 --gt Pb3(PO4)2 (solid)decreased Pb solubility
bull PolyphosphatesReversion partially (formation of orthophosphate)P3O10
-5 ---gt P2O7-3 + PO4
-3
polymer --gt monomersComplexation (soluble complexes)3Pb+2 + P3O10
-5 --gt PbP3O10-2
higher Pb solubility
Slid
e cr
edit
Mar
c Ed
war
ds
Conceptual possible reactions Actual reactants and products may bedifferent Polyphosphates cannot be easily quantified
20
Ortho-P Treatment for Pb+2
Effectiveness Depends on Dose DIC pH and ldquoCleanlinessrdquo of Pipe Surface
mg PO4L00 10 20 30 40 50
mg
PbL
000
005
010
015
020
025
030
035
040
48 mg CL
48 mg CL
pH = 70pH = 75pH = 80pH = 85
bull pH less critical at low DICbull pH less critical at high PO4bull Point of diminishing returns higher with high DIC
bull Faster Pb reduction at high PO4
Typical UK Dosages 4-6 mgL
Most PWSs with LSLs do not have corrosion control treatment that minimizes Pb release and exposure
Effect of pH and ortho-P on Pb release
21
DIC = 10 mg CL 1 mg PO4L
60 65 70 75 80 85 90 95
gL
b m
P
0001
001
01
1
10
US Action Level
Dissolved Pb with ortho-PTotal Pb with ortho-PDissolved Pb with CO3Total Pb with CO3
At low DIC orthophosphate improves lead release regardless of pH
pHSchock M R DeSantis M K Metz D H Welch M M Hyland R N Nadagouda M N Revisiting the pH Effect on the Orthophosphate Control of Plumbosolvency Proc AWWA Annual Conference and Exposition Atlanta GA 2008
Ortho-P at pH 90 (DIC 6 mgL)
22
Pb(μgL)
bull Low Pb at 1 mgL relative to higher doses needs to be further verified across water qualities
bull Must do dose optimization study for your own water quality especially at high pH
bull Ortho-P may precipitate with Ca
Miller S A Investigation of Lead Solubility and Orthophosphate Addition in High pH Low DIC Water Master of Science Department of Biomedical Chemical and Environmental Engineering College of Engineering and Applied Science University of Cincinnati Cincinnati OH 2014
Ortho-Ppoint of diminishing returns
bull Orthophosphate addition to where large increments result in small reductions in lead release
bull Key to cost-effective lead release control and exposure reduction
bull Becomes the rdquomaintenancerdquo dosage unless the scale ldquoagesrdquo
bull Background constituents such as aluminum hardness ions iron manganese and others interfere with optimum reduction of lead release
bull Varies with the background water chemistry from system to system
23
Sodium silicate
bull No systematic studies to look at pH carbonate silicate background chemistry relationships
bull Unclear mechanism pH increase Buffer intensity increase Protective film (thin amp amorphous)
bull Can sequester ironmanganese
bull Canrsquot be evaluated with fresh surfaces
24pH5 6 7 8 9 10 11
β ty
nsi
nte
Iuf
fer
B
00000
00005
00010TIC=48 mg CL20 mgL SiO250 mgL PO4β H2OTotal β
Silicate may contribute to Buffer Intensity
Carbonate orthophosphate silicate I=001 25ordmC
Chemical changes may reduce Pb+4 to Pb+2
10
8
6
4
)ts 2ol
h (v
E
ndash2
ndash4
ndash6
ndash8
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
pH
12
PbO2 (plattnerite)
Pb++ )s
deg
(2
CO
3
(OH
) -2 2) 3
Pb
-2
0 (CO
-- 2
0
) 43(C
O
Pb
3 (OH
)
DIC = 18 mg CL
Pb Pb Pb = 0010 mgL
Pb metal
-10
Drop in ORP from treatment change or DS
oxidant demand
Drop in pH at surface from treatment change
chemical reactions nitrification etc
C D
ngto
nhisa
W
Newark NJ
Disinfectant demand in DS must be controlled amp enough free chlorine consistently maintained throughout LSL area to keep protective Pb+4 corrosion scales in place
Switching disinfectants may reduce Pb+4 to Pb+2
Low
Switch from chlorine to chloramine disinfectant dissolved lead from pipe scales during the Washington DC ldquolead-in-water crisisrdquo in 2001-2004
Example profile of Pb(+4) (ie PbO2) scale house
27
20
L)
Cl1 Cincinnati OH
18
microg 16
n ( 14
o 12
tia 10
trn 8
e 6
nco 4
C 2
b P 0-1 0 1 2 3 4 5 6 7 8 9 10 11
Cumulative Water Volume (L)
CI-1 10 h 1209CI-1 10 h 0410
Plumbing sequence corresponding to water volumeFaucet LSLCu pipe Brass
Water MainMCu pipe 2
015
a
let
dou
illyaf
ant
irT
bull Pb release to water from PbO2 coated LSLs was found to be lowbull Be aware are there are no known cases of exhumed PbO2 coated pipes being able to
re-stabilize in pipe rigs
Chloride and sulfateCSMR a factor controlling galvanic corrosion (Oliphant 1983 Gregory 1986)- Chloride detrimental- Sulfate beneficial if sufficient to overcome chloride thus- CSMR ratio was created
[Cl- ] 12 mgL Cl-Chloride to Sulfate Mass Ratio (CSMR) = =[SO -2 2 = 06
4 ] 20 mgL SO-4
Confirmed empirically in case studies of water utilities (Dodrill and Edwards 1995 Edwards and Triantafyllidou 2007) - High CSMR gt 05-06 more corrosive- Low CSMR lt 05-06 less corrosive
28
Verified in laboratory experiments with fresh galvanic Pb junctions (eg Wang et al 2013 Triantafyllidou et al 2011 Nguyen et al 2010 2011)
Tria
ntaf
yllid
ou e
t al
201
1
Chloride and sulfate
29
CSMR considered contributing factor to the Flint water Pb crisis (Pieper et al 2017)
30
Chloride and sulfate
bull Absolute concentrations of chloride and sulfate important chloride considered particularly detrimental to galvanic Pb corrosion
bull Increasing chloride and sulfate concentrations at a constant CSMR of 1 increased total lead concentrations in galvanic corrosion experiments (Ng amp Lin 2016)
bull High CSMR ratio proven empirically important in aggravating galvanic lead corrosion (verified by experiments and modeling)
bull Water source or treatment changes that affect the CSMR must be evaluated prior to change
bull Absolute concentrations also important particularly for chloride
31
Manganese
Two distinct pipe scale categories in 5 LSLs (Lead 1-5 Pipes) excavated from same city
31
Schock et al 2014 JAWWAbull The outer pipe scale is in direct contact with the flowing water affecting
water lead contamination the mostbull Plattnerite an insoluble Pb(+4) scale that protects water from lead
contamination was inhibited by Mn (and Fe) presence
Manganese
In this case study (Schock et al JAWWA 2014)bull Where Mn was present in more abundance
- Pb scales had less PbO2- Pb associated with Mn sorbing onto Mn+Fe-rich surface deposit
bull Erratic Pb persisted after full LSLR in areas rich with Mn deposits (and to lesser extent Fe)
bull Galvanized interior plumbing also accumulated dissolved Pb while LSL present
32
33
Manganese
Other studiesbull Dong et al 2000 Mn acted as vector for particulate Pb movementbull Trueman et al 2019 Increasing Mn increased Pb in experimentbull McNeill and Edwards 2004 Mn sequestration by polyphosphate destabilized
Pb scales and caused Pb release
bull Pan et al 2019 Possible benefit by promoting formation of Pb(+4)
Reference Analysis Mn Impact on PbSchock et al 2014 Excavated LSLs
amp historical water Pb data in a US city
bull Promoted amorphous Pb scalebull Inhibited insoluble Pb(+4) scalebull Erratic Pb persisted after full LSLR in areas rich
with Mn deposits
Tully et al 2019 Excavated LSLs amp water Pb data in several Midwest case studies
bull Mn was a component of amorphous pipe scales in several systems
bull Amorphous scales not amenable to determining their adherence to pipe walls their responses to changes in various water chemistry variables or quantitatively calculating their lead release
Itrsquos not just manganese
34
Broad variety of coatings on LSLs analyzed at EPA
Al Mn Fe P and Ca-rich coatings may interfere withorthophosphate pH adjustment or other CCT
35
More water parameters affect metal release
bull pH amp AlkalinityDICbull Concentration and type of
oxidants (disinfectantsdissolved oxygen)
bull ORPcorrosion potentialbull Corrosion Inhibitorsbull Chloridebull Sulfatebull Manganese
bull Temperaturebull Sorptive surfaces downstream of
LSLs (ie galvanized interior pipe)bull Iron (deposition and corrosion)bull Calciumbull Aluminumbull NOM (type amount)bull Mixing from water treatment plants
or sourcesbull Ammoniabull Hydrogen Sulfidebull Microbial activity (nitrification amp
other)
Corrosion Control Treatment is intertwined with all treatments affecting water chemistry
Water Quality Parameters(WQP)
bull pH alkalinity calcium orthophosphate or silica othersbull WQP measurements are important for
bull Process control (measurement frequency needs to be HIGH)bull Keeping treatment from being turned offbull Keeping inhibitor dosages amp pH control from being cut to save $$$bull Keeping required repairs to treatment on tight schedulebull Preventing frequent variations and alternations of water sources
bull Effectiveness of Pb treatment only determinable from tap sampling by combination of different assessment approaches for different lead plumbing sources to which people can be exposed
36
bull OWQP measurements important for process control look at yourdata to understand what they mean and avoid problems
bull OWQP do not predict lead release There are no substitutes fordirectly monitoring lead release
37
M+
M+
M+
M+
Corrosion barrier film hypothesis
Inert barrier film (eg CaCO3)
Pipe Wall
Barrier film would allow tobull Protect ALL pipe materialsbull Monitor corrosion based on one chemical reaction
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
Corrosion and metal release
bull Pure corrosion (ie electrochemical oxidation-reduction reactions) becomes less important when the surface becomes covered with scale
bull Metal solubility or destabilization of the scale at the water contact will be more important (if the scale is porous corrosion may still occur at the base of the scale)
bull Metal solubility varies by factor of 5 to 10 or more across drinking water systems
CCT includes both pure corrosion and control of metal release from the pipe scale
Corrosion typesbull Uniform corrosion
- Materials degradation- Metal release (lead copper etc)
bull Non-uniform corrosion- Pinhole leaks (copper)- Dezincification (brass)- Tuberculation (iron galvanized brass)- Galvanic Corrosion
bull Galvanic corrosion- Soldered joints- Brass devices- Partial LSLs- Any coupling of different metals
Different leaded materials may undergo different corrosion mechanisms depending on composition configuration geometry
water quality and other factors
Pb pipe uniform
Pb pipe galvanic
DeSantis et al 2018
Important factors affecting corrosion and metal release
bull pH and AlkalinityDissolved Inorganic Carbon
bull Type and Concentration of Oxidants (chlorine species dissolved oxygen etc)OxidationReduction Potential
bull Corrosion Inhibitors
bull Chloride to Sulfate Mass Ratio
bull Manganese
pH is master variable
12
-1
0
1
2
0 2 4 6 8 10 12 14
E (V
OLT
S)
pH
IMMUNE
Simplified Pourbaix diagram (EH-pH Diagram)
Desirable pH-EH combinations allow passivation (ie formation ofprotective pipe scales)
pH is master variable
12
PbO2 (plattnerite)
Pb ++
deg s)(3O 2 -2C H) 2)b O 3P O
00
( 2 -- -2) C 43 ( )
CO
b HP O( ( DIC = 18 mg CL
3 b
Pb P Pb = 0010 mgL
Pb metal
-10
13
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
ndash8
ndash6
ndash4
ndash2
2
4
6
8
10
pH
Eh (v
olts
)
Dissolved Inorganic Carbon (DIC) and Total ALKalinity (TALK)
DIC = [CO32-] + [H2CO3
] + [HCO3- ]]
TALK = 2 [CO32-] + [HCO3
-] + [OH-] - [H+]
mg CL Dissolved Inorganic Carbon (DIC)0 10 20 30 40 50
mg
CaCO
3L
Tota
l Alka
linity
0255075
100125150175
200225250
pH 60pH 70pH 80pH 90pH 100
bull To understand corrosion it isimportant to keep up with thecarbonate system
bull DIC and TALK have linearrelationship but are not thesame thing
bullbullbullbullbull
Oxidants in drinking water
bull Disinfection (Free chlorine Chloramine Chlorine dioxide)Pre-(O3 H2O2 ClO2 KMnO4) Oxidative metal removal (eg As Fe Mn) Ammonia removalAeration (corrosion control VOCRn2S removal) Taste and odor control
bull Dissolved oxygen
Oxidants in drinking water
Oxidant Dosage (mgL)0 2 4 6 8 10
E H (V
olts
vs S
HE)
03
04
05
06
07
08
09
10
11
MCA Electrode 1 MCA Electrode 2 KMnO4 Electrode 1 KMnO4 Electrode 2 ClO2 Electrode 1 ClO2 Electrode 2 DO Electrode 1 DO Electrode 2 Cl2 Electrode 1Cl2 Electrode 2
ClO2 HOCldeg
KMnO4NH2Cl
DO
Different oxidants have different oxidizing power
Oxidation-Reduction Potential of several disinfectants in experimentJames et al 2004 (pH 7 10 mg CL 25degC)
17
ldquoClassicrdquo divalent Pb+2 solubility
pH6 7 8 9 10 11
mg
PbL
001
01
1
10
100 1 mg CL 5 mg CL 10 mg CL 20 mg CL 35 mg CL 50 mg CL 75 mg CL100 mg CL
DIC
Optimum pHDIC rangefor LSLs
High pH is needed to minimize Pb solubility
Lead ldquocorrosion Inhibitorsrdquo
bull Phosphoric acidbull Alkalai-metal orthophosphatebull Zinc orthophosphatebull Blended orthopolyphosphatesbull Polyphosphatesbull Sodium Silicate
Polyphosphates are not corrosion inhibitors but rather sequestering agentsEfficiency depends on sufficient doseconcentration for thebackground pH and other reactive water quality constituentswhat if the dose is too low to actually work 18
Illustrative reactions of P in pipebull OrthophosphatePrecipitation (consumption of orthophosphate)3Pb+2 + 2PO4
-3 --gt Pb3(PO4)2 (solid)decreased Pb solubility
bull PolyphosphatesReversion partially (formation of orthophosphate)P3O10
-5 ---gt P2O7-3 + PO4
-3
polymer --gt monomersComplexation (soluble complexes)3Pb+2 + P3O10
-5 --gt PbP3O10-2
higher Pb solubility
Slid
e cr
edit
Mar
c Ed
war
ds
Conceptual possible reactions Actual reactants and products may bedifferent Polyphosphates cannot be easily quantified
20
Ortho-P Treatment for Pb+2
Effectiveness Depends on Dose DIC pH and ldquoCleanlinessrdquo of Pipe Surface
mg PO4L00 10 20 30 40 50
mg
PbL
000
005
010
015
020
025
030
035
040
48 mg CL
48 mg CL
pH = 70pH = 75pH = 80pH = 85
bull pH less critical at low DICbull pH less critical at high PO4bull Point of diminishing returns higher with high DIC
bull Faster Pb reduction at high PO4
Typical UK Dosages 4-6 mgL
Most PWSs with LSLs do not have corrosion control treatment that minimizes Pb release and exposure
Effect of pH and ortho-P on Pb release
21
DIC = 10 mg CL 1 mg PO4L
60 65 70 75 80 85 90 95
gL
b m
P
0001
001
01
1
10
US Action Level
Dissolved Pb with ortho-PTotal Pb with ortho-PDissolved Pb with CO3Total Pb with CO3
At low DIC orthophosphate improves lead release regardless of pH
pHSchock M R DeSantis M K Metz D H Welch M M Hyland R N Nadagouda M N Revisiting the pH Effect on the Orthophosphate Control of Plumbosolvency Proc AWWA Annual Conference and Exposition Atlanta GA 2008
Ortho-P at pH 90 (DIC 6 mgL)
22
Pb(μgL)
bull Low Pb at 1 mgL relative to higher doses needs to be further verified across water qualities
bull Must do dose optimization study for your own water quality especially at high pH
bull Ortho-P may precipitate with Ca
Miller S A Investigation of Lead Solubility and Orthophosphate Addition in High pH Low DIC Water Master of Science Department of Biomedical Chemical and Environmental Engineering College of Engineering and Applied Science University of Cincinnati Cincinnati OH 2014
Ortho-Ppoint of diminishing returns
bull Orthophosphate addition to where large increments result in small reductions in lead release
bull Key to cost-effective lead release control and exposure reduction
bull Becomes the rdquomaintenancerdquo dosage unless the scale ldquoagesrdquo
bull Background constituents such as aluminum hardness ions iron manganese and others interfere with optimum reduction of lead release
bull Varies with the background water chemistry from system to system
23
Sodium silicate
bull No systematic studies to look at pH carbonate silicate background chemistry relationships
bull Unclear mechanism pH increase Buffer intensity increase Protective film (thin amp amorphous)
bull Can sequester ironmanganese
bull Canrsquot be evaluated with fresh surfaces
24pH5 6 7 8 9 10 11
β ty
nsi
nte
Iuf
fer
B
00000
00005
00010TIC=48 mg CL20 mgL SiO250 mgL PO4β H2OTotal β
Silicate may contribute to Buffer Intensity
Carbonate orthophosphate silicate I=001 25ordmC
Chemical changes may reduce Pb+4 to Pb+2
10
8
6
4
)ts 2ol
h (v
E
ndash2
ndash4
ndash6
ndash8
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
pH
12
PbO2 (plattnerite)
Pb++ )s
deg
(2
CO
3
(OH
) -2 2) 3
Pb
-2
0 (CO
-- 2
0
) 43(C
O
Pb
3 (OH
)
DIC = 18 mg CL
Pb Pb Pb = 0010 mgL
Pb metal
-10
Drop in ORP from treatment change or DS
oxidant demand
Drop in pH at surface from treatment change
chemical reactions nitrification etc
C D
ngto
nhisa
W
Newark NJ
Disinfectant demand in DS must be controlled amp enough free chlorine consistently maintained throughout LSL area to keep protective Pb+4 corrosion scales in place
Switching disinfectants may reduce Pb+4 to Pb+2
Low
Switch from chlorine to chloramine disinfectant dissolved lead from pipe scales during the Washington DC ldquolead-in-water crisisrdquo in 2001-2004
Example profile of Pb(+4) (ie PbO2) scale house
27
20
L)
Cl1 Cincinnati OH
18
microg 16
n ( 14
o 12
tia 10
trn 8
e 6
nco 4
C 2
b P 0-1 0 1 2 3 4 5 6 7 8 9 10 11
Cumulative Water Volume (L)
CI-1 10 h 1209CI-1 10 h 0410
Plumbing sequence corresponding to water volumeFaucet LSLCu pipe Brass
Water MainMCu pipe 2
015
a
let
dou
illyaf
ant
irT
bull Pb release to water from PbO2 coated LSLs was found to be lowbull Be aware are there are no known cases of exhumed PbO2 coated pipes being able to
re-stabilize in pipe rigs
Chloride and sulfateCSMR a factor controlling galvanic corrosion (Oliphant 1983 Gregory 1986)- Chloride detrimental- Sulfate beneficial if sufficient to overcome chloride thus- CSMR ratio was created
[Cl- ] 12 mgL Cl-Chloride to Sulfate Mass Ratio (CSMR) = =[SO -2 2 = 06
4 ] 20 mgL SO-4
Confirmed empirically in case studies of water utilities (Dodrill and Edwards 1995 Edwards and Triantafyllidou 2007) - High CSMR gt 05-06 more corrosive- Low CSMR lt 05-06 less corrosive
28
Verified in laboratory experiments with fresh galvanic Pb junctions (eg Wang et al 2013 Triantafyllidou et al 2011 Nguyen et al 2010 2011)
Tria
ntaf
yllid
ou e
t al
201
1
Chloride and sulfate
29
CSMR considered contributing factor to the Flint water Pb crisis (Pieper et al 2017)
30
Chloride and sulfate
bull Absolute concentrations of chloride and sulfate important chloride considered particularly detrimental to galvanic Pb corrosion
bull Increasing chloride and sulfate concentrations at a constant CSMR of 1 increased total lead concentrations in galvanic corrosion experiments (Ng amp Lin 2016)
bull High CSMR ratio proven empirically important in aggravating galvanic lead corrosion (verified by experiments and modeling)
bull Water source or treatment changes that affect the CSMR must be evaluated prior to change
bull Absolute concentrations also important particularly for chloride
31
Manganese
Two distinct pipe scale categories in 5 LSLs (Lead 1-5 Pipes) excavated from same city
31
Schock et al 2014 JAWWAbull The outer pipe scale is in direct contact with the flowing water affecting
water lead contamination the mostbull Plattnerite an insoluble Pb(+4) scale that protects water from lead
contamination was inhibited by Mn (and Fe) presence
Manganese
In this case study (Schock et al JAWWA 2014)bull Where Mn was present in more abundance
- Pb scales had less PbO2- Pb associated with Mn sorbing onto Mn+Fe-rich surface deposit
bull Erratic Pb persisted after full LSLR in areas rich with Mn deposits (and to lesser extent Fe)
bull Galvanized interior plumbing also accumulated dissolved Pb while LSL present
32
33
Manganese
Other studiesbull Dong et al 2000 Mn acted as vector for particulate Pb movementbull Trueman et al 2019 Increasing Mn increased Pb in experimentbull McNeill and Edwards 2004 Mn sequestration by polyphosphate destabilized
Pb scales and caused Pb release
bull Pan et al 2019 Possible benefit by promoting formation of Pb(+4)
Reference Analysis Mn Impact on PbSchock et al 2014 Excavated LSLs
amp historical water Pb data in a US city
bull Promoted amorphous Pb scalebull Inhibited insoluble Pb(+4) scalebull Erratic Pb persisted after full LSLR in areas rich
with Mn deposits
Tully et al 2019 Excavated LSLs amp water Pb data in several Midwest case studies
bull Mn was a component of amorphous pipe scales in several systems
bull Amorphous scales not amenable to determining their adherence to pipe walls their responses to changes in various water chemistry variables or quantitatively calculating their lead release
Itrsquos not just manganese
34
Broad variety of coatings on LSLs analyzed at EPA
Al Mn Fe P and Ca-rich coatings may interfere withorthophosphate pH adjustment or other CCT
35
More water parameters affect metal release
bull pH amp AlkalinityDICbull Concentration and type of
oxidants (disinfectantsdissolved oxygen)
bull ORPcorrosion potentialbull Corrosion Inhibitorsbull Chloridebull Sulfatebull Manganese
bull Temperaturebull Sorptive surfaces downstream of
LSLs (ie galvanized interior pipe)bull Iron (deposition and corrosion)bull Calciumbull Aluminumbull NOM (type amount)bull Mixing from water treatment plants
or sourcesbull Ammoniabull Hydrogen Sulfidebull Microbial activity (nitrification amp
other)
Corrosion Control Treatment is intertwined with all treatments affecting water chemistry
Water Quality Parameters(WQP)
bull pH alkalinity calcium orthophosphate or silica othersbull WQP measurements are important for
bull Process control (measurement frequency needs to be HIGH)bull Keeping treatment from being turned offbull Keeping inhibitor dosages amp pH control from being cut to save $$$bull Keeping required repairs to treatment on tight schedulebull Preventing frequent variations and alternations of water sources
bull Effectiveness of Pb treatment only determinable from tap sampling by combination of different assessment approaches for different lead plumbing sources to which people can be exposed
36
bull OWQP measurements important for process control look at yourdata to understand what they mean and avoid problems
bull OWQP do not predict lead release There are no substitutes fordirectly monitoring lead release
37
M+
M+
M+
M+
Corrosion barrier film hypothesis
Inert barrier film (eg CaCO3)
Pipe Wall
Barrier film would allow tobull Protect ALL pipe materialsbull Monitor corrosion based on one chemical reaction
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
Corrosion typesbull Uniform corrosion
- Materials degradation- Metal release (lead copper etc)
bull Non-uniform corrosion- Pinhole leaks (copper)- Dezincification (brass)- Tuberculation (iron galvanized brass)- Galvanic Corrosion
bull Galvanic corrosion- Soldered joints- Brass devices- Partial LSLs- Any coupling of different metals
Different leaded materials may undergo different corrosion mechanisms depending on composition configuration geometry
water quality and other factors
Pb pipe uniform
Pb pipe galvanic
DeSantis et al 2018
Important factors affecting corrosion and metal release
bull pH and AlkalinityDissolved Inorganic Carbon
bull Type and Concentration of Oxidants (chlorine species dissolved oxygen etc)OxidationReduction Potential
bull Corrosion Inhibitors
bull Chloride to Sulfate Mass Ratio
bull Manganese
pH is master variable
12
-1
0
1
2
0 2 4 6 8 10 12 14
E (V
OLT
S)
pH
IMMUNE
Simplified Pourbaix diagram (EH-pH Diagram)
Desirable pH-EH combinations allow passivation (ie formation ofprotective pipe scales)
pH is master variable
12
PbO2 (plattnerite)
Pb ++
deg s)(3O 2 -2C H) 2)b O 3P O
00
( 2 -- -2) C 43 ( )
CO
b HP O( ( DIC = 18 mg CL
3 b
Pb P Pb = 0010 mgL
Pb metal
-10
13
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
ndash8
ndash6
ndash4
ndash2
2
4
6
8
10
pH
Eh (v
olts
)
Dissolved Inorganic Carbon (DIC) and Total ALKalinity (TALK)
DIC = [CO32-] + [H2CO3
] + [HCO3- ]]
TALK = 2 [CO32-] + [HCO3
-] + [OH-] - [H+]
mg CL Dissolved Inorganic Carbon (DIC)0 10 20 30 40 50
mg
CaCO
3L
Tota
l Alka
linity
0255075
100125150175
200225250
pH 60pH 70pH 80pH 90pH 100
bull To understand corrosion it isimportant to keep up with thecarbonate system
bull DIC and TALK have linearrelationship but are not thesame thing
bullbullbullbullbull
Oxidants in drinking water
bull Disinfection (Free chlorine Chloramine Chlorine dioxide)Pre-(O3 H2O2 ClO2 KMnO4) Oxidative metal removal (eg As Fe Mn) Ammonia removalAeration (corrosion control VOCRn2S removal) Taste and odor control
bull Dissolved oxygen
Oxidants in drinking water
Oxidant Dosage (mgL)0 2 4 6 8 10
E H (V
olts
vs S
HE)
03
04
05
06
07
08
09
10
11
MCA Electrode 1 MCA Electrode 2 KMnO4 Electrode 1 KMnO4 Electrode 2 ClO2 Electrode 1 ClO2 Electrode 2 DO Electrode 1 DO Electrode 2 Cl2 Electrode 1Cl2 Electrode 2
ClO2 HOCldeg
KMnO4NH2Cl
DO
Different oxidants have different oxidizing power
Oxidation-Reduction Potential of several disinfectants in experimentJames et al 2004 (pH 7 10 mg CL 25degC)
17
ldquoClassicrdquo divalent Pb+2 solubility
pH6 7 8 9 10 11
mg
PbL
001
01
1
10
100 1 mg CL 5 mg CL 10 mg CL 20 mg CL 35 mg CL 50 mg CL 75 mg CL100 mg CL
DIC
Optimum pHDIC rangefor LSLs
High pH is needed to minimize Pb solubility
Lead ldquocorrosion Inhibitorsrdquo
bull Phosphoric acidbull Alkalai-metal orthophosphatebull Zinc orthophosphatebull Blended orthopolyphosphatesbull Polyphosphatesbull Sodium Silicate
Polyphosphates are not corrosion inhibitors but rather sequestering agentsEfficiency depends on sufficient doseconcentration for thebackground pH and other reactive water quality constituentswhat if the dose is too low to actually work 18
Illustrative reactions of P in pipebull OrthophosphatePrecipitation (consumption of orthophosphate)3Pb+2 + 2PO4
-3 --gt Pb3(PO4)2 (solid)decreased Pb solubility
bull PolyphosphatesReversion partially (formation of orthophosphate)P3O10
-5 ---gt P2O7-3 + PO4
-3
polymer --gt monomersComplexation (soluble complexes)3Pb+2 + P3O10
-5 --gt PbP3O10-2
higher Pb solubility
Slid
e cr
edit
Mar
c Ed
war
ds
Conceptual possible reactions Actual reactants and products may bedifferent Polyphosphates cannot be easily quantified
20
Ortho-P Treatment for Pb+2
Effectiveness Depends on Dose DIC pH and ldquoCleanlinessrdquo of Pipe Surface
mg PO4L00 10 20 30 40 50
mg
PbL
000
005
010
015
020
025
030
035
040
48 mg CL
48 mg CL
pH = 70pH = 75pH = 80pH = 85
bull pH less critical at low DICbull pH less critical at high PO4bull Point of diminishing returns higher with high DIC
bull Faster Pb reduction at high PO4
Typical UK Dosages 4-6 mgL
Most PWSs with LSLs do not have corrosion control treatment that minimizes Pb release and exposure
Effect of pH and ortho-P on Pb release
21
DIC = 10 mg CL 1 mg PO4L
60 65 70 75 80 85 90 95
gL
b m
P
0001
001
01
1
10
US Action Level
Dissolved Pb with ortho-PTotal Pb with ortho-PDissolved Pb with CO3Total Pb with CO3
At low DIC orthophosphate improves lead release regardless of pH
pHSchock M R DeSantis M K Metz D H Welch M M Hyland R N Nadagouda M N Revisiting the pH Effect on the Orthophosphate Control of Plumbosolvency Proc AWWA Annual Conference and Exposition Atlanta GA 2008
Ortho-P at pH 90 (DIC 6 mgL)
22
Pb(μgL)
bull Low Pb at 1 mgL relative to higher doses needs to be further verified across water qualities
bull Must do dose optimization study for your own water quality especially at high pH
bull Ortho-P may precipitate with Ca
Miller S A Investigation of Lead Solubility and Orthophosphate Addition in High pH Low DIC Water Master of Science Department of Biomedical Chemical and Environmental Engineering College of Engineering and Applied Science University of Cincinnati Cincinnati OH 2014
Ortho-Ppoint of diminishing returns
bull Orthophosphate addition to where large increments result in small reductions in lead release
bull Key to cost-effective lead release control and exposure reduction
bull Becomes the rdquomaintenancerdquo dosage unless the scale ldquoagesrdquo
bull Background constituents such as aluminum hardness ions iron manganese and others interfere with optimum reduction of lead release
bull Varies with the background water chemistry from system to system
23
Sodium silicate
bull No systematic studies to look at pH carbonate silicate background chemistry relationships
bull Unclear mechanism pH increase Buffer intensity increase Protective film (thin amp amorphous)
bull Can sequester ironmanganese
bull Canrsquot be evaluated with fresh surfaces
24pH5 6 7 8 9 10 11
β ty
nsi
nte
Iuf
fer
B
00000
00005
00010TIC=48 mg CL20 mgL SiO250 mgL PO4β H2OTotal β
Silicate may contribute to Buffer Intensity
Carbonate orthophosphate silicate I=001 25ordmC
Chemical changes may reduce Pb+4 to Pb+2
10
8
6
4
)ts 2ol
h (v
E
ndash2
ndash4
ndash6
ndash8
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
pH
12
PbO2 (plattnerite)
Pb++ )s
deg
(2
CO
3
(OH
) -2 2) 3
Pb
-2
0 (CO
-- 2
0
) 43(C
O
Pb
3 (OH
)
DIC = 18 mg CL
Pb Pb Pb = 0010 mgL
Pb metal
-10
Drop in ORP from treatment change or DS
oxidant demand
Drop in pH at surface from treatment change
chemical reactions nitrification etc
C D
ngto
nhisa
W
Newark NJ
Disinfectant demand in DS must be controlled amp enough free chlorine consistently maintained throughout LSL area to keep protective Pb+4 corrosion scales in place
Switching disinfectants may reduce Pb+4 to Pb+2
Low
Switch from chlorine to chloramine disinfectant dissolved lead from pipe scales during the Washington DC ldquolead-in-water crisisrdquo in 2001-2004
Example profile of Pb(+4) (ie PbO2) scale house
27
20
L)
Cl1 Cincinnati OH
18
microg 16
n ( 14
o 12
tia 10
trn 8
e 6
nco 4
C 2
b P 0-1 0 1 2 3 4 5 6 7 8 9 10 11
Cumulative Water Volume (L)
CI-1 10 h 1209CI-1 10 h 0410
Plumbing sequence corresponding to water volumeFaucet LSLCu pipe Brass
Water MainMCu pipe 2
015
a
let
dou
illyaf
ant
irT
bull Pb release to water from PbO2 coated LSLs was found to be lowbull Be aware are there are no known cases of exhumed PbO2 coated pipes being able to
re-stabilize in pipe rigs
Chloride and sulfateCSMR a factor controlling galvanic corrosion (Oliphant 1983 Gregory 1986)- Chloride detrimental- Sulfate beneficial if sufficient to overcome chloride thus- CSMR ratio was created
[Cl- ] 12 mgL Cl-Chloride to Sulfate Mass Ratio (CSMR) = =[SO -2 2 = 06
4 ] 20 mgL SO-4
Confirmed empirically in case studies of water utilities (Dodrill and Edwards 1995 Edwards and Triantafyllidou 2007) - High CSMR gt 05-06 more corrosive- Low CSMR lt 05-06 less corrosive
28
Verified in laboratory experiments with fresh galvanic Pb junctions (eg Wang et al 2013 Triantafyllidou et al 2011 Nguyen et al 2010 2011)
Tria
ntaf
yllid
ou e
t al
201
1
Chloride and sulfate
29
CSMR considered contributing factor to the Flint water Pb crisis (Pieper et al 2017)
30
Chloride and sulfate
bull Absolute concentrations of chloride and sulfate important chloride considered particularly detrimental to galvanic Pb corrosion
bull Increasing chloride and sulfate concentrations at a constant CSMR of 1 increased total lead concentrations in galvanic corrosion experiments (Ng amp Lin 2016)
bull High CSMR ratio proven empirically important in aggravating galvanic lead corrosion (verified by experiments and modeling)
bull Water source or treatment changes that affect the CSMR must be evaluated prior to change
bull Absolute concentrations also important particularly for chloride
31
Manganese
Two distinct pipe scale categories in 5 LSLs (Lead 1-5 Pipes) excavated from same city
31
Schock et al 2014 JAWWAbull The outer pipe scale is in direct contact with the flowing water affecting
water lead contamination the mostbull Plattnerite an insoluble Pb(+4) scale that protects water from lead
contamination was inhibited by Mn (and Fe) presence
Manganese
In this case study (Schock et al JAWWA 2014)bull Where Mn was present in more abundance
- Pb scales had less PbO2- Pb associated with Mn sorbing onto Mn+Fe-rich surface deposit
bull Erratic Pb persisted after full LSLR in areas rich with Mn deposits (and to lesser extent Fe)
bull Galvanized interior plumbing also accumulated dissolved Pb while LSL present
32
33
Manganese
Other studiesbull Dong et al 2000 Mn acted as vector for particulate Pb movementbull Trueman et al 2019 Increasing Mn increased Pb in experimentbull McNeill and Edwards 2004 Mn sequestration by polyphosphate destabilized
Pb scales and caused Pb release
bull Pan et al 2019 Possible benefit by promoting formation of Pb(+4)
Reference Analysis Mn Impact on PbSchock et al 2014 Excavated LSLs
amp historical water Pb data in a US city
bull Promoted amorphous Pb scalebull Inhibited insoluble Pb(+4) scalebull Erratic Pb persisted after full LSLR in areas rich
with Mn deposits
Tully et al 2019 Excavated LSLs amp water Pb data in several Midwest case studies
bull Mn was a component of amorphous pipe scales in several systems
bull Amorphous scales not amenable to determining their adherence to pipe walls their responses to changes in various water chemistry variables or quantitatively calculating their lead release
Itrsquos not just manganese
34
Broad variety of coatings on LSLs analyzed at EPA
Al Mn Fe P and Ca-rich coatings may interfere withorthophosphate pH adjustment or other CCT
35
More water parameters affect metal release
bull pH amp AlkalinityDICbull Concentration and type of
oxidants (disinfectantsdissolved oxygen)
bull ORPcorrosion potentialbull Corrosion Inhibitorsbull Chloridebull Sulfatebull Manganese
bull Temperaturebull Sorptive surfaces downstream of
LSLs (ie galvanized interior pipe)bull Iron (deposition and corrosion)bull Calciumbull Aluminumbull NOM (type amount)bull Mixing from water treatment plants
or sourcesbull Ammoniabull Hydrogen Sulfidebull Microbial activity (nitrification amp
other)
Corrosion Control Treatment is intertwined with all treatments affecting water chemistry
Water Quality Parameters(WQP)
bull pH alkalinity calcium orthophosphate or silica othersbull WQP measurements are important for
bull Process control (measurement frequency needs to be HIGH)bull Keeping treatment from being turned offbull Keeping inhibitor dosages amp pH control from being cut to save $$$bull Keeping required repairs to treatment on tight schedulebull Preventing frequent variations and alternations of water sources
bull Effectiveness of Pb treatment only determinable from tap sampling by combination of different assessment approaches for different lead plumbing sources to which people can be exposed
36
bull OWQP measurements important for process control look at yourdata to understand what they mean and avoid problems
bull OWQP do not predict lead release There are no substitutes fordirectly monitoring lead release
37
M+
M+
M+
M+
Corrosion barrier film hypothesis
Inert barrier film (eg CaCO3)
Pipe Wall
Barrier film would allow tobull Protect ALL pipe materialsbull Monitor corrosion based on one chemical reaction
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
Important factors affecting corrosion and metal release
bull pH and AlkalinityDissolved Inorganic Carbon
bull Type and Concentration of Oxidants (chlorine species dissolved oxygen etc)OxidationReduction Potential
bull Corrosion Inhibitors
bull Chloride to Sulfate Mass Ratio
bull Manganese
pH is master variable
12
-1
0
1
2
0 2 4 6 8 10 12 14
E (V
OLT
S)
pH
IMMUNE
Simplified Pourbaix diagram (EH-pH Diagram)
Desirable pH-EH combinations allow passivation (ie formation ofprotective pipe scales)
pH is master variable
12
PbO2 (plattnerite)
Pb ++
deg s)(3O 2 -2C H) 2)b O 3P O
00
( 2 -- -2) C 43 ( )
CO
b HP O( ( DIC = 18 mg CL
3 b
Pb P Pb = 0010 mgL
Pb metal
-10
13
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
ndash8
ndash6
ndash4
ndash2
2
4
6
8
10
pH
Eh (v
olts
)
Dissolved Inorganic Carbon (DIC) and Total ALKalinity (TALK)
DIC = [CO32-] + [H2CO3
] + [HCO3- ]]
TALK = 2 [CO32-] + [HCO3
-] + [OH-] - [H+]
mg CL Dissolved Inorganic Carbon (DIC)0 10 20 30 40 50
mg
CaCO
3L
Tota
l Alka
linity
0255075
100125150175
200225250
pH 60pH 70pH 80pH 90pH 100
bull To understand corrosion it isimportant to keep up with thecarbonate system
bull DIC and TALK have linearrelationship but are not thesame thing
bullbullbullbullbull
Oxidants in drinking water
bull Disinfection (Free chlorine Chloramine Chlorine dioxide)Pre-(O3 H2O2 ClO2 KMnO4) Oxidative metal removal (eg As Fe Mn) Ammonia removalAeration (corrosion control VOCRn2S removal) Taste and odor control
bull Dissolved oxygen
Oxidants in drinking water
Oxidant Dosage (mgL)0 2 4 6 8 10
E H (V
olts
vs S
HE)
03
04
05
06
07
08
09
10
11
MCA Electrode 1 MCA Electrode 2 KMnO4 Electrode 1 KMnO4 Electrode 2 ClO2 Electrode 1 ClO2 Electrode 2 DO Electrode 1 DO Electrode 2 Cl2 Electrode 1Cl2 Electrode 2
ClO2 HOCldeg
KMnO4NH2Cl
DO
Different oxidants have different oxidizing power
Oxidation-Reduction Potential of several disinfectants in experimentJames et al 2004 (pH 7 10 mg CL 25degC)
17
ldquoClassicrdquo divalent Pb+2 solubility
pH6 7 8 9 10 11
mg
PbL
001
01
1
10
100 1 mg CL 5 mg CL 10 mg CL 20 mg CL 35 mg CL 50 mg CL 75 mg CL100 mg CL
DIC
Optimum pHDIC rangefor LSLs
High pH is needed to minimize Pb solubility
Lead ldquocorrosion Inhibitorsrdquo
bull Phosphoric acidbull Alkalai-metal orthophosphatebull Zinc orthophosphatebull Blended orthopolyphosphatesbull Polyphosphatesbull Sodium Silicate
Polyphosphates are not corrosion inhibitors but rather sequestering agentsEfficiency depends on sufficient doseconcentration for thebackground pH and other reactive water quality constituentswhat if the dose is too low to actually work 18
Illustrative reactions of P in pipebull OrthophosphatePrecipitation (consumption of orthophosphate)3Pb+2 + 2PO4
-3 --gt Pb3(PO4)2 (solid)decreased Pb solubility
bull PolyphosphatesReversion partially (formation of orthophosphate)P3O10
-5 ---gt P2O7-3 + PO4
-3
polymer --gt monomersComplexation (soluble complexes)3Pb+2 + P3O10
-5 --gt PbP3O10-2
higher Pb solubility
Slid
e cr
edit
Mar
c Ed
war
ds
Conceptual possible reactions Actual reactants and products may bedifferent Polyphosphates cannot be easily quantified
20
Ortho-P Treatment for Pb+2
Effectiveness Depends on Dose DIC pH and ldquoCleanlinessrdquo of Pipe Surface
mg PO4L00 10 20 30 40 50
mg
PbL
000
005
010
015
020
025
030
035
040
48 mg CL
48 mg CL
pH = 70pH = 75pH = 80pH = 85
bull pH less critical at low DICbull pH less critical at high PO4bull Point of diminishing returns higher with high DIC
bull Faster Pb reduction at high PO4
Typical UK Dosages 4-6 mgL
Most PWSs with LSLs do not have corrosion control treatment that minimizes Pb release and exposure
Effect of pH and ortho-P on Pb release
21
DIC = 10 mg CL 1 mg PO4L
60 65 70 75 80 85 90 95
gL
b m
P
0001
001
01
1
10
US Action Level
Dissolved Pb with ortho-PTotal Pb with ortho-PDissolved Pb with CO3Total Pb with CO3
At low DIC orthophosphate improves lead release regardless of pH
pHSchock M R DeSantis M K Metz D H Welch M M Hyland R N Nadagouda M N Revisiting the pH Effect on the Orthophosphate Control of Plumbosolvency Proc AWWA Annual Conference and Exposition Atlanta GA 2008
Ortho-P at pH 90 (DIC 6 mgL)
22
Pb(μgL)
bull Low Pb at 1 mgL relative to higher doses needs to be further verified across water qualities
bull Must do dose optimization study for your own water quality especially at high pH
bull Ortho-P may precipitate with Ca
Miller S A Investigation of Lead Solubility and Orthophosphate Addition in High pH Low DIC Water Master of Science Department of Biomedical Chemical and Environmental Engineering College of Engineering and Applied Science University of Cincinnati Cincinnati OH 2014
Ortho-Ppoint of diminishing returns
bull Orthophosphate addition to where large increments result in small reductions in lead release
bull Key to cost-effective lead release control and exposure reduction
bull Becomes the rdquomaintenancerdquo dosage unless the scale ldquoagesrdquo
bull Background constituents such as aluminum hardness ions iron manganese and others interfere with optimum reduction of lead release
bull Varies with the background water chemistry from system to system
23
Sodium silicate
bull No systematic studies to look at pH carbonate silicate background chemistry relationships
bull Unclear mechanism pH increase Buffer intensity increase Protective film (thin amp amorphous)
bull Can sequester ironmanganese
bull Canrsquot be evaluated with fresh surfaces
24pH5 6 7 8 9 10 11
β ty
nsi
nte
Iuf
fer
B
00000
00005
00010TIC=48 mg CL20 mgL SiO250 mgL PO4β H2OTotal β
Silicate may contribute to Buffer Intensity
Carbonate orthophosphate silicate I=001 25ordmC
Chemical changes may reduce Pb+4 to Pb+2
10
8
6
4
)ts 2ol
h (v
E
ndash2
ndash4
ndash6
ndash8
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
pH
12
PbO2 (plattnerite)
Pb++ )s
deg
(2
CO
3
(OH
) -2 2) 3
Pb
-2
0 (CO
-- 2
0
) 43(C
O
Pb
3 (OH
)
DIC = 18 mg CL
Pb Pb Pb = 0010 mgL
Pb metal
-10
Drop in ORP from treatment change or DS
oxidant demand
Drop in pH at surface from treatment change
chemical reactions nitrification etc
C D
ngto
nhisa
W
Newark NJ
Disinfectant demand in DS must be controlled amp enough free chlorine consistently maintained throughout LSL area to keep protective Pb+4 corrosion scales in place
Switching disinfectants may reduce Pb+4 to Pb+2
Low
Switch from chlorine to chloramine disinfectant dissolved lead from pipe scales during the Washington DC ldquolead-in-water crisisrdquo in 2001-2004
Example profile of Pb(+4) (ie PbO2) scale house
27
20
L)
Cl1 Cincinnati OH
18
microg 16
n ( 14
o 12
tia 10
trn 8
e 6
nco 4
C 2
b P 0-1 0 1 2 3 4 5 6 7 8 9 10 11
Cumulative Water Volume (L)
CI-1 10 h 1209CI-1 10 h 0410
Plumbing sequence corresponding to water volumeFaucet LSLCu pipe Brass
Water MainMCu pipe 2
015
a
let
dou
illyaf
ant
irT
bull Pb release to water from PbO2 coated LSLs was found to be lowbull Be aware are there are no known cases of exhumed PbO2 coated pipes being able to
re-stabilize in pipe rigs
Chloride and sulfateCSMR a factor controlling galvanic corrosion (Oliphant 1983 Gregory 1986)- Chloride detrimental- Sulfate beneficial if sufficient to overcome chloride thus- CSMR ratio was created
[Cl- ] 12 mgL Cl-Chloride to Sulfate Mass Ratio (CSMR) = =[SO -2 2 = 06
4 ] 20 mgL SO-4
Confirmed empirically in case studies of water utilities (Dodrill and Edwards 1995 Edwards and Triantafyllidou 2007) - High CSMR gt 05-06 more corrosive- Low CSMR lt 05-06 less corrosive
28
Verified in laboratory experiments with fresh galvanic Pb junctions (eg Wang et al 2013 Triantafyllidou et al 2011 Nguyen et al 2010 2011)
Tria
ntaf
yllid
ou e
t al
201
1
Chloride and sulfate
29
CSMR considered contributing factor to the Flint water Pb crisis (Pieper et al 2017)
30
Chloride and sulfate
bull Absolute concentrations of chloride and sulfate important chloride considered particularly detrimental to galvanic Pb corrosion
bull Increasing chloride and sulfate concentrations at a constant CSMR of 1 increased total lead concentrations in galvanic corrosion experiments (Ng amp Lin 2016)
bull High CSMR ratio proven empirically important in aggravating galvanic lead corrosion (verified by experiments and modeling)
bull Water source or treatment changes that affect the CSMR must be evaluated prior to change
bull Absolute concentrations also important particularly for chloride
31
Manganese
Two distinct pipe scale categories in 5 LSLs (Lead 1-5 Pipes) excavated from same city
31
Schock et al 2014 JAWWAbull The outer pipe scale is in direct contact with the flowing water affecting
water lead contamination the mostbull Plattnerite an insoluble Pb(+4) scale that protects water from lead
contamination was inhibited by Mn (and Fe) presence
Manganese
In this case study (Schock et al JAWWA 2014)bull Where Mn was present in more abundance
- Pb scales had less PbO2- Pb associated with Mn sorbing onto Mn+Fe-rich surface deposit
bull Erratic Pb persisted after full LSLR in areas rich with Mn deposits (and to lesser extent Fe)
bull Galvanized interior plumbing also accumulated dissolved Pb while LSL present
32
33
Manganese
Other studiesbull Dong et al 2000 Mn acted as vector for particulate Pb movementbull Trueman et al 2019 Increasing Mn increased Pb in experimentbull McNeill and Edwards 2004 Mn sequestration by polyphosphate destabilized
Pb scales and caused Pb release
bull Pan et al 2019 Possible benefit by promoting formation of Pb(+4)
Reference Analysis Mn Impact on PbSchock et al 2014 Excavated LSLs
amp historical water Pb data in a US city
bull Promoted amorphous Pb scalebull Inhibited insoluble Pb(+4) scalebull Erratic Pb persisted after full LSLR in areas rich
with Mn deposits
Tully et al 2019 Excavated LSLs amp water Pb data in several Midwest case studies
bull Mn was a component of amorphous pipe scales in several systems
bull Amorphous scales not amenable to determining their adherence to pipe walls their responses to changes in various water chemistry variables or quantitatively calculating their lead release
Itrsquos not just manganese
34
Broad variety of coatings on LSLs analyzed at EPA
Al Mn Fe P and Ca-rich coatings may interfere withorthophosphate pH adjustment or other CCT
35
More water parameters affect metal release
bull pH amp AlkalinityDICbull Concentration and type of
oxidants (disinfectantsdissolved oxygen)
bull ORPcorrosion potentialbull Corrosion Inhibitorsbull Chloridebull Sulfatebull Manganese
bull Temperaturebull Sorptive surfaces downstream of
LSLs (ie galvanized interior pipe)bull Iron (deposition and corrosion)bull Calciumbull Aluminumbull NOM (type amount)bull Mixing from water treatment plants
or sourcesbull Ammoniabull Hydrogen Sulfidebull Microbial activity (nitrification amp
other)
Corrosion Control Treatment is intertwined with all treatments affecting water chemistry
Water Quality Parameters(WQP)
bull pH alkalinity calcium orthophosphate or silica othersbull WQP measurements are important for
bull Process control (measurement frequency needs to be HIGH)bull Keeping treatment from being turned offbull Keeping inhibitor dosages amp pH control from being cut to save $$$bull Keeping required repairs to treatment on tight schedulebull Preventing frequent variations and alternations of water sources
bull Effectiveness of Pb treatment only determinable from tap sampling by combination of different assessment approaches for different lead plumbing sources to which people can be exposed
36
bull OWQP measurements important for process control look at yourdata to understand what they mean and avoid problems
bull OWQP do not predict lead release There are no substitutes fordirectly monitoring lead release
37
M+
M+
M+
M+
Corrosion barrier film hypothesis
Inert barrier film (eg CaCO3)
Pipe Wall
Barrier film would allow tobull Protect ALL pipe materialsbull Monitor corrosion based on one chemical reaction
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
pH is master variable
12
-1
0
1
2
0 2 4 6 8 10 12 14
E (V
OLT
S)
pH
IMMUNE
Simplified Pourbaix diagram (EH-pH Diagram)
Desirable pH-EH combinations allow passivation (ie formation ofprotective pipe scales)
pH is master variable
12
PbO2 (plattnerite)
Pb ++
deg s)(3O 2 -2C H) 2)b O 3P O
00
( 2 -- -2) C 43 ( )
CO
b HP O( ( DIC = 18 mg CL
3 b
Pb P Pb = 0010 mgL
Pb metal
-10
13
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
ndash8
ndash6
ndash4
ndash2
2
4
6
8
10
pH
Eh (v
olts
)
Dissolved Inorganic Carbon (DIC) and Total ALKalinity (TALK)
DIC = [CO32-] + [H2CO3
] + [HCO3- ]]
TALK = 2 [CO32-] + [HCO3
-] + [OH-] - [H+]
mg CL Dissolved Inorganic Carbon (DIC)0 10 20 30 40 50
mg
CaCO
3L
Tota
l Alka
linity
0255075
100125150175
200225250
pH 60pH 70pH 80pH 90pH 100
bull To understand corrosion it isimportant to keep up with thecarbonate system
bull DIC and TALK have linearrelationship but are not thesame thing
bullbullbullbullbull
Oxidants in drinking water
bull Disinfection (Free chlorine Chloramine Chlorine dioxide)Pre-(O3 H2O2 ClO2 KMnO4) Oxidative metal removal (eg As Fe Mn) Ammonia removalAeration (corrosion control VOCRn2S removal) Taste and odor control
bull Dissolved oxygen
Oxidants in drinking water
Oxidant Dosage (mgL)0 2 4 6 8 10
E H (V
olts
vs S
HE)
03
04
05
06
07
08
09
10
11
MCA Electrode 1 MCA Electrode 2 KMnO4 Electrode 1 KMnO4 Electrode 2 ClO2 Electrode 1 ClO2 Electrode 2 DO Electrode 1 DO Electrode 2 Cl2 Electrode 1Cl2 Electrode 2
ClO2 HOCldeg
KMnO4NH2Cl
DO
Different oxidants have different oxidizing power
Oxidation-Reduction Potential of several disinfectants in experimentJames et al 2004 (pH 7 10 mg CL 25degC)
17
ldquoClassicrdquo divalent Pb+2 solubility
pH6 7 8 9 10 11
mg
PbL
001
01
1
10
100 1 mg CL 5 mg CL 10 mg CL 20 mg CL 35 mg CL 50 mg CL 75 mg CL100 mg CL
DIC
Optimum pHDIC rangefor LSLs
High pH is needed to minimize Pb solubility
Lead ldquocorrosion Inhibitorsrdquo
bull Phosphoric acidbull Alkalai-metal orthophosphatebull Zinc orthophosphatebull Blended orthopolyphosphatesbull Polyphosphatesbull Sodium Silicate
Polyphosphates are not corrosion inhibitors but rather sequestering agentsEfficiency depends on sufficient doseconcentration for thebackground pH and other reactive water quality constituentswhat if the dose is too low to actually work 18
Illustrative reactions of P in pipebull OrthophosphatePrecipitation (consumption of orthophosphate)3Pb+2 + 2PO4
-3 --gt Pb3(PO4)2 (solid)decreased Pb solubility
bull PolyphosphatesReversion partially (formation of orthophosphate)P3O10
-5 ---gt P2O7-3 + PO4
-3
polymer --gt monomersComplexation (soluble complexes)3Pb+2 + P3O10
-5 --gt PbP3O10-2
higher Pb solubility
Slid
e cr
edit
Mar
c Ed
war
ds
Conceptual possible reactions Actual reactants and products may bedifferent Polyphosphates cannot be easily quantified
20
Ortho-P Treatment for Pb+2
Effectiveness Depends on Dose DIC pH and ldquoCleanlinessrdquo of Pipe Surface
mg PO4L00 10 20 30 40 50
mg
PbL
000
005
010
015
020
025
030
035
040
48 mg CL
48 mg CL
pH = 70pH = 75pH = 80pH = 85
bull pH less critical at low DICbull pH less critical at high PO4bull Point of diminishing returns higher with high DIC
bull Faster Pb reduction at high PO4
Typical UK Dosages 4-6 mgL
Most PWSs with LSLs do not have corrosion control treatment that minimizes Pb release and exposure
Effect of pH and ortho-P on Pb release
21
DIC = 10 mg CL 1 mg PO4L
60 65 70 75 80 85 90 95
gL
b m
P
0001
001
01
1
10
US Action Level
Dissolved Pb with ortho-PTotal Pb with ortho-PDissolved Pb with CO3Total Pb with CO3
At low DIC orthophosphate improves lead release regardless of pH
pHSchock M R DeSantis M K Metz D H Welch M M Hyland R N Nadagouda M N Revisiting the pH Effect on the Orthophosphate Control of Plumbosolvency Proc AWWA Annual Conference and Exposition Atlanta GA 2008
Ortho-P at pH 90 (DIC 6 mgL)
22
Pb(μgL)
bull Low Pb at 1 mgL relative to higher doses needs to be further verified across water qualities
bull Must do dose optimization study for your own water quality especially at high pH
bull Ortho-P may precipitate with Ca
Miller S A Investigation of Lead Solubility and Orthophosphate Addition in High pH Low DIC Water Master of Science Department of Biomedical Chemical and Environmental Engineering College of Engineering and Applied Science University of Cincinnati Cincinnati OH 2014
Ortho-Ppoint of diminishing returns
bull Orthophosphate addition to where large increments result in small reductions in lead release
bull Key to cost-effective lead release control and exposure reduction
bull Becomes the rdquomaintenancerdquo dosage unless the scale ldquoagesrdquo
bull Background constituents such as aluminum hardness ions iron manganese and others interfere with optimum reduction of lead release
bull Varies with the background water chemistry from system to system
23
Sodium silicate
bull No systematic studies to look at pH carbonate silicate background chemistry relationships
bull Unclear mechanism pH increase Buffer intensity increase Protective film (thin amp amorphous)
bull Can sequester ironmanganese
bull Canrsquot be evaluated with fresh surfaces
24pH5 6 7 8 9 10 11
β ty
nsi
nte
Iuf
fer
B
00000
00005
00010TIC=48 mg CL20 mgL SiO250 mgL PO4β H2OTotal β
Silicate may contribute to Buffer Intensity
Carbonate orthophosphate silicate I=001 25ordmC
Chemical changes may reduce Pb+4 to Pb+2
10
8
6
4
)ts 2ol
h (v
E
ndash2
ndash4
ndash6
ndash8
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
pH
12
PbO2 (plattnerite)
Pb++ )s
deg
(2
CO
3
(OH
) -2 2) 3
Pb
-2
0 (CO
-- 2
0
) 43(C
O
Pb
3 (OH
)
DIC = 18 mg CL
Pb Pb Pb = 0010 mgL
Pb metal
-10
Drop in ORP from treatment change or DS
oxidant demand
Drop in pH at surface from treatment change
chemical reactions nitrification etc
C D
ngto
nhisa
W
Newark NJ
Disinfectant demand in DS must be controlled amp enough free chlorine consistently maintained throughout LSL area to keep protective Pb+4 corrosion scales in place
Switching disinfectants may reduce Pb+4 to Pb+2
Low
Switch from chlorine to chloramine disinfectant dissolved lead from pipe scales during the Washington DC ldquolead-in-water crisisrdquo in 2001-2004
Example profile of Pb(+4) (ie PbO2) scale house
27
20
L)
Cl1 Cincinnati OH
18
microg 16
n ( 14
o 12
tia 10
trn 8
e 6
nco 4
C 2
b P 0-1 0 1 2 3 4 5 6 7 8 9 10 11
Cumulative Water Volume (L)
CI-1 10 h 1209CI-1 10 h 0410
Plumbing sequence corresponding to water volumeFaucet LSLCu pipe Brass
Water MainMCu pipe 2
015
a
let
dou
illyaf
ant
irT
bull Pb release to water from PbO2 coated LSLs was found to be lowbull Be aware are there are no known cases of exhumed PbO2 coated pipes being able to
re-stabilize in pipe rigs
Chloride and sulfateCSMR a factor controlling galvanic corrosion (Oliphant 1983 Gregory 1986)- Chloride detrimental- Sulfate beneficial if sufficient to overcome chloride thus- CSMR ratio was created
[Cl- ] 12 mgL Cl-Chloride to Sulfate Mass Ratio (CSMR) = =[SO -2 2 = 06
4 ] 20 mgL SO-4
Confirmed empirically in case studies of water utilities (Dodrill and Edwards 1995 Edwards and Triantafyllidou 2007) - High CSMR gt 05-06 more corrosive- Low CSMR lt 05-06 less corrosive
28
Verified in laboratory experiments with fresh galvanic Pb junctions (eg Wang et al 2013 Triantafyllidou et al 2011 Nguyen et al 2010 2011)
Tria
ntaf
yllid
ou e
t al
201
1
Chloride and sulfate
29
CSMR considered contributing factor to the Flint water Pb crisis (Pieper et al 2017)
30
Chloride and sulfate
bull Absolute concentrations of chloride and sulfate important chloride considered particularly detrimental to galvanic Pb corrosion
bull Increasing chloride and sulfate concentrations at a constant CSMR of 1 increased total lead concentrations in galvanic corrosion experiments (Ng amp Lin 2016)
bull High CSMR ratio proven empirically important in aggravating galvanic lead corrosion (verified by experiments and modeling)
bull Water source or treatment changes that affect the CSMR must be evaluated prior to change
bull Absolute concentrations also important particularly for chloride
31
Manganese
Two distinct pipe scale categories in 5 LSLs (Lead 1-5 Pipes) excavated from same city
31
Schock et al 2014 JAWWAbull The outer pipe scale is in direct contact with the flowing water affecting
water lead contamination the mostbull Plattnerite an insoluble Pb(+4) scale that protects water from lead
contamination was inhibited by Mn (and Fe) presence
Manganese
In this case study (Schock et al JAWWA 2014)bull Where Mn was present in more abundance
- Pb scales had less PbO2- Pb associated with Mn sorbing onto Mn+Fe-rich surface deposit
bull Erratic Pb persisted after full LSLR in areas rich with Mn deposits (and to lesser extent Fe)
bull Galvanized interior plumbing also accumulated dissolved Pb while LSL present
32
33
Manganese
Other studiesbull Dong et al 2000 Mn acted as vector for particulate Pb movementbull Trueman et al 2019 Increasing Mn increased Pb in experimentbull McNeill and Edwards 2004 Mn sequestration by polyphosphate destabilized
Pb scales and caused Pb release
bull Pan et al 2019 Possible benefit by promoting formation of Pb(+4)
Reference Analysis Mn Impact on PbSchock et al 2014 Excavated LSLs
amp historical water Pb data in a US city
bull Promoted amorphous Pb scalebull Inhibited insoluble Pb(+4) scalebull Erratic Pb persisted after full LSLR in areas rich
with Mn deposits
Tully et al 2019 Excavated LSLs amp water Pb data in several Midwest case studies
bull Mn was a component of amorphous pipe scales in several systems
bull Amorphous scales not amenable to determining their adherence to pipe walls their responses to changes in various water chemistry variables or quantitatively calculating their lead release
Itrsquos not just manganese
34
Broad variety of coatings on LSLs analyzed at EPA
Al Mn Fe P and Ca-rich coatings may interfere withorthophosphate pH adjustment or other CCT
35
More water parameters affect metal release
bull pH amp AlkalinityDICbull Concentration and type of
oxidants (disinfectantsdissolved oxygen)
bull ORPcorrosion potentialbull Corrosion Inhibitorsbull Chloridebull Sulfatebull Manganese
bull Temperaturebull Sorptive surfaces downstream of
LSLs (ie galvanized interior pipe)bull Iron (deposition and corrosion)bull Calciumbull Aluminumbull NOM (type amount)bull Mixing from water treatment plants
or sourcesbull Ammoniabull Hydrogen Sulfidebull Microbial activity (nitrification amp
other)
Corrosion Control Treatment is intertwined with all treatments affecting water chemistry
Water Quality Parameters(WQP)
bull pH alkalinity calcium orthophosphate or silica othersbull WQP measurements are important for
bull Process control (measurement frequency needs to be HIGH)bull Keeping treatment from being turned offbull Keeping inhibitor dosages amp pH control from being cut to save $$$bull Keeping required repairs to treatment on tight schedulebull Preventing frequent variations and alternations of water sources
bull Effectiveness of Pb treatment only determinable from tap sampling by combination of different assessment approaches for different lead plumbing sources to which people can be exposed
36
bull OWQP measurements important for process control look at yourdata to understand what they mean and avoid problems
bull OWQP do not predict lead release There are no substitutes fordirectly monitoring lead release
37
M+
M+
M+
M+
Corrosion barrier film hypothesis
Inert barrier film (eg CaCO3)
Pipe Wall
Barrier film would allow tobull Protect ALL pipe materialsbull Monitor corrosion based on one chemical reaction
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
pH is master variable
12
PbO2 (plattnerite)
Pb ++
deg s)(3O 2 -2C H) 2)b O 3P O
00
( 2 -- -2) C 43 ( )
CO
b HP O( ( DIC = 18 mg CL
3 b
Pb P Pb = 0010 mgL
Pb metal
-10
13
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
ndash8
ndash6
ndash4
ndash2
2
4
6
8
10
pH
Eh (v
olts
)
Dissolved Inorganic Carbon (DIC) and Total ALKalinity (TALK)
DIC = [CO32-] + [H2CO3
] + [HCO3- ]]
TALK = 2 [CO32-] + [HCO3
-] + [OH-] - [H+]
mg CL Dissolved Inorganic Carbon (DIC)0 10 20 30 40 50
mg
CaCO
3L
Tota
l Alka
linity
0255075
100125150175
200225250
pH 60pH 70pH 80pH 90pH 100
bull To understand corrosion it isimportant to keep up with thecarbonate system
bull DIC and TALK have linearrelationship but are not thesame thing
bullbullbullbullbull
Oxidants in drinking water
bull Disinfection (Free chlorine Chloramine Chlorine dioxide)Pre-(O3 H2O2 ClO2 KMnO4) Oxidative metal removal (eg As Fe Mn) Ammonia removalAeration (corrosion control VOCRn2S removal) Taste and odor control
bull Dissolved oxygen
Oxidants in drinking water
Oxidant Dosage (mgL)0 2 4 6 8 10
E H (V
olts
vs S
HE)
03
04
05
06
07
08
09
10
11
MCA Electrode 1 MCA Electrode 2 KMnO4 Electrode 1 KMnO4 Electrode 2 ClO2 Electrode 1 ClO2 Electrode 2 DO Electrode 1 DO Electrode 2 Cl2 Electrode 1Cl2 Electrode 2
ClO2 HOCldeg
KMnO4NH2Cl
DO
Different oxidants have different oxidizing power
Oxidation-Reduction Potential of several disinfectants in experimentJames et al 2004 (pH 7 10 mg CL 25degC)
17
ldquoClassicrdquo divalent Pb+2 solubility
pH6 7 8 9 10 11
mg
PbL
001
01
1
10
100 1 mg CL 5 mg CL 10 mg CL 20 mg CL 35 mg CL 50 mg CL 75 mg CL100 mg CL
DIC
Optimum pHDIC rangefor LSLs
High pH is needed to minimize Pb solubility
Lead ldquocorrosion Inhibitorsrdquo
bull Phosphoric acidbull Alkalai-metal orthophosphatebull Zinc orthophosphatebull Blended orthopolyphosphatesbull Polyphosphatesbull Sodium Silicate
Polyphosphates are not corrosion inhibitors but rather sequestering agentsEfficiency depends on sufficient doseconcentration for thebackground pH and other reactive water quality constituentswhat if the dose is too low to actually work 18
Illustrative reactions of P in pipebull OrthophosphatePrecipitation (consumption of orthophosphate)3Pb+2 + 2PO4
-3 --gt Pb3(PO4)2 (solid)decreased Pb solubility
bull PolyphosphatesReversion partially (formation of orthophosphate)P3O10
-5 ---gt P2O7-3 + PO4
-3
polymer --gt monomersComplexation (soluble complexes)3Pb+2 + P3O10
-5 --gt PbP3O10-2
higher Pb solubility
Slid
e cr
edit
Mar
c Ed
war
ds
Conceptual possible reactions Actual reactants and products may bedifferent Polyphosphates cannot be easily quantified
20
Ortho-P Treatment for Pb+2
Effectiveness Depends on Dose DIC pH and ldquoCleanlinessrdquo of Pipe Surface
mg PO4L00 10 20 30 40 50
mg
PbL
000
005
010
015
020
025
030
035
040
48 mg CL
48 mg CL
pH = 70pH = 75pH = 80pH = 85
bull pH less critical at low DICbull pH less critical at high PO4bull Point of diminishing returns higher with high DIC
bull Faster Pb reduction at high PO4
Typical UK Dosages 4-6 mgL
Most PWSs with LSLs do not have corrosion control treatment that minimizes Pb release and exposure
Effect of pH and ortho-P on Pb release
21
DIC = 10 mg CL 1 mg PO4L
60 65 70 75 80 85 90 95
gL
b m
P
0001
001
01
1
10
US Action Level
Dissolved Pb with ortho-PTotal Pb with ortho-PDissolved Pb with CO3Total Pb with CO3
At low DIC orthophosphate improves lead release regardless of pH
pHSchock M R DeSantis M K Metz D H Welch M M Hyland R N Nadagouda M N Revisiting the pH Effect on the Orthophosphate Control of Plumbosolvency Proc AWWA Annual Conference and Exposition Atlanta GA 2008
Ortho-P at pH 90 (DIC 6 mgL)
22
Pb(μgL)
bull Low Pb at 1 mgL relative to higher doses needs to be further verified across water qualities
bull Must do dose optimization study for your own water quality especially at high pH
bull Ortho-P may precipitate with Ca
Miller S A Investigation of Lead Solubility and Orthophosphate Addition in High pH Low DIC Water Master of Science Department of Biomedical Chemical and Environmental Engineering College of Engineering and Applied Science University of Cincinnati Cincinnati OH 2014
Ortho-Ppoint of diminishing returns
bull Orthophosphate addition to where large increments result in small reductions in lead release
bull Key to cost-effective lead release control and exposure reduction
bull Becomes the rdquomaintenancerdquo dosage unless the scale ldquoagesrdquo
bull Background constituents such as aluminum hardness ions iron manganese and others interfere with optimum reduction of lead release
bull Varies with the background water chemistry from system to system
23
Sodium silicate
bull No systematic studies to look at pH carbonate silicate background chemistry relationships
bull Unclear mechanism pH increase Buffer intensity increase Protective film (thin amp amorphous)
bull Can sequester ironmanganese
bull Canrsquot be evaluated with fresh surfaces
24pH5 6 7 8 9 10 11
β ty
nsi
nte
Iuf
fer
B
00000
00005
00010TIC=48 mg CL20 mgL SiO250 mgL PO4β H2OTotal β
Silicate may contribute to Buffer Intensity
Carbonate orthophosphate silicate I=001 25ordmC
Chemical changes may reduce Pb+4 to Pb+2
10
8
6
4
)ts 2ol
h (v
E
ndash2
ndash4
ndash6
ndash8
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
pH
12
PbO2 (plattnerite)
Pb++ )s
deg
(2
CO
3
(OH
) -2 2) 3
Pb
-2
0 (CO
-- 2
0
) 43(C
O
Pb
3 (OH
)
DIC = 18 mg CL
Pb Pb Pb = 0010 mgL
Pb metal
-10
Drop in ORP from treatment change or DS
oxidant demand
Drop in pH at surface from treatment change
chemical reactions nitrification etc
C D
ngto
nhisa
W
Newark NJ
Disinfectant demand in DS must be controlled amp enough free chlorine consistently maintained throughout LSL area to keep protective Pb+4 corrosion scales in place
Switching disinfectants may reduce Pb+4 to Pb+2
Low
Switch from chlorine to chloramine disinfectant dissolved lead from pipe scales during the Washington DC ldquolead-in-water crisisrdquo in 2001-2004
Example profile of Pb(+4) (ie PbO2) scale house
27
20
L)
Cl1 Cincinnati OH
18
microg 16
n ( 14
o 12
tia 10
trn 8
e 6
nco 4
C 2
b P 0-1 0 1 2 3 4 5 6 7 8 9 10 11
Cumulative Water Volume (L)
CI-1 10 h 1209CI-1 10 h 0410
Plumbing sequence corresponding to water volumeFaucet LSLCu pipe Brass
Water MainMCu pipe 2
015
a
let
dou
illyaf
ant
irT
bull Pb release to water from PbO2 coated LSLs was found to be lowbull Be aware are there are no known cases of exhumed PbO2 coated pipes being able to
re-stabilize in pipe rigs
Chloride and sulfateCSMR a factor controlling galvanic corrosion (Oliphant 1983 Gregory 1986)- Chloride detrimental- Sulfate beneficial if sufficient to overcome chloride thus- CSMR ratio was created
[Cl- ] 12 mgL Cl-Chloride to Sulfate Mass Ratio (CSMR) = =[SO -2 2 = 06
4 ] 20 mgL SO-4
Confirmed empirically in case studies of water utilities (Dodrill and Edwards 1995 Edwards and Triantafyllidou 2007) - High CSMR gt 05-06 more corrosive- Low CSMR lt 05-06 less corrosive
28
Verified in laboratory experiments with fresh galvanic Pb junctions (eg Wang et al 2013 Triantafyllidou et al 2011 Nguyen et al 2010 2011)
Tria
ntaf
yllid
ou e
t al
201
1
Chloride and sulfate
29
CSMR considered contributing factor to the Flint water Pb crisis (Pieper et al 2017)
30
Chloride and sulfate
bull Absolute concentrations of chloride and sulfate important chloride considered particularly detrimental to galvanic Pb corrosion
bull Increasing chloride and sulfate concentrations at a constant CSMR of 1 increased total lead concentrations in galvanic corrosion experiments (Ng amp Lin 2016)
bull High CSMR ratio proven empirically important in aggravating galvanic lead corrosion (verified by experiments and modeling)
bull Water source or treatment changes that affect the CSMR must be evaluated prior to change
bull Absolute concentrations also important particularly for chloride
31
Manganese
Two distinct pipe scale categories in 5 LSLs (Lead 1-5 Pipes) excavated from same city
31
Schock et al 2014 JAWWAbull The outer pipe scale is in direct contact with the flowing water affecting
water lead contamination the mostbull Plattnerite an insoluble Pb(+4) scale that protects water from lead
contamination was inhibited by Mn (and Fe) presence
Manganese
In this case study (Schock et al JAWWA 2014)bull Where Mn was present in more abundance
- Pb scales had less PbO2- Pb associated with Mn sorbing onto Mn+Fe-rich surface deposit
bull Erratic Pb persisted after full LSLR in areas rich with Mn deposits (and to lesser extent Fe)
bull Galvanized interior plumbing also accumulated dissolved Pb while LSL present
32
33
Manganese
Other studiesbull Dong et al 2000 Mn acted as vector for particulate Pb movementbull Trueman et al 2019 Increasing Mn increased Pb in experimentbull McNeill and Edwards 2004 Mn sequestration by polyphosphate destabilized
Pb scales and caused Pb release
bull Pan et al 2019 Possible benefit by promoting formation of Pb(+4)
Reference Analysis Mn Impact on PbSchock et al 2014 Excavated LSLs
amp historical water Pb data in a US city
bull Promoted amorphous Pb scalebull Inhibited insoluble Pb(+4) scalebull Erratic Pb persisted after full LSLR in areas rich
with Mn deposits
Tully et al 2019 Excavated LSLs amp water Pb data in several Midwest case studies
bull Mn was a component of amorphous pipe scales in several systems
bull Amorphous scales not amenable to determining their adherence to pipe walls their responses to changes in various water chemistry variables or quantitatively calculating their lead release
Itrsquos not just manganese
34
Broad variety of coatings on LSLs analyzed at EPA
Al Mn Fe P and Ca-rich coatings may interfere withorthophosphate pH adjustment or other CCT
35
More water parameters affect metal release
bull pH amp AlkalinityDICbull Concentration and type of
oxidants (disinfectantsdissolved oxygen)
bull ORPcorrosion potentialbull Corrosion Inhibitorsbull Chloridebull Sulfatebull Manganese
bull Temperaturebull Sorptive surfaces downstream of
LSLs (ie galvanized interior pipe)bull Iron (deposition and corrosion)bull Calciumbull Aluminumbull NOM (type amount)bull Mixing from water treatment plants
or sourcesbull Ammoniabull Hydrogen Sulfidebull Microbial activity (nitrification amp
other)
Corrosion Control Treatment is intertwined with all treatments affecting water chemistry
Water Quality Parameters(WQP)
bull pH alkalinity calcium orthophosphate or silica othersbull WQP measurements are important for
bull Process control (measurement frequency needs to be HIGH)bull Keeping treatment from being turned offbull Keeping inhibitor dosages amp pH control from being cut to save $$$bull Keeping required repairs to treatment on tight schedulebull Preventing frequent variations and alternations of water sources
bull Effectiveness of Pb treatment only determinable from tap sampling by combination of different assessment approaches for different lead plumbing sources to which people can be exposed
36
bull OWQP measurements important for process control look at yourdata to understand what they mean and avoid problems
bull OWQP do not predict lead release There are no substitutes fordirectly monitoring lead release
37
M+
M+
M+
M+
Corrosion barrier film hypothesis
Inert barrier film (eg CaCO3)
Pipe Wall
Barrier film would allow tobull Protect ALL pipe materialsbull Monitor corrosion based on one chemical reaction
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
Dissolved Inorganic Carbon (DIC) and Total ALKalinity (TALK)
DIC = [CO32-] + [H2CO3
] + [HCO3- ]]
TALK = 2 [CO32-] + [HCO3
-] + [OH-] - [H+]
mg CL Dissolved Inorganic Carbon (DIC)0 10 20 30 40 50
mg
CaCO
3L
Tota
l Alka
linity
0255075
100125150175
200225250
pH 60pH 70pH 80pH 90pH 100
bull To understand corrosion it isimportant to keep up with thecarbonate system
bull DIC and TALK have linearrelationship but are not thesame thing
bullbullbullbullbull
Oxidants in drinking water
bull Disinfection (Free chlorine Chloramine Chlorine dioxide)Pre-(O3 H2O2 ClO2 KMnO4) Oxidative metal removal (eg As Fe Mn) Ammonia removalAeration (corrosion control VOCRn2S removal) Taste and odor control
bull Dissolved oxygen
Oxidants in drinking water
Oxidant Dosage (mgL)0 2 4 6 8 10
E H (V
olts
vs S
HE)
03
04
05
06
07
08
09
10
11
MCA Electrode 1 MCA Electrode 2 KMnO4 Electrode 1 KMnO4 Electrode 2 ClO2 Electrode 1 ClO2 Electrode 2 DO Electrode 1 DO Electrode 2 Cl2 Electrode 1Cl2 Electrode 2
ClO2 HOCldeg
KMnO4NH2Cl
DO
Different oxidants have different oxidizing power
Oxidation-Reduction Potential of several disinfectants in experimentJames et al 2004 (pH 7 10 mg CL 25degC)
17
ldquoClassicrdquo divalent Pb+2 solubility
pH6 7 8 9 10 11
mg
PbL
001
01
1
10
100 1 mg CL 5 mg CL 10 mg CL 20 mg CL 35 mg CL 50 mg CL 75 mg CL100 mg CL
DIC
Optimum pHDIC rangefor LSLs
High pH is needed to minimize Pb solubility
Lead ldquocorrosion Inhibitorsrdquo
bull Phosphoric acidbull Alkalai-metal orthophosphatebull Zinc orthophosphatebull Blended orthopolyphosphatesbull Polyphosphatesbull Sodium Silicate
Polyphosphates are not corrosion inhibitors but rather sequestering agentsEfficiency depends on sufficient doseconcentration for thebackground pH and other reactive water quality constituentswhat if the dose is too low to actually work 18
Illustrative reactions of P in pipebull OrthophosphatePrecipitation (consumption of orthophosphate)3Pb+2 + 2PO4
-3 --gt Pb3(PO4)2 (solid)decreased Pb solubility
bull PolyphosphatesReversion partially (formation of orthophosphate)P3O10
-5 ---gt P2O7-3 + PO4
-3
polymer --gt monomersComplexation (soluble complexes)3Pb+2 + P3O10
-5 --gt PbP3O10-2
higher Pb solubility
Slid
e cr
edit
Mar
c Ed
war
ds
Conceptual possible reactions Actual reactants and products may bedifferent Polyphosphates cannot be easily quantified
20
Ortho-P Treatment for Pb+2
Effectiveness Depends on Dose DIC pH and ldquoCleanlinessrdquo of Pipe Surface
mg PO4L00 10 20 30 40 50
mg
PbL
000
005
010
015
020
025
030
035
040
48 mg CL
48 mg CL
pH = 70pH = 75pH = 80pH = 85
bull pH less critical at low DICbull pH less critical at high PO4bull Point of diminishing returns higher with high DIC
bull Faster Pb reduction at high PO4
Typical UK Dosages 4-6 mgL
Most PWSs with LSLs do not have corrosion control treatment that minimizes Pb release and exposure
Effect of pH and ortho-P on Pb release
21
DIC = 10 mg CL 1 mg PO4L
60 65 70 75 80 85 90 95
gL
b m
P
0001
001
01
1
10
US Action Level
Dissolved Pb with ortho-PTotal Pb with ortho-PDissolved Pb with CO3Total Pb with CO3
At low DIC orthophosphate improves lead release regardless of pH
pHSchock M R DeSantis M K Metz D H Welch M M Hyland R N Nadagouda M N Revisiting the pH Effect on the Orthophosphate Control of Plumbosolvency Proc AWWA Annual Conference and Exposition Atlanta GA 2008
Ortho-P at pH 90 (DIC 6 mgL)
22
Pb(μgL)
bull Low Pb at 1 mgL relative to higher doses needs to be further verified across water qualities
bull Must do dose optimization study for your own water quality especially at high pH
bull Ortho-P may precipitate with Ca
Miller S A Investigation of Lead Solubility and Orthophosphate Addition in High pH Low DIC Water Master of Science Department of Biomedical Chemical and Environmental Engineering College of Engineering and Applied Science University of Cincinnati Cincinnati OH 2014
Ortho-Ppoint of diminishing returns
bull Orthophosphate addition to where large increments result in small reductions in lead release
bull Key to cost-effective lead release control and exposure reduction
bull Becomes the rdquomaintenancerdquo dosage unless the scale ldquoagesrdquo
bull Background constituents such as aluminum hardness ions iron manganese and others interfere with optimum reduction of lead release
bull Varies with the background water chemistry from system to system
23
Sodium silicate
bull No systematic studies to look at pH carbonate silicate background chemistry relationships
bull Unclear mechanism pH increase Buffer intensity increase Protective film (thin amp amorphous)
bull Can sequester ironmanganese
bull Canrsquot be evaluated with fresh surfaces
24pH5 6 7 8 9 10 11
β ty
nsi
nte
Iuf
fer
B
00000
00005
00010TIC=48 mg CL20 mgL SiO250 mgL PO4β H2OTotal β
Silicate may contribute to Buffer Intensity
Carbonate orthophosphate silicate I=001 25ordmC
Chemical changes may reduce Pb+4 to Pb+2
10
8
6
4
)ts 2ol
h (v
E
ndash2
ndash4
ndash6
ndash8
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
pH
12
PbO2 (plattnerite)
Pb++ )s
deg
(2
CO
3
(OH
) -2 2) 3
Pb
-2
0 (CO
-- 2
0
) 43(C
O
Pb
3 (OH
)
DIC = 18 mg CL
Pb Pb Pb = 0010 mgL
Pb metal
-10
Drop in ORP from treatment change or DS
oxidant demand
Drop in pH at surface from treatment change
chemical reactions nitrification etc
C D
ngto
nhisa
W
Newark NJ
Disinfectant demand in DS must be controlled amp enough free chlorine consistently maintained throughout LSL area to keep protective Pb+4 corrosion scales in place
Switching disinfectants may reduce Pb+4 to Pb+2
Low
Switch from chlorine to chloramine disinfectant dissolved lead from pipe scales during the Washington DC ldquolead-in-water crisisrdquo in 2001-2004
Example profile of Pb(+4) (ie PbO2) scale house
27
20
L)
Cl1 Cincinnati OH
18
microg 16
n ( 14
o 12
tia 10
trn 8
e 6
nco 4
C 2
b P 0-1 0 1 2 3 4 5 6 7 8 9 10 11
Cumulative Water Volume (L)
CI-1 10 h 1209CI-1 10 h 0410
Plumbing sequence corresponding to water volumeFaucet LSLCu pipe Brass
Water MainMCu pipe 2
015
a
let
dou
illyaf
ant
irT
bull Pb release to water from PbO2 coated LSLs was found to be lowbull Be aware are there are no known cases of exhumed PbO2 coated pipes being able to
re-stabilize in pipe rigs
Chloride and sulfateCSMR a factor controlling galvanic corrosion (Oliphant 1983 Gregory 1986)- Chloride detrimental- Sulfate beneficial if sufficient to overcome chloride thus- CSMR ratio was created
[Cl- ] 12 mgL Cl-Chloride to Sulfate Mass Ratio (CSMR) = =[SO -2 2 = 06
4 ] 20 mgL SO-4
Confirmed empirically in case studies of water utilities (Dodrill and Edwards 1995 Edwards and Triantafyllidou 2007) - High CSMR gt 05-06 more corrosive- Low CSMR lt 05-06 less corrosive
28
Verified in laboratory experiments with fresh galvanic Pb junctions (eg Wang et al 2013 Triantafyllidou et al 2011 Nguyen et al 2010 2011)
Tria
ntaf
yllid
ou e
t al
201
1
Chloride and sulfate
29
CSMR considered contributing factor to the Flint water Pb crisis (Pieper et al 2017)
30
Chloride and sulfate
bull Absolute concentrations of chloride and sulfate important chloride considered particularly detrimental to galvanic Pb corrosion
bull Increasing chloride and sulfate concentrations at a constant CSMR of 1 increased total lead concentrations in galvanic corrosion experiments (Ng amp Lin 2016)
bull High CSMR ratio proven empirically important in aggravating galvanic lead corrosion (verified by experiments and modeling)
bull Water source or treatment changes that affect the CSMR must be evaluated prior to change
bull Absolute concentrations also important particularly for chloride
31
Manganese
Two distinct pipe scale categories in 5 LSLs (Lead 1-5 Pipes) excavated from same city
31
Schock et al 2014 JAWWAbull The outer pipe scale is in direct contact with the flowing water affecting
water lead contamination the mostbull Plattnerite an insoluble Pb(+4) scale that protects water from lead
contamination was inhibited by Mn (and Fe) presence
Manganese
In this case study (Schock et al JAWWA 2014)bull Where Mn was present in more abundance
- Pb scales had less PbO2- Pb associated with Mn sorbing onto Mn+Fe-rich surface deposit
bull Erratic Pb persisted after full LSLR in areas rich with Mn deposits (and to lesser extent Fe)
bull Galvanized interior plumbing also accumulated dissolved Pb while LSL present
32
33
Manganese
Other studiesbull Dong et al 2000 Mn acted as vector for particulate Pb movementbull Trueman et al 2019 Increasing Mn increased Pb in experimentbull McNeill and Edwards 2004 Mn sequestration by polyphosphate destabilized
Pb scales and caused Pb release
bull Pan et al 2019 Possible benefit by promoting formation of Pb(+4)
Reference Analysis Mn Impact on PbSchock et al 2014 Excavated LSLs
amp historical water Pb data in a US city
bull Promoted amorphous Pb scalebull Inhibited insoluble Pb(+4) scalebull Erratic Pb persisted after full LSLR in areas rich
with Mn deposits
Tully et al 2019 Excavated LSLs amp water Pb data in several Midwest case studies
bull Mn was a component of amorphous pipe scales in several systems
bull Amorphous scales not amenable to determining their adherence to pipe walls their responses to changes in various water chemistry variables or quantitatively calculating their lead release
Itrsquos not just manganese
34
Broad variety of coatings on LSLs analyzed at EPA
Al Mn Fe P and Ca-rich coatings may interfere withorthophosphate pH adjustment or other CCT
35
More water parameters affect metal release
bull pH amp AlkalinityDICbull Concentration and type of
oxidants (disinfectantsdissolved oxygen)
bull ORPcorrosion potentialbull Corrosion Inhibitorsbull Chloridebull Sulfatebull Manganese
bull Temperaturebull Sorptive surfaces downstream of
LSLs (ie galvanized interior pipe)bull Iron (deposition and corrosion)bull Calciumbull Aluminumbull NOM (type amount)bull Mixing from water treatment plants
or sourcesbull Ammoniabull Hydrogen Sulfidebull Microbial activity (nitrification amp
other)
Corrosion Control Treatment is intertwined with all treatments affecting water chemistry
Water Quality Parameters(WQP)
bull pH alkalinity calcium orthophosphate or silica othersbull WQP measurements are important for
bull Process control (measurement frequency needs to be HIGH)bull Keeping treatment from being turned offbull Keeping inhibitor dosages amp pH control from being cut to save $$$bull Keeping required repairs to treatment on tight schedulebull Preventing frequent variations and alternations of water sources
bull Effectiveness of Pb treatment only determinable from tap sampling by combination of different assessment approaches for different lead plumbing sources to which people can be exposed
36
bull OWQP measurements important for process control look at yourdata to understand what they mean and avoid problems
bull OWQP do not predict lead release There are no substitutes fordirectly monitoring lead release
37
M+
M+
M+
M+
Corrosion barrier film hypothesis
Inert barrier film (eg CaCO3)
Pipe Wall
Barrier film would allow tobull Protect ALL pipe materialsbull Monitor corrosion based on one chemical reaction
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
bullbullbullbullbull
Oxidants in drinking water
bull Disinfection (Free chlorine Chloramine Chlorine dioxide)Pre-(O3 H2O2 ClO2 KMnO4) Oxidative metal removal (eg As Fe Mn) Ammonia removalAeration (corrosion control VOCRn2S removal) Taste and odor control
bull Dissolved oxygen
Oxidants in drinking water
Oxidant Dosage (mgL)0 2 4 6 8 10
E H (V
olts
vs S
HE)
03
04
05
06
07
08
09
10
11
MCA Electrode 1 MCA Electrode 2 KMnO4 Electrode 1 KMnO4 Electrode 2 ClO2 Electrode 1 ClO2 Electrode 2 DO Electrode 1 DO Electrode 2 Cl2 Electrode 1Cl2 Electrode 2
ClO2 HOCldeg
KMnO4NH2Cl
DO
Different oxidants have different oxidizing power
Oxidation-Reduction Potential of several disinfectants in experimentJames et al 2004 (pH 7 10 mg CL 25degC)
17
ldquoClassicrdquo divalent Pb+2 solubility
pH6 7 8 9 10 11
mg
PbL
001
01
1
10
100 1 mg CL 5 mg CL 10 mg CL 20 mg CL 35 mg CL 50 mg CL 75 mg CL100 mg CL
DIC
Optimum pHDIC rangefor LSLs
High pH is needed to minimize Pb solubility
Lead ldquocorrosion Inhibitorsrdquo
bull Phosphoric acidbull Alkalai-metal orthophosphatebull Zinc orthophosphatebull Blended orthopolyphosphatesbull Polyphosphatesbull Sodium Silicate
Polyphosphates are not corrosion inhibitors but rather sequestering agentsEfficiency depends on sufficient doseconcentration for thebackground pH and other reactive water quality constituentswhat if the dose is too low to actually work 18
Illustrative reactions of P in pipebull OrthophosphatePrecipitation (consumption of orthophosphate)3Pb+2 + 2PO4
-3 --gt Pb3(PO4)2 (solid)decreased Pb solubility
bull PolyphosphatesReversion partially (formation of orthophosphate)P3O10
-5 ---gt P2O7-3 + PO4
-3
polymer --gt monomersComplexation (soluble complexes)3Pb+2 + P3O10
-5 --gt PbP3O10-2
higher Pb solubility
Slid
e cr
edit
Mar
c Ed
war
ds
Conceptual possible reactions Actual reactants and products may bedifferent Polyphosphates cannot be easily quantified
20
Ortho-P Treatment for Pb+2
Effectiveness Depends on Dose DIC pH and ldquoCleanlinessrdquo of Pipe Surface
mg PO4L00 10 20 30 40 50
mg
PbL
000
005
010
015
020
025
030
035
040
48 mg CL
48 mg CL
pH = 70pH = 75pH = 80pH = 85
bull pH less critical at low DICbull pH less critical at high PO4bull Point of diminishing returns higher with high DIC
bull Faster Pb reduction at high PO4
Typical UK Dosages 4-6 mgL
Most PWSs with LSLs do not have corrosion control treatment that minimizes Pb release and exposure
Effect of pH and ortho-P on Pb release
21
DIC = 10 mg CL 1 mg PO4L
60 65 70 75 80 85 90 95
gL
b m
P
0001
001
01
1
10
US Action Level
Dissolved Pb with ortho-PTotal Pb with ortho-PDissolved Pb with CO3Total Pb with CO3
At low DIC orthophosphate improves lead release regardless of pH
pHSchock M R DeSantis M K Metz D H Welch M M Hyland R N Nadagouda M N Revisiting the pH Effect on the Orthophosphate Control of Plumbosolvency Proc AWWA Annual Conference and Exposition Atlanta GA 2008
Ortho-P at pH 90 (DIC 6 mgL)
22
Pb(μgL)
bull Low Pb at 1 mgL relative to higher doses needs to be further verified across water qualities
bull Must do dose optimization study for your own water quality especially at high pH
bull Ortho-P may precipitate with Ca
Miller S A Investigation of Lead Solubility and Orthophosphate Addition in High pH Low DIC Water Master of Science Department of Biomedical Chemical and Environmental Engineering College of Engineering and Applied Science University of Cincinnati Cincinnati OH 2014
Ortho-Ppoint of diminishing returns
bull Orthophosphate addition to where large increments result in small reductions in lead release
bull Key to cost-effective lead release control and exposure reduction
bull Becomes the rdquomaintenancerdquo dosage unless the scale ldquoagesrdquo
bull Background constituents such as aluminum hardness ions iron manganese and others interfere with optimum reduction of lead release
bull Varies with the background water chemistry from system to system
23
Sodium silicate
bull No systematic studies to look at pH carbonate silicate background chemistry relationships
bull Unclear mechanism pH increase Buffer intensity increase Protective film (thin amp amorphous)
bull Can sequester ironmanganese
bull Canrsquot be evaluated with fresh surfaces
24pH5 6 7 8 9 10 11
β ty
nsi
nte
Iuf
fer
B
00000
00005
00010TIC=48 mg CL20 mgL SiO250 mgL PO4β H2OTotal β
Silicate may contribute to Buffer Intensity
Carbonate orthophosphate silicate I=001 25ordmC
Chemical changes may reduce Pb+4 to Pb+2
10
8
6
4
)ts 2ol
h (v
E
ndash2
ndash4
ndash6
ndash8
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
pH
12
PbO2 (plattnerite)
Pb++ )s
deg
(2
CO
3
(OH
) -2 2) 3
Pb
-2
0 (CO
-- 2
0
) 43(C
O
Pb
3 (OH
)
DIC = 18 mg CL
Pb Pb Pb = 0010 mgL
Pb metal
-10
Drop in ORP from treatment change or DS
oxidant demand
Drop in pH at surface from treatment change
chemical reactions nitrification etc
C D
ngto
nhisa
W
Newark NJ
Disinfectant demand in DS must be controlled amp enough free chlorine consistently maintained throughout LSL area to keep protective Pb+4 corrosion scales in place
Switching disinfectants may reduce Pb+4 to Pb+2
Low
Switch from chlorine to chloramine disinfectant dissolved lead from pipe scales during the Washington DC ldquolead-in-water crisisrdquo in 2001-2004
Example profile of Pb(+4) (ie PbO2) scale house
27
20
L)
Cl1 Cincinnati OH
18
microg 16
n ( 14
o 12
tia 10
trn 8
e 6
nco 4
C 2
b P 0-1 0 1 2 3 4 5 6 7 8 9 10 11
Cumulative Water Volume (L)
CI-1 10 h 1209CI-1 10 h 0410
Plumbing sequence corresponding to water volumeFaucet LSLCu pipe Brass
Water MainMCu pipe 2
015
a
let
dou
illyaf
ant
irT
bull Pb release to water from PbO2 coated LSLs was found to be lowbull Be aware are there are no known cases of exhumed PbO2 coated pipes being able to
re-stabilize in pipe rigs
Chloride and sulfateCSMR a factor controlling galvanic corrosion (Oliphant 1983 Gregory 1986)- Chloride detrimental- Sulfate beneficial if sufficient to overcome chloride thus- CSMR ratio was created
[Cl- ] 12 mgL Cl-Chloride to Sulfate Mass Ratio (CSMR) = =[SO -2 2 = 06
4 ] 20 mgL SO-4
Confirmed empirically in case studies of water utilities (Dodrill and Edwards 1995 Edwards and Triantafyllidou 2007) - High CSMR gt 05-06 more corrosive- Low CSMR lt 05-06 less corrosive
28
Verified in laboratory experiments with fresh galvanic Pb junctions (eg Wang et al 2013 Triantafyllidou et al 2011 Nguyen et al 2010 2011)
Tria
ntaf
yllid
ou e
t al
201
1
Chloride and sulfate
29
CSMR considered contributing factor to the Flint water Pb crisis (Pieper et al 2017)
30
Chloride and sulfate
bull Absolute concentrations of chloride and sulfate important chloride considered particularly detrimental to galvanic Pb corrosion
bull Increasing chloride and sulfate concentrations at a constant CSMR of 1 increased total lead concentrations in galvanic corrosion experiments (Ng amp Lin 2016)
bull High CSMR ratio proven empirically important in aggravating galvanic lead corrosion (verified by experiments and modeling)
bull Water source or treatment changes that affect the CSMR must be evaluated prior to change
bull Absolute concentrations also important particularly for chloride
31
Manganese
Two distinct pipe scale categories in 5 LSLs (Lead 1-5 Pipes) excavated from same city
31
Schock et al 2014 JAWWAbull The outer pipe scale is in direct contact with the flowing water affecting
water lead contamination the mostbull Plattnerite an insoluble Pb(+4) scale that protects water from lead
contamination was inhibited by Mn (and Fe) presence
Manganese
In this case study (Schock et al JAWWA 2014)bull Where Mn was present in more abundance
- Pb scales had less PbO2- Pb associated with Mn sorbing onto Mn+Fe-rich surface deposit
bull Erratic Pb persisted after full LSLR in areas rich with Mn deposits (and to lesser extent Fe)
bull Galvanized interior plumbing also accumulated dissolved Pb while LSL present
32
33
Manganese
Other studiesbull Dong et al 2000 Mn acted as vector for particulate Pb movementbull Trueman et al 2019 Increasing Mn increased Pb in experimentbull McNeill and Edwards 2004 Mn sequestration by polyphosphate destabilized
Pb scales and caused Pb release
bull Pan et al 2019 Possible benefit by promoting formation of Pb(+4)
Reference Analysis Mn Impact on PbSchock et al 2014 Excavated LSLs
amp historical water Pb data in a US city
bull Promoted amorphous Pb scalebull Inhibited insoluble Pb(+4) scalebull Erratic Pb persisted after full LSLR in areas rich
with Mn deposits
Tully et al 2019 Excavated LSLs amp water Pb data in several Midwest case studies
bull Mn was a component of amorphous pipe scales in several systems
bull Amorphous scales not amenable to determining their adherence to pipe walls their responses to changes in various water chemistry variables or quantitatively calculating their lead release
Itrsquos not just manganese
34
Broad variety of coatings on LSLs analyzed at EPA
Al Mn Fe P and Ca-rich coatings may interfere withorthophosphate pH adjustment or other CCT
35
More water parameters affect metal release
bull pH amp AlkalinityDICbull Concentration and type of
oxidants (disinfectantsdissolved oxygen)
bull ORPcorrosion potentialbull Corrosion Inhibitorsbull Chloridebull Sulfatebull Manganese
bull Temperaturebull Sorptive surfaces downstream of
LSLs (ie galvanized interior pipe)bull Iron (deposition and corrosion)bull Calciumbull Aluminumbull NOM (type amount)bull Mixing from water treatment plants
or sourcesbull Ammoniabull Hydrogen Sulfidebull Microbial activity (nitrification amp
other)
Corrosion Control Treatment is intertwined with all treatments affecting water chemistry
Water Quality Parameters(WQP)
bull pH alkalinity calcium orthophosphate or silica othersbull WQP measurements are important for
bull Process control (measurement frequency needs to be HIGH)bull Keeping treatment from being turned offbull Keeping inhibitor dosages amp pH control from being cut to save $$$bull Keeping required repairs to treatment on tight schedulebull Preventing frequent variations and alternations of water sources
bull Effectiveness of Pb treatment only determinable from tap sampling by combination of different assessment approaches for different lead plumbing sources to which people can be exposed
36
bull OWQP measurements important for process control look at yourdata to understand what they mean and avoid problems
bull OWQP do not predict lead release There are no substitutes fordirectly monitoring lead release
37
M+
M+
M+
M+
Corrosion barrier film hypothesis
Inert barrier film (eg CaCO3)
Pipe Wall
Barrier film would allow tobull Protect ALL pipe materialsbull Monitor corrosion based on one chemical reaction
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
Oxidants in drinking water
Oxidant Dosage (mgL)0 2 4 6 8 10
E H (V
olts
vs S
HE)
03
04
05
06
07
08
09
10
11
MCA Electrode 1 MCA Electrode 2 KMnO4 Electrode 1 KMnO4 Electrode 2 ClO2 Electrode 1 ClO2 Electrode 2 DO Electrode 1 DO Electrode 2 Cl2 Electrode 1Cl2 Electrode 2
ClO2 HOCldeg
KMnO4NH2Cl
DO
Different oxidants have different oxidizing power
Oxidation-Reduction Potential of several disinfectants in experimentJames et al 2004 (pH 7 10 mg CL 25degC)
17
ldquoClassicrdquo divalent Pb+2 solubility
pH6 7 8 9 10 11
mg
PbL
001
01
1
10
100 1 mg CL 5 mg CL 10 mg CL 20 mg CL 35 mg CL 50 mg CL 75 mg CL100 mg CL
DIC
Optimum pHDIC rangefor LSLs
High pH is needed to minimize Pb solubility
Lead ldquocorrosion Inhibitorsrdquo
bull Phosphoric acidbull Alkalai-metal orthophosphatebull Zinc orthophosphatebull Blended orthopolyphosphatesbull Polyphosphatesbull Sodium Silicate
Polyphosphates are not corrosion inhibitors but rather sequestering agentsEfficiency depends on sufficient doseconcentration for thebackground pH and other reactive water quality constituentswhat if the dose is too low to actually work 18
Illustrative reactions of P in pipebull OrthophosphatePrecipitation (consumption of orthophosphate)3Pb+2 + 2PO4
-3 --gt Pb3(PO4)2 (solid)decreased Pb solubility
bull PolyphosphatesReversion partially (formation of orthophosphate)P3O10
-5 ---gt P2O7-3 + PO4
-3
polymer --gt monomersComplexation (soluble complexes)3Pb+2 + P3O10
-5 --gt PbP3O10-2
higher Pb solubility
Slid
e cr
edit
Mar
c Ed
war
ds
Conceptual possible reactions Actual reactants and products may bedifferent Polyphosphates cannot be easily quantified
20
Ortho-P Treatment for Pb+2
Effectiveness Depends on Dose DIC pH and ldquoCleanlinessrdquo of Pipe Surface
mg PO4L00 10 20 30 40 50
mg
PbL
000
005
010
015
020
025
030
035
040
48 mg CL
48 mg CL
pH = 70pH = 75pH = 80pH = 85
bull pH less critical at low DICbull pH less critical at high PO4bull Point of diminishing returns higher with high DIC
bull Faster Pb reduction at high PO4
Typical UK Dosages 4-6 mgL
Most PWSs with LSLs do not have corrosion control treatment that minimizes Pb release and exposure
Effect of pH and ortho-P on Pb release
21
DIC = 10 mg CL 1 mg PO4L
60 65 70 75 80 85 90 95
gL
b m
P
0001
001
01
1
10
US Action Level
Dissolved Pb with ortho-PTotal Pb with ortho-PDissolved Pb with CO3Total Pb with CO3
At low DIC orthophosphate improves lead release regardless of pH
pHSchock M R DeSantis M K Metz D H Welch M M Hyland R N Nadagouda M N Revisiting the pH Effect on the Orthophosphate Control of Plumbosolvency Proc AWWA Annual Conference and Exposition Atlanta GA 2008
Ortho-P at pH 90 (DIC 6 mgL)
22
Pb(μgL)
bull Low Pb at 1 mgL relative to higher doses needs to be further verified across water qualities
bull Must do dose optimization study for your own water quality especially at high pH
bull Ortho-P may precipitate with Ca
Miller S A Investigation of Lead Solubility and Orthophosphate Addition in High pH Low DIC Water Master of Science Department of Biomedical Chemical and Environmental Engineering College of Engineering and Applied Science University of Cincinnati Cincinnati OH 2014
Ortho-Ppoint of diminishing returns
bull Orthophosphate addition to where large increments result in small reductions in lead release
bull Key to cost-effective lead release control and exposure reduction
bull Becomes the rdquomaintenancerdquo dosage unless the scale ldquoagesrdquo
bull Background constituents such as aluminum hardness ions iron manganese and others interfere with optimum reduction of lead release
bull Varies with the background water chemistry from system to system
23
Sodium silicate
bull No systematic studies to look at pH carbonate silicate background chemistry relationships
bull Unclear mechanism pH increase Buffer intensity increase Protective film (thin amp amorphous)
bull Can sequester ironmanganese
bull Canrsquot be evaluated with fresh surfaces
24pH5 6 7 8 9 10 11
β ty
nsi
nte
Iuf
fer
B
00000
00005
00010TIC=48 mg CL20 mgL SiO250 mgL PO4β H2OTotal β
Silicate may contribute to Buffer Intensity
Carbonate orthophosphate silicate I=001 25ordmC
Chemical changes may reduce Pb+4 to Pb+2
10
8
6
4
)ts 2ol
h (v
E
ndash2
ndash4
ndash6
ndash8
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
pH
12
PbO2 (plattnerite)
Pb++ )s
deg
(2
CO
3
(OH
) -2 2) 3
Pb
-2
0 (CO
-- 2
0
) 43(C
O
Pb
3 (OH
)
DIC = 18 mg CL
Pb Pb Pb = 0010 mgL
Pb metal
-10
Drop in ORP from treatment change or DS
oxidant demand
Drop in pH at surface from treatment change
chemical reactions nitrification etc
C D
ngto
nhisa
W
Newark NJ
Disinfectant demand in DS must be controlled amp enough free chlorine consistently maintained throughout LSL area to keep protective Pb+4 corrosion scales in place
Switching disinfectants may reduce Pb+4 to Pb+2
Low
Switch from chlorine to chloramine disinfectant dissolved lead from pipe scales during the Washington DC ldquolead-in-water crisisrdquo in 2001-2004
Example profile of Pb(+4) (ie PbO2) scale house
27
20
L)
Cl1 Cincinnati OH
18
microg 16
n ( 14
o 12
tia 10
trn 8
e 6
nco 4
C 2
b P 0-1 0 1 2 3 4 5 6 7 8 9 10 11
Cumulative Water Volume (L)
CI-1 10 h 1209CI-1 10 h 0410
Plumbing sequence corresponding to water volumeFaucet LSLCu pipe Brass
Water MainMCu pipe 2
015
a
let
dou
illyaf
ant
irT
bull Pb release to water from PbO2 coated LSLs was found to be lowbull Be aware are there are no known cases of exhumed PbO2 coated pipes being able to
re-stabilize in pipe rigs
Chloride and sulfateCSMR a factor controlling galvanic corrosion (Oliphant 1983 Gregory 1986)- Chloride detrimental- Sulfate beneficial if sufficient to overcome chloride thus- CSMR ratio was created
[Cl- ] 12 mgL Cl-Chloride to Sulfate Mass Ratio (CSMR) = =[SO -2 2 = 06
4 ] 20 mgL SO-4
Confirmed empirically in case studies of water utilities (Dodrill and Edwards 1995 Edwards and Triantafyllidou 2007) - High CSMR gt 05-06 more corrosive- Low CSMR lt 05-06 less corrosive
28
Verified in laboratory experiments with fresh galvanic Pb junctions (eg Wang et al 2013 Triantafyllidou et al 2011 Nguyen et al 2010 2011)
Tria
ntaf
yllid
ou e
t al
201
1
Chloride and sulfate
29
CSMR considered contributing factor to the Flint water Pb crisis (Pieper et al 2017)
30
Chloride and sulfate
bull Absolute concentrations of chloride and sulfate important chloride considered particularly detrimental to galvanic Pb corrosion
bull Increasing chloride and sulfate concentrations at a constant CSMR of 1 increased total lead concentrations in galvanic corrosion experiments (Ng amp Lin 2016)
bull High CSMR ratio proven empirically important in aggravating galvanic lead corrosion (verified by experiments and modeling)
bull Water source or treatment changes that affect the CSMR must be evaluated prior to change
bull Absolute concentrations also important particularly for chloride
31
Manganese
Two distinct pipe scale categories in 5 LSLs (Lead 1-5 Pipes) excavated from same city
31
Schock et al 2014 JAWWAbull The outer pipe scale is in direct contact with the flowing water affecting
water lead contamination the mostbull Plattnerite an insoluble Pb(+4) scale that protects water from lead
contamination was inhibited by Mn (and Fe) presence
Manganese
In this case study (Schock et al JAWWA 2014)bull Where Mn was present in more abundance
- Pb scales had less PbO2- Pb associated with Mn sorbing onto Mn+Fe-rich surface deposit
bull Erratic Pb persisted after full LSLR in areas rich with Mn deposits (and to lesser extent Fe)
bull Galvanized interior plumbing also accumulated dissolved Pb while LSL present
32
33
Manganese
Other studiesbull Dong et al 2000 Mn acted as vector for particulate Pb movementbull Trueman et al 2019 Increasing Mn increased Pb in experimentbull McNeill and Edwards 2004 Mn sequestration by polyphosphate destabilized
Pb scales and caused Pb release
bull Pan et al 2019 Possible benefit by promoting formation of Pb(+4)
Reference Analysis Mn Impact on PbSchock et al 2014 Excavated LSLs
amp historical water Pb data in a US city
bull Promoted amorphous Pb scalebull Inhibited insoluble Pb(+4) scalebull Erratic Pb persisted after full LSLR in areas rich
with Mn deposits
Tully et al 2019 Excavated LSLs amp water Pb data in several Midwest case studies
bull Mn was a component of amorphous pipe scales in several systems
bull Amorphous scales not amenable to determining their adherence to pipe walls their responses to changes in various water chemistry variables or quantitatively calculating their lead release
Itrsquos not just manganese
34
Broad variety of coatings on LSLs analyzed at EPA
Al Mn Fe P and Ca-rich coatings may interfere withorthophosphate pH adjustment or other CCT
35
More water parameters affect metal release
bull pH amp AlkalinityDICbull Concentration and type of
oxidants (disinfectantsdissolved oxygen)
bull ORPcorrosion potentialbull Corrosion Inhibitorsbull Chloridebull Sulfatebull Manganese
bull Temperaturebull Sorptive surfaces downstream of
LSLs (ie galvanized interior pipe)bull Iron (deposition and corrosion)bull Calciumbull Aluminumbull NOM (type amount)bull Mixing from water treatment plants
or sourcesbull Ammoniabull Hydrogen Sulfidebull Microbial activity (nitrification amp
other)
Corrosion Control Treatment is intertwined with all treatments affecting water chemistry
Water Quality Parameters(WQP)
bull pH alkalinity calcium orthophosphate or silica othersbull WQP measurements are important for
bull Process control (measurement frequency needs to be HIGH)bull Keeping treatment from being turned offbull Keeping inhibitor dosages amp pH control from being cut to save $$$bull Keeping required repairs to treatment on tight schedulebull Preventing frequent variations and alternations of water sources
bull Effectiveness of Pb treatment only determinable from tap sampling by combination of different assessment approaches for different lead plumbing sources to which people can be exposed
36
bull OWQP measurements important for process control look at yourdata to understand what they mean and avoid problems
bull OWQP do not predict lead release There are no substitutes fordirectly monitoring lead release
37
M+
M+
M+
M+
Corrosion barrier film hypothesis
Inert barrier film (eg CaCO3)
Pipe Wall
Barrier film would allow tobull Protect ALL pipe materialsbull Monitor corrosion based on one chemical reaction
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
17
ldquoClassicrdquo divalent Pb+2 solubility
pH6 7 8 9 10 11
mg
PbL
001
01
1
10
100 1 mg CL 5 mg CL 10 mg CL 20 mg CL 35 mg CL 50 mg CL 75 mg CL100 mg CL
DIC
Optimum pHDIC rangefor LSLs
High pH is needed to minimize Pb solubility
Lead ldquocorrosion Inhibitorsrdquo
bull Phosphoric acidbull Alkalai-metal orthophosphatebull Zinc orthophosphatebull Blended orthopolyphosphatesbull Polyphosphatesbull Sodium Silicate
Polyphosphates are not corrosion inhibitors but rather sequestering agentsEfficiency depends on sufficient doseconcentration for thebackground pH and other reactive water quality constituentswhat if the dose is too low to actually work 18
Illustrative reactions of P in pipebull OrthophosphatePrecipitation (consumption of orthophosphate)3Pb+2 + 2PO4
-3 --gt Pb3(PO4)2 (solid)decreased Pb solubility
bull PolyphosphatesReversion partially (formation of orthophosphate)P3O10
-5 ---gt P2O7-3 + PO4
-3
polymer --gt monomersComplexation (soluble complexes)3Pb+2 + P3O10
-5 --gt PbP3O10-2
higher Pb solubility
Slid
e cr
edit
Mar
c Ed
war
ds
Conceptual possible reactions Actual reactants and products may bedifferent Polyphosphates cannot be easily quantified
20
Ortho-P Treatment for Pb+2
Effectiveness Depends on Dose DIC pH and ldquoCleanlinessrdquo of Pipe Surface
mg PO4L00 10 20 30 40 50
mg
PbL
000
005
010
015
020
025
030
035
040
48 mg CL
48 mg CL
pH = 70pH = 75pH = 80pH = 85
bull pH less critical at low DICbull pH less critical at high PO4bull Point of diminishing returns higher with high DIC
bull Faster Pb reduction at high PO4
Typical UK Dosages 4-6 mgL
Most PWSs with LSLs do not have corrosion control treatment that minimizes Pb release and exposure
Effect of pH and ortho-P on Pb release
21
DIC = 10 mg CL 1 mg PO4L
60 65 70 75 80 85 90 95
gL
b m
P
0001
001
01
1
10
US Action Level
Dissolved Pb with ortho-PTotal Pb with ortho-PDissolved Pb with CO3Total Pb with CO3
At low DIC orthophosphate improves lead release regardless of pH
pHSchock M R DeSantis M K Metz D H Welch M M Hyland R N Nadagouda M N Revisiting the pH Effect on the Orthophosphate Control of Plumbosolvency Proc AWWA Annual Conference and Exposition Atlanta GA 2008
Ortho-P at pH 90 (DIC 6 mgL)
22
Pb(μgL)
bull Low Pb at 1 mgL relative to higher doses needs to be further verified across water qualities
bull Must do dose optimization study for your own water quality especially at high pH
bull Ortho-P may precipitate with Ca
Miller S A Investigation of Lead Solubility and Orthophosphate Addition in High pH Low DIC Water Master of Science Department of Biomedical Chemical and Environmental Engineering College of Engineering and Applied Science University of Cincinnati Cincinnati OH 2014
Ortho-Ppoint of diminishing returns
bull Orthophosphate addition to where large increments result in small reductions in lead release
bull Key to cost-effective lead release control and exposure reduction
bull Becomes the rdquomaintenancerdquo dosage unless the scale ldquoagesrdquo
bull Background constituents such as aluminum hardness ions iron manganese and others interfere with optimum reduction of lead release
bull Varies with the background water chemistry from system to system
23
Sodium silicate
bull No systematic studies to look at pH carbonate silicate background chemistry relationships
bull Unclear mechanism pH increase Buffer intensity increase Protective film (thin amp amorphous)
bull Can sequester ironmanganese
bull Canrsquot be evaluated with fresh surfaces
24pH5 6 7 8 9 10 11
β ty
nsi
nte
Iuf
fer
B
00000
00005
00010TIC=48 mg CL20 mgL SiO250 mgL PO4β H2OTotal β
Silicate may contribute to Buffer Intensity
Carbonate orthophosphate silicate I=001 25ordmC
Chemical changes may reduce Pb+4 to Pb+2
10
8
6
4
)ts 2ol
h (v
E
ndash2
ndash4
ndash6
ndash8
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
pH
12
PbO2 (plattnerite)
Pb++ )s
deg
(2
CO
3
(OH
) -2 2) 3
Pb
-2
0 (CO
-- 2
0
) 43(C
O
Pb
3 (OH
)
DIC = 18 mg CL
Pb Pb Pb = 0010 mgL
Pb metal
-10
Drop in ORP from treatment change or DS
oxidant demand
Drop in pH at surface from treatment change
chemical reactions nitrification etc
C D
ngto
nhisa
W
Newark NJ
Disinfectant demand in DS must be controlled amp enough free chlorine consistently maintained throughout LSL area to keep protective Pb+4 corrosion scales in place
Switching disinfectants may reduce Pb+4 to Pb+2
Low
Switch from chlorine to chloramine disinfectant dissolved lead from pipe scales during the Washington DC ldquolead-in-water crisisrdquo in 2001-2004
Example profile of Pb(+4) (ie PbO2) scale house
27
20
L)
Cl1 Cincinnati OH
18
microg 16
n ( 14
o 12
tia 10
trn 8
e 6
nco 4
C 2
b P 0-1 0 1 2 3 4 5 6 7 8 9 10 11
Cumulative Water Volume (L)
CI-1 10 h 1209CI-1 10 h 0410
Plumbing sequence corresponding to water volumeFaucet LSLCu pipe Brass
Water MainMCu pipe 2
015
a
let
dou
illyaf
ant
irT
bull Pb release to water from PbO2 coated LSLs was found to be lowbull Be aware are there are no known cases of exhumed PbO2 coated pipes being able to
re-stabilize in pipe rigs
Chloride and sulfateCSMR a factor controlling galvanic corrosion (Oliphant 1983 Gregory 1986)- Chloride detrimental- Sulfate beneficial if sufficient to overcome chloride thus- CSMR ratio was created
[Cl- ] 12 mgL Cl-Chloride to Sulfate Mass Ratio (CSMR) = =[SO -2 2 = 06
4 ] 20 mgL SO-4
Confirmed empirically in case studies of water utilities (Dodrill and Edwards 1995 Edwards and Triantafyllidou 2007) - High CSMR gt 05-06 more corrosive- Low CSMR lt 05-06 less corrosive
28
Verified in laboratory experiments with fresh galvanic Pb junctions (eg Wang et al 2013 Triantafyllidou et al 2011 Nguyen et al 2010 2011)
Tria
ntaf
yllid
ou e
t al
201
1
Chloride and sulfate
29
CSMR considered contributing factor to the Flint water Pb crisis (Pieper et al 2017)
30
Chloride and sulfate
bull Absolute concentrations of chloride and sulfate important chloride considered particularly detrimental to galvanic Pb corrosion
bull Increasing chloride and sulfate concentrations at a constant CSMR of 1 increased total lead concentrations in galvanic corrosion experiments (Ng amp Lin 2016)
bull High CSMR ratio proven empirically important in aggravating galvanic lead corrosion (verified by experiments and modeling)
bull Water source or treatment changes that affect the CSMR must be evaluated prior to change
bull Absolute concentrations also important particularly for chloride
31
Manganese
Two distinct pipe scale categories in 5 LSLs (Lead 1-5 Pipes) excavated from same city
31
Schock et al 2014 JAWWAbull The outer pipe scale is in direct contact with the flowing water affecting
water lead contamination the mostbull Plattnerite an insoluble Pb(+4) scale that protects water from lead
contamination was inhibited by Mn (and Fe) presence
Manganese
In this case study (Schock et al JAWWA 2014)bull Where Mn was present in more abundance
- Pb scales had less PbO2- Pb associated with Mn sorbing onto Mn+Fe-rich surface deposit
bull Erratic Pb persisted after full LSLR in areas rich with Mn deposits (and to lesser extent Fe)
bull Galvanized interior plumbing also accumulated dissolved Pb while LSL present
32
33
Manganese
Other studiesbull Dong et al 2000 Mn acted as vector for particulate Pb movementbull Trueman et al 2019 Increasing Mn increased Pb in experimentbull McNeill and Edwards 2004 Mn sequestration by polyphosphate destabilized
Pb scales and caused Pb release
bull Pan et al 2019 Possible benefit by promoting formation of Pb(+4)
Reference Analysis Mn Impact on PbSchock et al 2014 Excavated LSLs
amp historical water Pb data in a US city
bull Promoted amorphous Pb scalebull Inhibited insoluble Pb(+4) scalebull Erratic Pb persisted after full LSLR in areas rich
with Mn deposits
Tully et al 2019 Excavated LSLs amp water Pb data in several Midwest case studies
bull Mn was a component of amorphous pipe scales in several systems
bull Amorphous scales not amenable to determining their adherence to pipe walls their responses to changes in various water chemistry variables or quantitatively calculating their lead release
Itrsquos not just manganese
34
Broad variety of coatings on LSLs analyzed at EPA
Al Mn Fe P and Ca-rich coatings may interfere withorthophosphate pH adjustment or other CCT
35
More water parameters affect metal release
bull pH amp AlkalinityDICbull Concentration and type of
oxidants (disinfectantsdissolved oxygen)
bull ORPcorrosion potentialbull Corrosion Inhibitorsbull Chloridebull Sulfatebull Manganese
bull Temperaturebull Sorptive surfaces downstream of
LSLs (ie galvanized interior pipe)bull Iron (deposition and corrosion)bull Calciumbull Aluminumbull NOM (type amount)bull Mixing from water treatment plants
or sourcesbull Ammoniabull Hydrogen Sulfidebull Microbial activity (nitrification amp
other)
Corrosion Control Treatment is intertwined with all treatments affecting water chemistry
Water Quality Parameters(WQP)
bull pH alkalinity calcium orthophosphate or silica othersbull WQP measurements are important for
bull Process control (measurement frequency needs to be HIGH)bull Keeping treatment from being turned offbull Keeping inhibitor dosages amp pH control from being cut to save $$$bull Keeping required repairs to treatment on tight schedulebull Preventing frequent variations and alternations of water sources
bull Effectiveness of Pb treatment only determinable from tap sampling by combination of different assessment approaches for different lead plumbing sources to which people can be exposed
36
bull OWQP measurements important for process control look at yourdata to understand what they mean and avoid problems
bull OWQP do not predict lead release There are no substitutes fordirectly monitoring lead release
37
M+
M+
M+
M+
Corrosion barrier film hypothesis
Inert barrier film (eg CaCO3)
Pipe Wall
Barrier film would allow tobull Protect ALL pipe materialsbull Monitor corrosion based on one chemical reaction
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
Lead ldquocorrosion Inhibitorsrdquo
bull Phosphoric acidbull Alkalai-metal orthophosphatebull Zinc orthophosphatebull Blended orthopolyphosphatesbull Polyphosphatesbull Sodium Silicate
Polyphosphates are not corrosion inhibitors but rather sequestering agentsEfficiency depends on sufficient doseconcentration for thebackground pH and other reactive water quality constituentswhat if the dose is too low to actually work 18
Illustrative reactions of P in pipebull OrthophosphatePrecipitation (consumption of orthophosphate)3Pb+2 + 2PO4
-3 --gt Pb3(PO4)2 (solid)decreased Pb solubility
bull PolyphosphatesReversion partially (formation of orthophosphate)P3O10
-5 ---gt P2O7-3 + PO4
-3
polymer --gt monomersComplexation (soluble complexes)3Pb+2 + P3O10
-5 --gt PbP3O10-2
higher Pb solubility
Slid
e cr
edit
Mar
c Ed
war
ds
Conceptual possible reactions Actual reactants and products may bedifferent Polyphosphates cannot be easily quantified
20
Ortho-P Treatment for Pb+2
Effectiveness Depends on Dose DIC pH and ldquoCleanlinessrdquo of Pipe Surface
mg PO4L00 10 20 30 40 50
mg
PbL
000
005
010
015
020
025
030
035
040
48 mg CL
48 mg CL
pH = 70pH = 75pH = 80pH = 85
bull pH less critical at low DICbull pH less critical at high PO4bull Point of diminishing returns higher with high DIC
bull Faster Pb reduction at high PO4
Typical UK Dosages 4-6 mgL
Most PWSs with LSLs do not have corrosion control treatment that minimizes Pb release and exposure
Effect of pH and ortho-P on Pb release
21
DIC = 10 mg CL 1 mg PO4L
60 65 70 75 80 85 90 95
gL
b m
P
0001
001
01
1
10
US Action Level
Dissolved Pb with ortho-PTotal Pb with ortho-PDissolved Pb with CO3Total Pb with CO3
At low DIC orthophosphate improves lead release regardless of pH
pHSchock M R DeSantis M K Metz D H Welch M M Hyland R N Nadagouda M N Revisiting the pH Effect on the Orthophosphate Control of Plumbosolvency Proc AWWA Annual Conference and Exposition Atlanta GA 2008
Ortho-P at pH 90 (DIC 6 mgL)
22
Pb(μgL)
bull Low Pb at 1 mgL relative to higher doses needs to be further verified across water qualities
bull Must do dose optimization study for your own water quality especially at high pH
bull Ortho-P may precipitate with Ca
Miller S A Investigation of Lead Solubility and Orthophosphate Addition in High pH Low DIC Water Master of Science Department of Biomedical Chemical and Environmental Engineering College of Engineering and Applied Science University of Cincinnati Cincinnati OH 2014
Ortho-Ppoint of diminishing returns
bull Orthophosphate addition to where large increments result in small reductions in lead release
bull Key to cost-effective lead release control and exposure reduction
bull Becomes the rdquomaintenancerdquo dosage unless the scale ldquoagesrdquo
bull Background constituents such as aluminum hardness ions iron manganese and others interfere with optimum reduction of lead release
bull Varies with the background water chemistry from system to system
23
Sodium silicate
bull No systematic studies to look at pH carbonate silicate background chemistry relationships
bull Unclear mechanism pH increase Buffer intensity increase Protective film (thin amp amorphous)
bull Can sequester ironmanganese
bull Canrsquot be evaluated with fresh surfaces
24pH5 6 7 8 9 10 11
β ty
nsi
nte
Iuf
fer
B
00000
00005
00010TIC=48 mg CL20 mgL SiO250 mgL PO4β H2OTotal β
Silicate may contribute to Buffer Intensity
Carbonate orthophosphate silicate I=001 25ordmC
Chemical changes may reduce Pb+4 to Pb+2
10
8
6
4
)ts 2ol
h (v
E
ndash2
ndash4
ndash6
ndash8
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
pH
12
PbO2 (plattnerite)
Pb++ )s
deg
(2
CO
3
(OH
) -2 2) 3
Pb
-2
0 (CO
-- 2
0
) 43(C
O
Pb
3 (OH
)
DIC = 18 mg CL
Pb Pb Pb = 0010 mgL
Pb metal
-10
Drop in ORP from treatment change or DS
oxidant demand
Drop in pH at surface from treatment change
chemical reactions nitrification etc
C D
ngto
nhisa
W
Newark NJ
Disinfectant demand in DS must be controlled amp enough free chlorine consistently maintained throughout LSL area to keep protective Pb+4 corrosion scales in place
Switching disinfectants may reduce Pb+4 to Pb+2
Low
Switch from chlorine to chloramine disinfectant dissolved lead from pipe scales during the Washington DC ldquolead-in-water crisisrdquo in 2001-2004
Example profile of Pb(+4) (ie PbO2) scale house
27
20
L)
Cl1 Cincinnati OH
18
microg 16
n ( 14
o 12
tia 10
trn 8
e 6
nco 4
C 2
b P 0-1 0 1 2 3 4 5 6 7 8 9 10 11
Cumulative Water Volume (L)
CI-1 10 h 1209CI-1 10 h 0410
Plumbing sequence corresponding to water volumeFaucet LSLCu pipe Brass
Water MainMCu pipe 2
015
a
let
dou
illyaf
ant
irT
bull Pb release to water from PbO2 coated LSLs was found to be lowbull Be aware are there are no known cases of exhumed PbO2 coated pipes being able to
re-stabilize in pipe rigs
Chloride and sulfateCSMR a factor controlling galvanic corrosion (Oliphant 1983 Gregory 1986)- Chloride detrimental- Sulfate beneficial if sufficient to overcome chloride thus- CSMR ratio was created
[Cl- ] 12 mgL Cl-Chloride to Sulfate Mass Ratio (CSMR) = =[SO -2 2 = 06
4 ] 20 mgL SO-4
Confirmed empirically in case studies of water utilities (Dodrill and Edwards 1995 Edwards and Triantafyllidou 2007) - High CSMR gt 05-06 more corrosive- Low CSMR lt 05-06 less corrosive
28
Verified in laboratory experiments with fresh galvanic Pb junctions (eg Wang et al 2013 Triantafyllidou et al 2011 Nguyen et al 2010 2011)
Tria
ntaf
yllid
ou e
t al
201
1
Chloride and sulfate
29
CSMR considered contributing factor to the Flint water Pb crisis (Pieper et al 2017)
30
Chloride and sulfate
bull Absolute concentrations of chloride and sulfate important chloride considered particularly detrimental to galvanic Pb corrosion
bull Increasing chloride and sulfate concentrations at a constant CSMR of 1 increased total lead concentrations in galvanic corrosion experiments (Ng amp Lin 2016)
bull High CSMR ratio proven empirically important in aggravating galvanic lead corrosion (verified by experiments and modeling)
bull Water source or treatment changes that affect the CSMR must be evaluated prior to change
bull Absolute concentrations also important particularly for chloride
31
Manganese
Two distinct pipe scale categories in 5 LSLs (Lead 1-5 Pipes) excavated from same city
31
Schock et al 2014 JAWWAbull The outer pipe scale is in direct contact with the flowing water affecting
water lead contamination the mostbull Plattnerite an insoluble Pb(+4) scale that protects water from lead
contamination was inhibited by Mn (and Fe) presence
Manganese
In this case study (Schock et al JAWWA 2014)bull Where Mn was present in more abundance
- Pb scales had less PbO2- Pb associated with Mn sorbing onto Mn+Fe-rich surface deposit
bull Erratic Pb persisted after full LSLR in areas rich with Mn deposits (and to lesser extent Fe)
bull Galvanized interior plumbing also accumulated dissolved Pb while LSL present
32
33
Manganese
Other studiesbull Dong et al 2000 Mn acted as vector for particulate Pb movementbull Trueman et al 2019 Increasing Mn increased Pb in experimentbull McNeill and Edwards 2004 Mn sequestration by polyphosphate destabilized
Pb scales and caused Pb release
bull Pan et al 2019 Possible benefit by promoting formation of Pb(+4)
Reference Analysis Mn Impact on PbSchock et al 2014 Excavated LSLs
amp historical water Pb data in a US city
bull Promoted amorphous Pb scalebull Inhibited insoluble Pb(+4) scalebull Erratic Pb persisted after full LSLR in areas rich
with Mn deposits
Tully et al 2019 Excavated LSLs amp water Pb data in several Midwest case studies
bull Mn was a component of amorphous pipe scales in several systems
bull Amorphous scales not amenable to determining their adherence to pipe walls their responses to changes in various water chemistry variables or quantitatively calculating their lead release
Itrsquos not just manganese
34
Broad variety of coatings on LSLs analyzed at EPA
Al Mn Fe P and Ca-rich coatings may interfere withorthophosphate pH adjustment or other CCT
35
More water parameters affect metal release
bull pH amp AlkalinityDICbull Concentration and type of
oxidants (disinfectantsdissolved oxygen)
bull ORPcorrosion potentialbull Corrosion Inhibitorsbull Chloridebull Sulfatebull Manganese
bull Temperaturebull Sorptive surfaces downstream of
LSLs (ie galvanized interior pipe)bull Iron (deposition and corrosion)bull Calciumbull Aluminumbull NOM (type amount)bull Mixing from water treatment plants
or sourcesbull Ammoniabull Hydrogen Sulfidebull Microbial activity (nitrification amp
other)
Corrosion Control Treatment is intertwined with all treatments affecting water chemistry
Water Quality Parameters(WQP)
bull pH alkalinity calcium orthophosphate or silica othersbull WQP measurements are important for
bull Process control (measurement frequency needs to be HIGH)bull Keeping treatment from being turned offbull Keeping inhibitor dosages amp pH control from being cut to save $$$bull Keeping required repairs to treatment on tight schedulebull Preventing frequent variations and alternations of water sources
bull Effectiveness of Pb treatment only determinable from tap sampling by combination of different assessment approaches for different lead plumbing sources to which people can be exposed
36
bull OWQP measurements important for process control look at yourdata to understand what they mean and avoid problems
bull OWQP do not predict lead release There are no substitutes fordirectly monitoring lead release
37
M+
M+
M+
M+
Corrosion barrier film hypothesis
Inert barrier film (eg CaCO3)
Pipe Wall
Barrier film would allow tobull Protect ALL pipe materialsbull Monitor corrosion based on one chemical reaction
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
Illustrative reactions of P in pipebull OrthophosphatePrecipitation (consumption of orthophosphate)3Pb+2 + 2PO4
-3 --gt Pb3(PO4)2 (solid)decreased Pb solubility
bull PolyphosphatesReversion partially (formation of orthophosphate)P3O10
-5 ---gt P2O7-3 + PO4
-3
polymer --gt monomersComplexation (soluble complexes)3Pb+2 + P3O10
-5 --gt PbP3O10-2
higher Pb solubility
Slid
e cr
edit
Mar
c Ed
war
ds
Conceptual possible reactions Actual reactants and products may bedifferent Polyphosphates cannot be easily quantified
20
Ortho-P Treatment for Pb+2
Effectiveness Depends on Dose DIC pH and ldquoCleanlinessrdquo of Pipe Surface
mg PO4L00 10 20 30 40 50
mg
PbL
000
005
010
015
020
025
030
035
040
48 mg CL
48 mg CL
pH = 70pH = 75pH = 80pH = 85
bull pH less critical at low DICbull pH less critical at high PO4bull Point of diminishing returns higher with high DIC
bull Faster Pb reduction at high PO4
Typical UK Dosages 4-6 mgL
Most PWSs with LSLs do not have corrosion control treatment that minimizes Pb release and exposure
Effect of pH and ortho-P on Pb release
21
DIC = 10 mg CL 1 mg PO4L
60 65 70 75 80 85 90 95
gL
b m
P
0001
001
01
1
10
US Action Level
Dissolved Pb with ortho-PTotal Pb with ortho-PDissolved Pb with CO3Total Pb with CO3
At low DIC orthophosphate improves lead release regardless of pH
pHSchock M R DeSantis M K Metz D H Welch M M Hyland R N Nadagouda M N Revisiting the pH Effect on the Orthophosphate Control of Plumbosolvency Proc AWWA Annual Conference and Exposition Atlanta GA 2008
Ortho-P at pH 90 (DIC 6 mgL)
22
Pb(μgL)
bull Low Pb at 1 mgL relative to higher doses needs to be further verified across water qualities
bull Must do dose optimization study for your own water quality especially at high pH
bull Ortho-P may precipitate with Ca
Miller S A Investigation of Lead Solubility and Orthophosphate Addition in High pH Low DIC Water Master of Science Department of Biomedical Chemical and Environmental Engineering College of Engineering and Applied Science University of Cincinnati Cincinnati OH 2014
Ortho-Ppoint of diminishing returns
bull Orthophosphate addition to where large increments result in small reductions in lead release
bull Key to cost-effective lead release control and exposure reduction
bull Becomes the rdquomaintenancerdquo dosage unless the scale ldquoagesrdquo
bull Background constituents such as aluminum hardness ions iron manganese and others interfere with optimum reduction of lead release
bull Varies with the background water chemistry from system to system
23
Sodium silicate
bull No systematic studies to look at pH carbonate silicate background chemistry relationships
bull Unclear mechanism pH increase Buffer intensity increase Protective film (thin amp amorphous)
bull Can sequester ironmanganese
bull Canrsquot be evaluated with fresh surfaces
24pH5 6 7 8 9 10 11
β ty
nsi
nte
Iuf
fer
B
00000
00005
00010TIC=48 mg CL20 mgL SiO250 mgL PO4β H2OTotal β
Silicate may contribute to Buffer Intensity
Carbonate orthophosphate silicate I=001 25ordmC
Chemical changes may reduce Pb+4 to Pb+2
10
8
6
4
)ts 2ol
h (v
E
ndash2
ndash4
ndash6
ndash8
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
pH
12
PbO2 (plattnerite)
Pb++ )s
deg
(2
CO
3
(OH
) -2 2) 3
Pb
-2
0 (CO
-- 2
0
) 43(C
O
Pb
3 (OH
)
DIC = 18 mg CL
Pb Pb Pb = 0010 mgL
Pb metal
-10
Drop in ORP from treatment change or DS
oxidant demand
Drop in pH at surface from treatment change
chemical reactions nitrification etc
C D
ngto
nhisa
W
Newark NJ
Disinfectant demand in DS must be controlled amp enough free chlorine consistently maintained throughout LSL area to keep protective Pb+4 corrosion scales in place
Switching disinfectants may reduce Pb+4 to Pb+2
Low
Switch from chlorine to chloramine disinfectant dissolved lead from pipe scales during the Washington DC ldquolead-in-water crisisrdquo in 2001-2004
Example profile of Pb(+4) (ie PbO2) scale house
27
20
L)
Cl1 Cincinnati OH
18
microg 16
n ( 14
o 12
tia 10
trn 8
e 6
nco 4
C 2
b P 0-1 0 1 2 3 4 5 6 7 8 9 10 11
Cumulative Water Volume (L)
CI-1 10 h 1209CI-1 10 h 0410
Plumbing sequence corresponding to water volumeFaucet LSLCu pipe Brass
Water MainMCu pipe 2
015
a
let
dou
illyaf
ant
irT
bull Pb release to water from PbO2 coated LSLs was found to be lowbull Be aware are there are no known cases of exhumed PbO2 coated pipes being able to
re-stabilize in pipe rigs
Chloride and sulfateCSMR a factor controlling galvanic corrosion (Oliphant 1983 Gregory 1986)- Chloride detrimental- Sulfate beneficial if sufficient to overcome chloride thus- CSMR ratio was created
[Cl- ] 12 mgL Cl-Chloride to Sulfate Mass Ratio (CSMR) = =[SO -2 2 = 06
4 ] 20 mgL SO-4
Confirmed empirically in case studies of water utilities (Dodrill and Edwards 1995 Edwards and Triantafyllidou 2007) - High CSMR gt 05-06 more corrosive- Low CSMR lt 05-06 less corrosive
28
Verified in laboratory experiments with fresh galvanic Pb junctions (eg Wang et al 2013 Triantafyllidou et al 2011 Nguyen et al 2010 2011)
Tria
ntaf
yllid
ou e
t al
201
1
Chloride and sulfate
29
CSMR considered contributing factor to the Flint water Pb crisis (Pieper et al 2017)
30
Chloride and sulfate
bull Absolute concentrations of chloride and sulfate important chloride considered particularly detrimental to galvanic Pb corrosion
bull Increasing chloride and sulfate concentrations at a constant CSMR of 1 increased total lead concentrations in galvanic corrosion experiments (Ng amp Lin 2016)
bull High CSMR ratio proven empirically important in aggravating galvanic lead corrosion (verified by experiments and modeling)
bull Water source or treatment changes that affect the CSMR must be evaluated prior to change
bull Absolute concentrations also important particularly for chloride
31
Manganese
Two distinct pipe scale categories in 5 LSLs (Lead 1-5 Pipes) excavated from same city
31
Schock et al 2014 JAWWAbull The outer pipe scale is in direct contact with the flowing water affecting
water lead contamination the mostbull Plattnerite an insoluble Pb(+4) scale that protects water from lead
contamination was inhibited by Mn (and Fe) presence
Manganese
In this case study (Schock et al JAWWA 2014)bull Where Mn was present in more abundance
- Pb scales had less PbO2- Pb associated with Mn sorbing onto Mn+Fe-rich surface deposit
bull Erratic Pb persisted after full LSLR in areas rich with Mn deposits (and to lesser extent Fe)
bull Galvanized interior plumbing also accumulated dissolved Pb while LSL present
32
33
Manganese
Other studiesbull Dong et al 2000 Mn acted as vector for particulate Pb movementbull Trueman et al 2019 Increasing Mn increased Pb in experimentbull McNeill and Edwards 2004 Mn sequestration by polyphosphate destabilized
Pb scales and caused Pb release
bull Pan et al 2019 Possible benefit by promoting formation of Pb(+4)
Reference Analysis Mn Impact on PbSchock et al 2014 Excavated LSLs
amp historical water Pb data in a US city
bull Promoted amorphous Pb scalebull Inhibited insoluble Pb(+4) scalebull Erratic Pb persisted after full LSLR in areas rich
with Mn deposits
Tully et al 2019 Excavated LSLs amp water Pb data in several Midwest case studies
bull Mn was a component of amorphous pipe scales in several systems
bull Amorphous scales not amenable to determining their adherence to pipe walls their responses to changes in various water chemistry variables or quantitatively calculating their lead release
Itrsquos not just manganese
34
Broad variety of coatings on LSLs analyzed at EPA
Al Mn Fe P and Ca-rich coatings may interfere withorthophosphate pH adjustment or other CCT
35
More water parameters affect metal release
bull pH amp AlkalinityDICbull Concentration and type of
oxidants (disinfectantsdissolved oxygen)
bull ORPcorrosion potentialbull Corrosion Inhibitorsbull Chloridebull Sulfatebull Manganese
bull Temperaturebull Sorptive surfaces downstream of
LSLs (ie galvanized interior pipe)bull Iron (deposition and corrosion)bull Calciumbull Aluminumbull NOM (type amount)bull Mixing from water treatment plants
or sourcesbull Ammoniabull Hydrogen Sulfidebull Microbial activity (nitrification amp
other)
Corrosion Control Treatment is intertwined with all treatments affecting water chemistry
Water Quality Parameters(WQP)
bull pH alkalinity calcium orthophosphate or silica othersbull WQP measurements are important for
bull Process control (measurement frequency needs to be HIGH)bull Keeping treatment from being turned offbull Keeping inhibitor dosages amp pH control from being cut to save $$$bull Keeping required repairs to treatment on tight schedulebull Preventing frequent variations and alternations of water sources
bull Effectiveness of Pb treatment only determinable from tap sampling by combination of different assessment approaches for different lead plumbing sources to which people can be exposed
36
bull OWQP measurements important for process control look at yourdata to understand what they mean and avoid problems
bull OWQP do not predict lead release There are no substitutes fordirectly monitoring lead release
37
M+
M+
M+
M+
Corrosion barrier film hypothesis
Inert barrier film (eg CaCO3)
Pipe Wall
Barrier film would allow tobull Protect ALL pipe materialsbull Monitor corrosion based on one chemical reaction
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
20
Ortho-P Treatment for Pb+2
Effectiveness Depends on Dose DIC pH and ldquoCleanlinessrdquo of Pipe Surface
mg PO4L00 10 20 30 40 50
mg
PbL
000
005
010
015
020
025
030
035
040
48 mg CL
48 mg CL
pH = 70pH = 75pH = 80pH = 85
bull pH less critical at low DICbull pH less critical at high PO4bull Point of diminishing returns higher with high DIC
bull Faster Pb reduction at high PO4
Typical UK Dosages 4-6 mgL
Most PWSs with LSLs do not have corrosion control treatment that minimizes Pb release and exposure
Effect of pH and ortho-P on Pb release
21
DIC = 10 mg CL 1 mg PO4L
60 65 70 75 80 85 90 95
gL
b m
P
0001
001
01
1
10
US Action Level
Dissolved Pb with ortho-PTotal Pb with ortho-PDissolved Pb with CO3Total Pb with CO3
At low DIC orthophosphate improves lead release regardless of pH
pHSchock M R DeSantis M K Metz D H Welch M M Hyland R N Nadagouda M N Revisiting the pH Effect on the Orthophosphate Control of Plumbosolvency Proc AWWA Annual Conference and Exposition Atlanta GA 2008
Ortho-P at pH 90 (DIC 6 mgL)
22
Pb(μgL)
bull Low Pb at 1 mgL relative to higher doses needs to be further verified across water qualities
bull Must do dose optimization study for your own water quality especially at high pH
bull Ortho-P may precipitate with Ca
Miller S A Investigation of Lead Solubility and Orthophosphate Addition in High pH Low DIC Water Master of Science Department of Biomedical Chemical and Environmental Engineering College of Engineering and Applied Science University of Cincinnati Cincinnati OH 2014
Ortho-Ppoint of diminishing returns
bull Orthophosphate addition to where large increments result in small reductions in lead release
bull Key to cost-effective lead release control and exposure reduction
bull Becomes the rdquomaintenancerdquo dosage unless the scale ldquoagesrdquo
bull Background constituents such as aluminum hardness ions iron manganese and others interfere with optimum reduction of lead release
bull Varies with the background water chemistry from system to system
23
Sodium silicate
bull No systematic studies to look at pH carbonate silicate background chemistry relationships
bull Unclear mechanism pH increase Buffer intensity increase Protective film (thin amp amorphous)
bull Can sequester ironmanganese
bull Canrsquot be evaluated with fresh surfaces
24pH5 6 7 8 9 10 11
β ty
nsi
nte
Iuf
fer
B
00000
00005
00010TIC=48 mg CL20 mgL SiO250 mgL PO4β H2OTotal β
Silicate may contribute to Buffer Intensity
Carbonate orthophosphate silicate I=001 25ordmC
Chemical changes may reduce Pb+4 to Pb+2
10
8
6
4
)ts 2ol
h (v
E
ndash2
ndash4
ndash6
ndash8
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
pH
12
PbO2 (plattnerite)
Pb++ )s
deg
(2
CO
3
(OH
) -2 2) 3
Pb
-2
0 (CO
-- 2
0
) 43(C
O
Pb
3 (OH
)
DIC = 18 mg CL
Pb Pb Pb = 0010 mgL
Pb metal
-10
Drop in ORP from treatment change or DS
oxidant demand
Drop in pH at surface from treatment change
chemical reactions nitrification etc
C D
ngto
nhisa
W
Newark NJ
Disinfectant demand in DS must be controlled amp enough free chlorine consistently maintained throughout LSL area to keep protective Pb+4 corrosion scales in place
Switching disinfectants may reduce Pb+4 to Pb+2
Low
Switch from chlorine to chloramine disinfectant dissolved lead from pipe scales during the Washington DC ldquolead-in-water crisisrdquo in 2001-2004
Example profile of Pb(+4) (ie PbO2) scale house
27
20
L)
Cl1 Cincinnati OH
18
microg 16
n ( 14
o 12
tia 10
trn 8
e 6
nco 4
C 2
b P 0-1 0 1 2 3 4 5 6 7 8 9 10 11
Cumulative Water Volume (L)
CI-1 10 h 1209CI-1 10 h 0410
Plumbing sequence corresponding to water volumeFaucet LSLCu pipe Brass
Water MainMCu pipe 2
015
a
let
dou
illyaf
ant
irT
bull Pb release to water from PbO2 coated LSLs was found to be lowbull Be aware are there are no known cases of exhumed PbO2 coated pipes being able to
re-stabilize in pipe rigs
Chloride and sulfateCSMR a factor controlling galvanic corrosion (Oliphant 1983 Gregory 1986)- Chloride detrimental- Sulfate beneficial if sufficient to overcome chloride thus- CSMR ratio was created
[Cl- ] 12 mgL Cl-Chloride to Sulfate Mass Ratio (CSMR) = =[SO -2 2 = 06
4 ] 20 mgL SO-4
Confirmed empirically in case studies of water utilities (Dodrill and Edwards 1995 Edwards and Triantafyllidou 2007) - High CSMR gt 05-06 more corrosive- Low CSMR lt 05-06 less corrosive
28
Verified in laboratory experiments with fresh galvanic Pb junctions (eg Wang et al 2013 Triantafyllidou et al 2011 Nguyen et al 2010 2011)
Tria
ntaf
yllid
ou e
t al
201
1
Chloride and sulfate
29
CSMR considered contributing factor to the Flint water Pb crisis (Pieper et al 2017)
30
Chloride and sulfate
bull Absolute concentrations of chloride and sulfate important chloride considered particularly detrimental to galvanic Pb corrosion
bull Increasing chloride and sulfate concentrations at a constant CSMR of 1 increased total lead concentrations in galvanic corrosion experiments (Ng amp Lin 2016)
bull High CSMR ratio proven empirically important in aggravating galvanic lead corrosion (verified by experiments and modeling)
bull Water source or treatment changes that affect the CSMR must be evaluated prior to change
bull Absolute concentrations also important particularly for chloride
31
Manganese
Two distinct pipe scale categories in 5 LSLs (Lead 1-5 Pipes) excavated from same city
31
Schock et al 2014 JAWWAbull The outer pipe scale is in direct contact with the flowing water affecting
water lead contamination the mostbull Plattnerite an insoluble Pb(+4) scale that protects water from lead
contamination was inhibited by Mn (and Fe) presence
Manganese
In this case study (Schock et al JAWWA 2014)bull Where Mn was present in more abundance
- Pb scales had less PbO2- Pb associated with Mn sorbing onto Mn+Fe-rich surface deposit
bull Erratic Pb persisted after full LSLR in areas rich with Mn deposits (and to lesser extent Fe)
bull Galvanized interior plumbing also accumulated dissolved Pb while LSL present
32
33
Manganese
Other studiesbull Dong et al 2000 Mn acted as vector for particulate Pb movementbull Trueman et al 2019 Increasing Mn increased Pb in experimentbull McNeill and Edwards 2004 Mn sequestration by polyphosphate destabilized
Pb scales and caused Pb release
bull Pan et al 2019 Possible benefit by promoting formation of Pb(+4)
Reference Analysis Mn Impact on PbSchock et al 2014 Excavated LSLs
amp historical water Pb data in a US city
bull Promoted amorphous Pb scalebull Inhibited insoluble Pb(+4) scalebull Erratic Pb persisted after full LSLR in areas rich
with Mn deposits
Tully et al 2019 Excavated LSLs amp water Pb data in several Midwest case studies
bull Mn was a component of amorphous pipe scales in several systems
bull Amorphous scales not amenable to determining their adherence to pipe walls their responses to changes in various water chemistry variables or quantitatively calculating their lead release
Itrsquos not just manganese
34
Broad variety of coatings on LSLs analyzed at EPA
Al Mn Fe P and Ca-rich coatings may interfere withorthophosphate pH adjustment or other CCT
35
More water parameters affect metal release
bull pH amp AlkalinityDICbull Concentration and type of
oxidants (disinfectantsdissolved oxygen)
bull ORPcorrosion potentialbull Corrosion Inhibitorsbull Chloridebull Sulfatebull Manganese
bull Temperaturebull Sorptive surfaces downstream of
LSLs (ie galvanized interior pipe)bull Iron (deposition and corrosion)bull Calciumbull Aluminumbull NOM (type amount)bull Mixing from water treatment plants
or sourcesbull Ammoniabull Hydrogen Sulfidebull Microbial activity (nitrification amp
other)
Corrosion Control Treatment is intertwined with all treatments affecting water chemistry
Water Quality Parameters(WQP)
bull pH alkalinity calcium orthophosphate or silica othersbull WQP measurements are important for
bull Process control (measurement frequency needs to be HIGH)bull Keeping treatment from being turned offbull Keeping inhibitor dosages amp pH control from being cut to save $$$bull Keeping required repairs to treatment on tight schedulebull Preventing frequent variations and alternations of water sources
bull Effectiveness of Pb treatment only determinable from tap sampling by combination of different assessment approaches for different lead plumbing sources to which people can be exposed
36
bull OWQP measurements important for process control look at yourdata to understand what they mean and avoid problems
bull OWQP do not predict lead release There are no substitutes fordirectly monitoring lead release
37
M+
M+
M+
M+
Corrosion barrier film hypothesis
Inert barrier film (eg CaCO3)
Pipe Wall
Barrier film would allow tobull Protect ALL pipe materialsbull Monitor corrosion based on one chemical reaction
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
Effect of pH and ortho-P on Pb release
21
DIC = 10 mg CL 1 mg PO4L
60 65 70 75 80 85 90 95
gL
b m
P
0001
001
01
1
10
US Action Level
Dissolved Pb with ortho-PTotal Pb with ortho-PDissolved Pb with CO3Total Pb with CO3
At low DIC orthophosphate improves lead release regardless of pH
pHSchock M R DeSantis M K Metz D H Welch M M Hyland R N Nadagouda M N Revisiting the pH Effect on the Orthophosphate Control of Plumbosolvency Proc AWWA Annual Conference and Exposition Atlanta GA 2008
Ortho-P at pH 90 (DIC 6 mgL)
22
Pb(μgL)
bull Low Pb at 1 mgL relative to higher doses needs to be further verified across water qualities
bull Must do dose optimization study for your own water quality especially at high pH
bull Ortho-P may precipitate with Ca
Miller S A Investigation of Lead Solubility and Orthophosphate Addition in High pH Low DIC Water Master of Science Department of Biomedical Chemical and Environmental Engineering College of Engineering and Applied Science University of Cincinnati Cincinnati OH 2014
Ortho-Ppoint of diminishing returns
bull Orthophosphate addition to where large increments result in small reductions in lead release
bull Key to cost-effective lead release control and exposure reduction
bull Becomes the rdquomaintenancerdquo dosage unless the scale ldquoagesrdquo
bull Background constituents such as aluminum hardness ions iron manganese and others interfere with optimum reduction of lead release
bull Varies with the background water chemistry from system to system
23
Sodium silicate
bull No systematic studies to look at pH carbonate silicate background chemistry relationships
bull Unclear mechanism pH increase Buffer intensity increase Protective film (thin amp amorphous)
bull Can sequester ironmanganese
bull Canrsquot be evaluated with fresh surfaces
24pH5 6 7 8 9 10 11
β ty
nsi
nte
Iuf
fer
B
00000
00005
00010TIC=48 mg CL20 mgL SiO250 mgL PO4β H2OTotal β
Silicate may contribute to Buffer Intensity
Carbonate orthophosphate silicate I=001 25ordmC
Chemical changes may reduce Pb+4 to Pb+2
10
8
6
4
)ts 2ol
h (v
E
ndash2
ndash4
ndash6
ndash8
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
pH
12
PbO2 (plattnerite)
Pb++ )s
deg
(2
CO
3
(OH
) -2 2) 3
Pb
-2
0 (CO
-- 2
0
) 43(C
O
Pb
3 (OH
)
DIC = 18 mg CL
Pb Pb Pb = 0010 mgL
Pb metal
-10
Drop in ORP from treatment change or DS
oxidant demand
Drop in pH at surface from treatment change
chemical reactions nitrification etc
C D
ngto
nhisa
W
Newark NJ
Disinfectant demand in DS must be controlled amp enough free chlorine consistently maintained throughout LSL area to keep protective Pb+4 corrosion scales in place
Switching disinfectants may reduce Pb+4 to Pb+2
Low
Switch from chlorine to chloramine disinfectant dissolved lead from pipe scales during the Washington DC ldquolead-in-water crisisrdquo in 2001-2004
Example profile of Pb(+4) (ie PbO2) scale house
27
20
L)
Cl1 Cincinnati OH
18
microg 16
n ( 14
o 12
tia 10
trn 8
e 6
nco 4
C 2
b P 0-1 0 1 2 3 4 5 6 7 8 9 10 11
Cumulative Water Volume (L)
CI-1 10 h 1209CI-1 10 h 0410
Plumbing sequence corresponding to water volumeFaucet LSLCu pipe Brass
Water MainMCu pipe 2
015
a
let
dou
illyaf
ant
irT
bull Pb release to water from PbO2 coated LSLs was found to be lowbull Be aware are there are no known cases of exhumed PbO2 coated pipes being able to
re-stabilize in pipe rigs
Chloride and sulfateCSMR a factor controlling galvanic corrosion (Oliphant 1983 Gregory 1986)- Chloride detrimental- Sulfate beneficial if sufficient to overcome chloride thus- CSMR ratio was created
[Cl- ] 12 mgL Cl-Chloride to Sulfate Mass Ratio (CSMR) = =[SO -2 2 = 06
4 ] 20 mgL SO-4
Confirmed empirically in case studies of water utilities (Dodrill and Edwards 1995 Edwards and Triantafyllidou 2007) - High CSMR gt 05-06 more corrosive- Low CSMR lt 05-06 less corrosive
28
Verified in laboratory experiments with fresh galvanic Pb junctions (eg Wang et al 2013 Triantafyllidou et al 2011 Nguyen et al 2010 2011)
Tria
ntaf
yllid
ou e
t al
201
1
Chloride and sulfate
29
CSMR considered contributing factor to the Flint water Pb crisis (Pieper et al 2017)
30
Chloride and sulfate
bull Absolute concentrations of chloride and sulfate important chloride considered particularly detrimental to galvanic Pb corrosion
bull Increasing chloride and sulfate concentrations at a constant CSMR of 1 increased total lead concentrations in galvanic corrosion experiments (Ng amp Lin 2016)
bull High CSMR ratio proven empirically important in aggravating galvanic lead corrosion (verified by experiments and modeling)
bull Water source or treatment changes that affect the CSMR must be evaluated prior to change
bull Absolute concentrations also important particularly for chloride
31
Manganese
Two distinct pipe scale categories in 5 LSLs (Lead 1-5 Pipes) excavated from same city
31
Schock et al 2014 JAWWAbull The outer pipe scale is in direct contact with the flowing water affecting
water lead contamination the mostbull Plattnerite an insoluble Pb(+4) scale that protects water from lead
contamination was inhibited by Mn (and Fe) presence
Manganese
In this case study (Schock et al JAWWA 2014)bull Where Mn was present in more abundance
- Pb scales had less PbO2- Pb associated with Mn sorbing onto Mn+Fe-rich surface deposit
bull Erratic Pb persisted after full LSLR in areas rich with Mn deposits (and to lesser extent Fe)
bull Galvanized interior plumbing also accumulated dissolved Pb while LSL present
32
33
Manganese
Other studiesbull Dong et al 2000 Mn acted as vector for particulate Pb movementbull Trueman et al 2019 Increasing Mn increased Pb in experimentbull McNeill and Edwards 2004 Mn sequestration by polyphosphate destabilized
Pb scales and caused Pb release
bull Pan et al 2019 Possible benefit by promoting formation of Pb(+4)
Reference Analysis Mn Impact on PbSchock et al 2014 Excavated LSLs
amp historical water Pb data in a US city
bull Promoted amorphous Pb scalebull Inhibited insoluble Pb(+4) scalebull Erratic Pb persisted after full LSLR in areas rich
with Mn deposits
Tully et al 2019 Excavated LSLs amp water Pb data in several Midwest case studies
bull Mn was a component of amorphous pipe scales in several systems
bull Amorphous scales not amenable to determining their adherence to pipe walls their responses to changes in various water chemistry variables or quantitatively calculating their lead release
Itrsquos not just manganese
34
Broad variety of coatings on LSLs analyzed at EPA
Al Mn Fe P and Ca-rich coatings may interfere withorthophosphate pH adjustment or other CCT
35
More water parameters affect metal release
bull pH amp AlkalinityDICbull Concentration and type of
oxidants (disinfectantsdissolved oxygen)
bull ORPcorrosion potentialbull Corrosion Inhibitorsbull Chloridebull Sulfatebull Manganese
bull Temperaturebull Sorptive surfaces downstream of
LSLs (ie galvanized interior pipe)bull Iron (deposition and corrosion)bull Calciumbull Aluminumbull NOM (type amount)bull Mixing from water treatment plants
or sourcesbull Ammoniabull Hydrogen Sulfidebull Microbial activity (nitrification amp
other)
Corrosion Control Treatment is intertwined with all treatments affecting water chemistry
Water Quality Parameters(WQP)
bull pH alkalinity calcium orthophosphate or silica othersbull WQP measurements are important for
bull Process control (measurement frequency needs to be HIGH)bull Keeping treatment from being turned offbull Keeping inhibitor dosages amp pH control from being cut to save $$$bull Keeping required repairs to treatment on tight schedulebull Preventing frequent variations and alternations of water sources
bull Effectiveness of Pb treatment only determinable from tap sampling by combination of different assessment approaches for different lead plumbing sources to which people can be exposed
36
bull OWQP measurements important for process control look at yourdata to understand what they mean and avoid problems
bull OWQP do not predict lead release There are no substitutes fordirectly monitoring lead release
37
M+
M+
M+
M+
Corrosion barrier film hypothesis
Inert barrier film (eg CaCO3)
Pipe Wall
Barrier film would allow tobull Protect ALL pipe materialsbull Monitor corrosion based on one chemical reaction
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
Ortho-P at pH 90 (DIC 6 mgL)
22
Pb(μgL)
bull Low Pb at 1 mgL relative to higher doses needs to be further verified across water qualities
bull Must do dose optimization study for your own water quality especially at high pH
bull Ortho-P may precipitate with Ca
Miller S A Investigation of Lead Solubility and Orthophosphate Addition in High pH Low DIC Water Master of Science Department of Biomedical Chemical and Environmental Engineering College of Engineering and Applied Science University of Cincinnati Cincinnati OH 2014
Ortho-Ppoint of diminishing returns
bull Orthophosphate addition to where large increments result in small reductions in lead release
bull Key to cost-effective lead release control and exposure reduction
bull Becomes the rdquomaintenancerdquo dosage unless the scale ldquoagesrdquo
bull Background constituents such as aluminum hardness ions iron manganese and others interfere with optimum reduction of lead release
bull Varies with the background water chemistry from system to system
23
Sodium silicate
bull No systematic studies to look at pH carbonate silicate background chemistry relationships
bull Unclear mechanism pH increase Buffer intensity increase Protective film (thin amp amorphous)
bull Can sequester ironmanganese
bull Canrsquot be evaluated with fresh surfaces
24pH5 6 7 8 9 10 11
β ty
nsi
nte
Iuf
fer
B
00000
00005
00010TIC=48 mg CL20 mgL SiO250 mgL PO4β H2OTotal β
Silicate may contribute to Buffer Intensity
Carbonate orthophosphate silicate I=001 25ordmC
Chemical changes may reduce Pb+4 to Pb+2
10
8
6
4
)ts 2ol
h (v
E
ndash2
ndash4
ndash6
ndash8
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
pH
12
PbO2 (plattnerite)
Pb++ )s
deg
(2
CO
3
(OH
) -2 2) 3
Pb
-2
0 (CO
-- 2
0
) 43(C
O
Pb
3 (OH
)
DIC = 18 mg CL
Pb Pb Pb = 0010 mgL
Pb metal
-10
Drop in ORP from treatment change or DS
oxidant demand
Drop in pH at surface from treatment change
chemical reactions nitrification etc
C D
ngto
nhisa
W
Newark NJ
Disinfectant demand in DS must be controlled amp enough free chlorine consistently maintained throughout LSL area to keep protective Pb+4 corrosion scales in place
Switching disinfectants may reduce Pb+4 to Pb+2
Low
Switch from chlorine to chloramine disinfectant dissolved lead from pipe scales during the Washington DC ldquolead-in-water crisisrdquo in 2001-2004
Example profile of Pb(+4) (ie PbO2) scale house
27
20
L)
Cl1 Cincinnati OH
18
microg 16
n ( 14
o 12
tia 10
trn 8
e 6
nco 4
C 2
b P 0-1 0 1 2 3 4 5 6 7 8 9 10 11
Cumulative Water Volume (L)
CI-1 10 h 1209CI-1 10 h 0410
Plumbing sequence corresponding to water volumeFaucet LSLCu pipe Brass
Water MainMCu pipe 2
015
a
let
dou
illyaf
ant
irT
bull Pb release to water from PbO2 coated LSLs was found to be lowbull Be aware are there are no known cases of exhumed PbO2 coated pipes being able to
re-stabilize in pipe rigs
Chloride and sulfateCSMR a factor controlling galvanic corrosion (Oliphant 1983 Gregory 1986)- Chloride detrimental- Sulfate beneficial if sufficient to overcome chloride thus- CSMR ratio was created
[Cl- ] 12 mgL Cl-Chloride to Sulfate Mass Ratio (CSMR) = =[SO -2 2 = 06
4 ] 20 mgL SO-4
Confirmed empirically in case studies of water utilities (Dodrill and Edwards 1995 Edwards and Triantafyllidou 2007) - High CSMR gt 05-06 more corrosive- Low CSMR lt 05-06 less corrosive
28
Verified in laboratory experiments with fresh galvanic Pb junctions (eg Wang et al 2013 Triantafyllidou et al 2011 Nguyen et al 2010 2011)
Tria
ntaf
yllid
ou e
t al
201
1
Chloride and sulfate
29
CSMR considered contributing factor to the Flint water Pb crisis (Pieper et al 2017)
30
Chloride and sulfate
bull Absolute concentrations of chloride and sulfate important chloride considered particularly detrimental to galvanic Pb corrosion
bull Increasing chloride and sulfate concentrations at a constant CSMR of 1 increased total lead concentrations in galvanic corrosion experiments (Ng amp Lin 2016)
bull High CSMR ratio proven empirically important in aggravating galvanic lead corrosion (verified by experiments and modeling)
bull Water source or treatment changes that affect the CSMR must be evaluated prior to change
bull Absolute concentrations also important particularly for chloride
31
Manganese
Two distinct pipe scale categories in 5 LSLs (Lead 1-5 Pipes) excavated from same city
31
Schock et al 2014 JAWWAbull The outer pipe scale is in direct contact with the flowing water affecting
water lead contamination the mostbull Plattnerite an insoluble Pb(+4) scale that protects water from lead
contamination was inhibited by Mn (and Fe) presence
Manganese
In this case study (Schock et al JAWWA 2014)bull Where Mn was present in more abundance
- Pb scales had less PbO2- Pb associated with Mn sorbing onto Mn+Fe-rich surface deposit
bull Erratic Pb persisted after full LSLR in areas rich with Mn deposits (and to lesser extent Fe)
bull Galvanized interior plumbing also accumulated dissolved Pb while LSL present
32
33
Manganese
Other studiesbull Dong et al 2000 Mn acted as vector for particulate Pb movementbull Trueman et al 2019 Increasing Mn increased Pb in experimentbull McNeill and Edwards 2004 Mn sequestration by polyphosphate destabilized
Pb scales and caused Pb release
bull Pan et al 2019 Possible benefit by promoting formation of Pb(+4)
Reference Analysis Mn Impact on PbSchock et al 2014 Excavated LSLs
amp historical water Pb data in a US city
bull Promoted amorphous Pb scalebull Inhibited insoluble Pb(+4) scalebull Erratic Pb persisted after full LSLR in areas rich
with Mn deposits
Tully et al 2019 Excavated LSLs amp water Pb data in several Midwest case studies
bull Mn was a component of amorphous pipe scales in several systems
bull Amorphous scales not amenable to determining their adherence to pipe walls their responses to changes in various water chemistry variables or quantitatively calculating their lead release
Itrsquos not just manganese
34
Broad variety of coatings on LSLs analyzed at EPA
Al Mn Fe P and Ca-rich coatings may interfere withorthophosphate pH adjustment or other CCT
35
More water parameters affect metal release
bull pH amp AlkalinityDICbull Concentration and type of
oxidants (disinfectantsdissolved oxygen)
bull ORPcorrosion potentialbull Corrosion Inhibitorsbull Chloridebull Sulfatebull Manganese
bull Temperaturebull Sorptive surfaces downstream of
LSLs (ie galvanized interior pipe)bull Iron (deposition and corrosion)bull Calciumbull Aluminumbull NOM (type amount)bull Mixing from water treatment plants
or sourcesbull Ammoniabull Hydrogen Sulfidebull Microbial activity (nitrification amp
other)
Corrosion Control Treatment is intertwined with all treatments affecting water chemistry
Water Quality Parameters(WQP)
bull pH alkalinity calcium orthophosphate or silica othersbull WQP measurements are important for
bull Process control (measurement frequency needs to be HIGH)bull Keeping treatment from being turned offbull Keeping inhibitor dosages amp pH control from being cut to save $$$bull Keeping required repairs to treatment on tight schedulebull Preventing frequent variations and alternations of water sources
bull Effectiveness of Pb treatment only determinable from tap sampling by combination of different assessment approaches for different lead plumbing sources to which people can be exposed
36
bull OWQP measurements important for process control look at yourdata to understand what they mean and avoid problems
bull OWQP do not predict lead release There are no substitutes fordirectly monitoring lead release
37
M+
M+
M+
M+
Corrosion barrier film hypothesis
Inert barrier film (eg CaCO3)
Pipe Wall
Barrier film would allow tobull Protect ALL pipe materialsbull Monitor corrosion based on one chemical reaction
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
Ortho-Ppoint of diminishing returns
bull Orthophosphate addition to where large increments result in small reductions in lead release
bull Key to cost-effective lead release control and exposure reduction
bull Becomes the rdquomaintenancerdquo dosage unless the scale ldquoagesrdquo
bull Background constituents such as aluminum hardness ions iron manganese and others interfere with optimum reduction of lead release
bull Varies with the background water chemistry from system to system
23
Sodium silicate
bull No systematic studies to look at pH carbonate silicate background chemistry relationships
bull Unclear mechanism pH increase Buffer intensity increase Protective film (thin amp amorphous)
bull Can sequester ironmanganese
bull Canrsquot be evaluated with fresh surfaces
24pH5 6 7 8 9 10 11
β ty
nsi
nte
Iuf
fer
B
00000
00005
00010TIC=48 mg CL20 mgL SiO250 mgL PO4β H2OTotal β
Silicate may contribute to Buffer Intensity
Carbonate orthophosphate silicate I=001 25ordmC
Chemical changes may reduce Pb+4 to Pb+2
10
8
6
4
)ts 2ol
h (v
E
ndash2
ndash4
ndash6
ndash8
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
pH
12
PbO2 (plattnerite)
Pb++ )s
deg
(2
CO
3
(OH
) -2 2) 3
Pb
-2
0 (CO
-- 2
0
) 43(C
O
Pb
3 (OH
)
DIC = 18 mg CL
Pb Pb Pb = 0010 mgL
Pb metal
-10
Drop in ORP from treatment change or DS
oxidant demand
Drop in pH at surface from treatment change
chemical reactions nitrification etc
C D
ngto
nhisa
W
Newark NJ
Disinfectant demand in DS must be controlled amp enough free chlorine consistently maintained throughout LSL area to keep protective Pb+4 corrosion scales in place
Switching disinfectants may reduce Pb+4 to Pb+2
Low
Switch from chlorine to chloramine disinfectant dissolved lead from pipe scales during the Washington DC ldquolead-in-water crisisrdquo in 2001-2004
Example profile of Pb(+4) (ie PbO2) scale house
27
20
L)
Cl1 Cincinnati OH
18
microg 16
n ( 14
o 12
tia 10
trn 8
e 6
nco 4
C 2
b P 0-1 0 1 2 3 4 5 6 7 8 9 10 11
Cumulative Water Volume (L)
CI-1 10 h 1209CI-1 10 h 0410
Plumbing sequence corresponding to water volumeFaucet LSLCu pipe Brass
Water MainMCu pipe 2
015
a
let
dou
illyaf
ant
irT
bull Pb release to water from PbO2 coated LSLs was found to be lowbull Be aware are there are no known cases of exhumed PbO2 coated pipes being able to
re-stabilize in pipe rigs
Chloride and sulfateCSMR a factor controlling galvanic corrosion (Oliphant 1983 Gregory 1986)- Chloride detrimental- Sulfate beneficial if sufficient to overcome chloride thus- CSMR ratio was created
[Cl- ] 12 mgL Cl-Chloride to Sulfate Mass Ratio (CSMR) = =[SO -2 2 = 06
4 ] 20 mgL SO-4
Confirmed empirically in case studies of water utilities (Dodrill and Edwards 1995 Edwards and Triantafyllidou 2007) - High CSMR gt 05-06 more corrosive- Low CSMR lt 05-06 less corrosive
28
Verified in laboratory experiments with fresh galvanic Pb junctions (eg Wang et al 2013 Triantafyllidou et al 2011 Nguyen et al 2010 2011)
Tria
ntaf
yllid
ou e
t al
201
1
Chloride and sulfate
29
CSMR considered contributing factor to the Flint water Pb crisis (Pieper et al 2017)
30
Chloride and sulfate
bull Absolute concentrations of chloride and sulfate important chloride considered particularly detrimental to galvanic Pb corrosion
bull Increasing chloride and sulfate concentrations at a constant CSMR of 1 increased total lead concentrations in galvanic corrosion experiments (Ng amp Lin 2016)
bull High CSMR ratio proven empirically important in aggravating galvanic lead corrosion (verified by experiments and modeling)
bull Water source or treatment changes that affect the CSMR must be evaluated prior to change
bull Absolute concentrations also important particularly for chloride
31
Manganese
Two distinct pipe scale categories in 5 LSLs (Lead 1-5 Pipes) excavated from same city
31
Schock et al 2014 JAWWAbull The outer pipe scale is in direct contact with the flowing water affecting
water lead contamination the mostbull Plattnerite an insoluble Pb(+4) scale that protects water from lead
contamination was inhibited by Mn (and Fe) presence
Manganese
In this case study (Schock et al JAWWA 2014)bull Where Mn was present in more abundance
- Pb scales had less PbO2- Pb associated with Mn sorbing onto Mn+Fe-rich surface deposit
bull Erratic Pb persisted after full LSLR in areas rich with Mn deposits (and to lesser extent Fe)
bull Galvanized interior plumbing also accumulated dissolved Pb while LSL present
32
33
Manganese
Other studiesbull Dong et al 2000 Mn acted as vector for particulate Pb movementbull Trueman et al 2019 Increasing Mn increased Pb in experimentbull McNeill and Edwards 2004 Mn sequestration by polyphosphate destabilized
Pb scales and caused Pb release
bull Pan et al 2019 Possible benefit by promoting formation of Pb(+4)
Reference Analysis Mn Impact on PbSchock et al 2014 Excavated LSLs
amp historical water Pb data in a US city
bull Promoted amorphous Pb scalebull Inhibited insoluble Pb(+4) scalebull Erratic Pb persisted after full LSLR in areas rich
with Mn deposits
Tully et al 2019 Excavated LSLs amp water Pb data in several Midwest case studies
bull Mn was a component of amorphous pipe scales in several systems
bull Amorphous scales not amenable to determining their adherence to pipe walls their responses to changes in various water chemistry variables or quantitatively calculating their lead release
Itrsquos not just manganese
34
Broad variety of coatings on LSLs analyzed at EPA
Al Mn Fe P and Ca-rich coatings may interfere withorthophosphate pH adjustment or other CCT
35
More water parameters affect metal release
bull pH amp AlkalinityDICbull Concentration and type of
oxidants (disinfectantsdissolved oxygen)
bull ORPcorrosion potentialbull Corrosion Inhibitorsbull Chloridebull Sulfatebull Manganese
bull Temperaturebull Sorptive surfaces downstream of
LSLs (ie galvanized interior pipe)bull Iron (deposition and corrosion)bull Calciumbull Aluminumbull NOM (type amount)bull Mixing from water treatment plants
or sourcesbull Ammoniabull Hydrogen Sulfidebull Microbial activity (nitrification amp
other)
Corrosion Control Treatment is intertwined with all treatments affecting water chemistry
Water Quality Parameters(WQP)
bull pH alkalinity calcium orthophosphate or silica othersbull WQP measurements are important for
bull Process control (measurement frequency needs to be HIGH)bull Keeping treatment from being turned offbull Keeping inhibitor dosages amp pH control from being cut to save $$$bull Keeping required repairs to treatment on tight schedulebull Preventing frequent variations and alternations of water sources
bull Effectiveness of Pb treatment only determinable from tap sampling by combination of different assessment approaches for different lead plumbing sources to which people can be exposed
36
bull OWQP measurements important for process control look at yourdata to understand what they mean and avoid problems
bull OWQP do not predict lead release There are no substitutes fordirectly monitoring lead release
37
M+
M+
M+
M+
Corrosion barrier film hypothesis
Inert barrier film (eg CaCO3)
Pipe Wall
Barrier film would allow tobull Protect ALL pipe materialsbull Monitor corrosion based on one chemical reaction
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
Sodium silicate
bull No systematic studies to look at pH carbonate silicate background chemistry relationships
bull Unclear mechanism pH increase Buffer intensity increase Protective film (thin amp amorphous)
bull Can sequester ironmanganese
bull Canrsquot be evaluated with fresh surfaces
24pH5 6 7 8 9 10 11
β ty
nsi
nte
Iuf
fer
B
00000
00005
00010TIC=48 mg CL20 mgL SiO250 mgL PO4β H2OTotal β
Silicate may contribute to Buffer Intensity
Carbonate orthophosphate silicate I=001 25ordmC
Chemical changes may reduce Pb+4 to Pb+2
10
8
6
4
)ts 2ol
h (v
E
ndash2
ndash4
ndash6
ndash8
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
pH
12
PbO2 (plattnerite)
Pb++ )s
deg
(2
CO
3
(OH
) -2 2) 3
Pb
-2
0 (CO
-- 2
0
) 43(C
O
Pb
3 (OH
)
DIC = 18 mg CL
Pb Pb Pb = 0010 mgL
Pb metal
-10
Drop in ORP from treatment change or DS
oxidant demand
Drop in pH at surface from treatment change
chemical reactions nitrification etc
C D
ngto
nhisa
W
Newark NJ
Disinfectant demand in DS must be controlled amp enough free chlorine consistently maintained throughout LSL area to keep protective Pb+4 corrosion scales in place
Switching disinfectants may reduce Pb+4 to Pb+2
Low
Switch from chlorine to chloramine disinfectant dissolved lead from pipe scales during the Washington DC ldquolead-in-water crisisrdquo in 2001-2004
Example profile of Pb(+4) (ie PbO2) scale house
27
20
L)
Cl1 Cincinnati OH
18
microg 16
n ( 14
o 12
tia 10
trn 8
e 6
nco 4
C 2
b P 0-1 0 1 2 3 4 5 6 7 8 9 10 11
Cumulative Water Volume (L)
CI-1 10 h 1209CI-1 10 h 0410
Plumbing sequence corresponding to water volumeFaucet LSLCu pipe Brass
Water MainMCu pipe 2
015
a
let
dou
illyaf
ant
irT
bull Pb release to water from PbO2 coated LSLs was found to be lowbull Be aware are there are no known cases of exhumed PbO2 coated pipes being able to
re-stabilize in pipe rigs
Chloride and sulfateCSMR a factor controlling galvanic corrosion (Oliphant 1983 Gregory 1986)- Chloride detrimental- Sulfate beneficial if sufficient to overcome chloride thus- CSMR ratio was created
[Cl- ] 12 mgL Cl-Chloride to Sulfate Mass Ratio (CSMR) = =[SO -2 2 = 06
4 ] 20 mgL SO-4
Confirmed empirically in case studies of water utilities (Dodrill and Edwards 1995 Edwards and Triantafyllidou 2007) - High CSMR gt 05-06 more corrosive- Low CSMR lt 05-06 less corrosive
28
Verified in laboratory experiments with fresh galvanic Pb junctions (eg Wang et al 2013 Triantafyllidou et al 2011 Nguyen et al 2010 2011)
Tria
ntaf
yllid
ou e
t al
201
1
Chloride and sulfate
29
CSMR considered contributing factor to the Flint water Pb crisis (Pieper et al 2017)
30
Chloride and sulfate
bull Absolute concentrations of chloride and sulfate important chloride considered particularly detrimental to galvanic Pb corrosion
bull Increasing chloride and sulfate concentrations at a constant CSMR of 1 increased total lead concentrations in galvanic corrosion experiments (Ng amp Lin 2016)
bull High CSMR ratio proven empirically important in aggravating galvanic lead corrosion (verified by experiments and modeling)
bull Water source or treatment changes that affect the CSMR must be evaluated prior to change
bull Absolute concentrations also important particularly for chloride
31
Manganese
Two distinct pipe scale categories in 5 LSLs (Lead 1-5 Pipes) excavated from same city
31
Schock et al 2014 JAWWAbull The outer pipe scale is in direct contact with the flowing water affecting
water lead contamination the mostbull Plattnerite an insoluble Pb(+4) scale that protects water from lead
contamination was inhibited by Mn (and Fe) presence
Manganese
In this case study (Schock et al JAWWA 2014)bull Where Mn was present in more abundance
- Pb scales had less PbO2- Pb associated with Mn sorbing onto Mn+Fe-rich surface deposit
bull Erratic Pb persisted after full LSLR in areas rich with Mn deposits (and to lesser extent Fe)
bull Galvanized interior plumbing also accumulated dissolved Pb while LSL present
32
33
Manganese
Other studiesbull Dong et al 2000 Mn acted as vector for particulate Pb movementbull Trueman et al 2019 Increasing Mn increased Pb in experimentbull McNeill and Edwards 2004 Mn sequestration by polyphosphate destabilized
Pb scales and caused Pb release
bull Pan et al 2019 Possible benefit by promoting formation of Pb(+4)
Reference Analysis Mn Impact on PbSchock et al 2014 Excavated LSLs
amp historical water Pb data in a US city
bull Promoted amorphous Pb scalebull Inhibited insoluble Pb(+4) scalebull Erratic Pb persisted after full LSLR in areas rich
with Mn deposits
Tully et al 2019 Excavated LSLs amp water Pb data in several Midwest case studies
bull Mn was a component of amorphous pipe scales in several systems
bull Amorphous scales not amenable to determining their adherence to pipe walls their responses to changes in various water chemistry variables or quantitatively calculating their lead release
Itrsquos not just manganese
34
Broad variety of coatings on LSLs analyzed at EPA
Al Mn Fe P and Ca-rich coatings may interfere withorthophosphate pH adjustment or other CCT
35
More water parameters affect metal release
bull pH amp AlkalinityDICbull Concentration and type of
oxidants (disinfectantsdissolved oxygen)
bull ORPcorrosion potentialbull Corrosion Inhibitorsbull Chloridebull Sulfatebull Manganese
bull Temperaturebull Sorptive surfaces downstream of
LSLs (ie galvanized interior pipe)bull Iron (deposition and corrosion)bull Calciumbull Aluminumbull NOM (type amount)bull Mixing from water treatment plants
or sourcesbull Ammoniabull Hydrogen Sulfidebull Microbial activity (nitrification amp
other)
Corrosion Control Treatment is intertwined with all treatments affecting water chemistry
Water Quality Parameters(WQP)
bull pH alkalinity calcium orthophosphate or silica othersbull WQP measurements are important for
bull Process control (measurement frequency needs to be HIGH)bull Keeping treatment from being turned offbull Keeping inhibitor dosages amp pH control from being cut to save $$$bull Keeping required repairs to treatment on tight schedulebull Preventing frequent variations and alternations of water sources
bull Effectiveness of Pb treatment only determinable from tap sampling by combination of different assessment approaches for different lead plumbing sources to which people can be exposed
36
bull OWQP measurements important for process control look at yourdata to understand what they mean and avoid problems
bull OWQP do not predict lead release There are no substitutes fordirectly monitoring lead release
37
M+
M+
M+
M+
Corrosion barrier film hypothesis
Inert barrier film (eg CaCO3)
Pipe Wall
Barrier film would allow tobull Protect ALL pipe materialsbull Monitor corrosion based on one chemical reaction
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
Chemical changes may reduce Pb+4 to Pb+2
10
8
6
4
)ts 2ol
h (v
E
ndash2
ndash4
ndash6
ndash8
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
pH
12
PbO2 (plattnerite)
Pb++ )s
deg
(2
CO
3
(OH
) -2 2) 3
Pb
-2
0 (CO
-- 2
0
) 43(C
O
Pb
3 (OH
)
DIC = 18 mg CL
Pb Pb Pb = 0010 mgL
Pb metal
-10
Drop in ORP from treatment change or DS
oxidant demand
Drop in pH at surface from treatment change
chemical reactions nitrification etc
C D
ngto
nhisa
W
Newark NJ
Disinfectant demand in DS must be controlled amp enough free chlorine consistently maintained throughout LSL area to keep protective Pb+4 corrosion scales in place
Switching disinfectants may reduce Pb+4 to Pb+2
Low
Switch from chlorine to chloramine disinfectant dissolved lead from pipe scales during the Washington DC ldquolead-in-water crisisrdquo in 2001-2004
Example profile of Pb(+4) (ie PbO2) scale house
27
20
L)
Cl1 Cincinnati OH
18
microg 16
n ( 14
o 12
tia 10
trn 8
e 6
nco 4
C 2
b P 0-1 0 1 2 3 4 5 6 7 8 9 10 11
Cumulative Water Volume (L)
CI-1 10 h 1209CI-1 10 h 0410
Plumbing sequence corresponding to water volumeFaucet LSLCu pipe Brass
Water MainMCu pipe 2
015
a
let
dou
illyaf
ant
irT
bull Pb release to water from PbO2 coated LSLs was found to be lowbull Be aware are there are no known cases of exhumed PbO2 coated pipes being able to
re-stabilize in pipe rigs
Chloride and sulfateCSMR a factor controlling galvanic corrosion (Oliphant 1983 Gregory 1986)- Chloride detrimental- Sulfate beneficial if sufficient to overcome chloride thus- CSMR ratio was created
[Cl- ] 12 mgL Cl-Chloride to Sulfate Mass Ratio (CSMR) = =[SO -2 2 = 06
4 ] 20 mgL SO-4
Confirmed empirically in case studies of water utilities (Dodrill and Edwards 1995 Edwards and Triantafyllidou 2007) - High CSMR gt 05-06 more corrosive- Low CSMR lt 05-06 less corrosive
28
Verified in laboratory experiments with fresh galvanic Pb junctions (eg Wang et al 2013 Triantafyllidou et al 2011 Nguyen et al 2010 2011)
Tria
ntaf
yllid
ou e
t al
201
1
Chloride and sulfate
29
CSMR considered contributing factor to the Flint water Pb crisis (Pieper et al 2017)
30
Chloride and sulfate
bull Absolute concentrations of chloride and sulfate important chloride considered particularly detrimental to galvanic Pb corrosion
bull Increasing chloride and sulfate concentrations at a constant CSMR of 1 increased total lead concentrations in galvanic corrosion experiments (Ng amp Lin 2016)
bull High CSMR ratio proven empirically important in aggravating galvanic lead corrosion (verified by experiments and modeling)
bull Water source or treatment changes that affect the CSMR must be evaluated prior to change
bull Absolute concentrations also important particularly for chloride
31
Manganese
Two distinct pipe scale categories in 5 LSLs (Lead 1-5 Pipes) excavated from same city
31
Schock et al 2014 JAWWAbull The outer pipe scale is in direct contact with the flowing water affecting
water lead contamination the mostbull Plattnerite an insoluble Pb(+4) scale that protects water from lead
contamination was inhibited by Mn (and Fe) presence
Manganese
In this case study (Schock et al JAWWA 2014)bull Where Mn was present in more abundance
- Pb scales had less PbO2- Pb associated with Mn sorbing onto Mn+Fe-rich surface deposit
bull Erratic Pb persisted after full LSLR in areas rich with Mn deposits (and to lesser extent Fe)
bull Galvanized interior plumbing also accumulated dissolved Pb while LSL present
32
33
Manganese
Other studiesbull Dong et al 2000 Mn acted as vector for particulate Pb movementbull Trueman et al 2019 Increasing Mn increased Pb in experimentbull McNeill and Edwards 2004 Mn sequestration by polyphosphate destabilized
Pb scales and caused Pb release
bull Pan et al 2019 Possible benefit by promoting formation of Pb(+4)
Reference Analysis Mn Impact on PbSchock et al 2014 Excavated LSLs
amp historical water Pb data in a US city
bull Promoted amorphous Pb scalebull Inhibited insoluble Pb(+4) scalebull Erratic Pb persisted after full LSLR in areas rich
with Mn deposits
Tully et al 2019 Excavated LSLs amp water Pb data in several Midwest case studies
bull Mn was a component of amorphous pipe scales in several systems
bull Amorphous scales not amenable to determining their adherence to pipe walls their responses to changes in various water chemistry variables or quantitatively calculating their lead release
Itrsquos not just manganese
34
Broad variety of coatings on LSLs analyzed at EPA
Al Mn Fe P and Ca-rich coatings may interfere withorthophosphate pH adjustment or other CCT
35
More water parameters affect metal release
bull pH amp AlkalinityDICbull Concentration and type of
oxidants (disinfectantsdissolved oxygen)
bull ORPcorrosion potentialbull Corrosion Inhibitorsbull Chloridebull Sulfatebull Manganese
bull Temperaturebull Sorptive surfaces downstream of
LSLs (ie galvanized interior pipe)bull Iron (deposition and corrosion)bull Calciumbull Aluminumbull NOM (type amount)bull Mixing from water treatment plants
or sourcesbull Ammoniabull Hydrogen Sulfidebull Microbial activity (nitrification amp
other)
Corrosion Control Treatment is intertwined with all treatments affecting water chemistry
Water Quality Parameters(WQP)
bull pH alkalinity calcium orthophosphate or silica othersbull WQP measurements are important for
bull Process control (measurement frequency needs to be HIGH)bull Keeping treatment from being turned offbull Keeping inhibitor dosages amp pH control from being cut to save $$$bull Keeping required repairs to treatment on tight schedulebull Preventing frequent variations and alternations of water sources
bull Effectiveness of Pb treatment only determinable from tap sampling by combination of different assessment approaches for different lead plumbing sources to which people can be exposed
36
bull OWQP measurements important for process control look at yourdata to understand what they mean and avoid problems
bull OWQP do not predict lead release There are no substitutes fordirectly monitoring lead release
37
M+
M+
M+
M+
Corrosion barrier film hypothesis
Inert barrier film (eg CaCO3)
Pipe Wall
Barrier film would allow tobull Protect ALL pipe materialsbull Monitor corrosion based on one chemical reaction
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
Switching disinfectants may reduce Pb+4 to Pb+2
Low
Switch from chlorine to chloramine disinfectant dissolved lead from pipe scales during the Washington DC ldquolead-in-water crisisrdquo in 2001-2004
Example profile of Pb(+4) (ie PbO2) scale house
27
20
L)
Cl1 Cincinnati OH
18
microg 16
n ( 14
o 12
tia 10
trn 8
e 6
nco 4
C 2
b P 0-1 0 1 2 3 4 5 6 7 8 9 10 11
Cumulative Water Volume (L)
CI-1 10 h 1209CI-1 10 h 0410
Plumbing sequence corresponding to water volumeFaucet LSLCu pipe Brass
Water MainMCu pipe 2
015
a
let
dou
illyaf
ant
irT
bull Pb release to water from PbO2 coated LSLs was found to be lowbull Be aware are there are no known cases of exhumed PbO2 coated pipes being able to
re-stabilize in pipe rigs
Chloride and sulfateCSMR a factor controlling galvanic corrosion (Oliphant 1983 Gregory 1986)- Chloride detrimental- Sulfate beneficial if sufficient to overcome chloride thus- CSMR ratio was created
[Cl- ] 12 mgL Cl-Chloride to Sulfate Mass Ratio (CSMR) = =[SO -2 2 = 06
4 ] 20 mgL SO-4
Confirmed empirically in case studies of water utilities (Dodrill and Edwards 1995 Edwards and Triantafyllidou 2007) - High CSMR gt 05-06 more corrosive- Low CSMR lt 05-06 less corrosive
28
Verified in laboratory experiments with fresh galvanic Pb junctions (eg Wang et al 2013 Triantafyllidou et al 2011 Nguyen et al 2010 2011)
Tria
ntaf
yllid
ou e
t al
201
1
Chloride and sulfate
29
CSMR considered contributing factor to the Flint water Pb crisis (Pieper et al 2017)
30
Chloride and sulfate
bull Absolute concentrations of chloride and sulfate important chloride considered particularly detrimental to galvanic Pb corrosion
bull Increasing chloride and sulfate concentrations at a constant CSMR of 1 increased total lead concentrations in galvanic corrosion experiments (Ng amp Lin 2016)
bull High CSMR ratio proven empirically important in aggravating galvanic lead corrosion (verified by experiments and modeling)
bull Water source or treatment changes that affect the CSMR must be evaluated prior to change
bull Absolute concentrations also important particularly for chloride
31
Manganese
Two distinct pipe scale categories in 5 LSLs (Lead 1-5 Pipes) excavated from same city
31
Schock et al 2014 JAWWAbull The outer pipe scale is in direct contact with the flowing water affecting
water lead contamination the mostbull Plattnerite an insoluble Pb(+4) scale that protects water from lead
contamination was inhibited by Mn (and Fe) presence
Manganese
In this case study (Schock et al JAWWA 2014)bull Where Mn was present in more abundance
- Pb scales had less PbO2- Pb associated with Mn sorbing onto Mn+Fe-rich surface deposit
bull Erratic Pb persisted after full LSLR in areas rich with Mn deposits (and to lesser extent Fe)
bull Galvanized interior plumbing also accumulated dissolved Pb while LSL present
32
33
Manganese
Other studiesbull Dong et al 2000 Mn acted as vector for particulate Pb movementbull Trueman et al 2019 Increasing Mn increased Pb in experimentbull McNeill and Edwards 2004 Mn sequestration by polyphosphate destabilized
Pb scales and caused Pb release
bull Pan et al 2019 Possible benefit by promoting formation of Pb(+4)
Reference Analysis Mn Impact on PbSchock et al 2014 Excavated LSLs
amp historical water Pb data in a US city
bull Promoted amorphous Pb scalebull Inhibited insoluble Pb(+4) scalebull Erratic Pb persisted after full LSLR in areas rich
with Mn deposits
Tully et al 2019 Excavated LSLs amp water Pb data in several Midwest case studies
bull Mn was a component of amorphous pipe scales in several systems
bull Amorphous scales not amenable to determining their adherence to pipe walls their responses to changes in various water chemistry variables or quantitatively calculating their lead release
Itrsquos not just manganese
34
Broad variety of coatings on LSLs analyzed at EPA
Al Mn Fe P and Ca-rich coatings may interfere withorthophosphate pH adjustment or other CCT
35
More water parameters affect metal release
bull pH amp AlkalinityDICbull Concentration and type of
oxidants (disinfectantsdissolved oxygen)
bull ORPcorrosion potentialbull Corrosion Inhibitorsbull Chloridebull Sulfatebull Manganese
bull Temperaturebull Sorptive surfaces downstream of
LSLs (ie galvanized interior pipe)bull Iron (deposition and corrosion)bull Calciumbull Aluminumbull NOM (type amount)bull Mixing from water treatment plants
or sourcesbull Ammoniabull Hydrogen Sulfidebull Microbial activity (nitrification amp
other)
Corrosion Control Treatment is intertwined with all treatments affecting water chemistry
Water Quality Parameters(WQP)
bull pH alkalinity calcium orthophosphate or silica othersbull WQP measurements are important for
bull Process control (measurement frequency needs to be HIGH)bull Keeping treatment from being turned offbull Keeping inhibitor dosages amp pH control from being cut to save $$$bull Keeping required repairs to treatment on tight schedulebull Preventing frequent variations and alternations of water sources
bull Effectiveness of Pb treatment only determinable from tap sampling by combination of different assessment approaches for different lead plumbing sources to which people can be exposed
36
bull OWQP measurements important for process control look at yourdata to understand what they mean and avoid problems
bull OWQP do not predict lead release There are no substitutes fordirectly monitoring lead release
37
M+
M+
M+
M+
Corrosion barrier film hypothesis
Inert barrier film (eg CaCO3)
Pipe Wall
Barrier film would allow tobull Protect ALL pipe materialsbull Monitor corrosion based on one chemical reaction
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
Example profile of Pb(+4) (ie PbO2) scale house
27
20
L)
Cl1 Cincinnati OH
18
microg 16
n ( 14
o 12
tia 10
trn 8
e 6
nco 4
C 2
b P 0-1 0 1 2 3 4 5 6 7 8 9 10 11
Cumulative Water Volume (L)
CI-1 10 h 1209CI-1 10 h 0410
Plumbing sequence corresponding to water volumeFaucet LSLCu pipe Brass
Water MainMCu pipe 2
015
a
let
dou
illyaf
ant
irT
bull Pb release to water from PbO2 coated LSLs was found to be lowbull Be aware are there are no known cases of exhumed PbO2 coated pipes being able to
re-stabilize in pipe rigs
Chloride and sulfateCSMR a factor controlling galvanic corrosion (Oliphant 1983 Gregory 1986)- Chloride detrimental- Sulfate beneficial if sufficient to overcome chloride thus- CSMR ratio was created
[Cl- ] 12 mgL Cl-Chloride to Sulfate Mass Ratio (CSMR) = =[SO -2 2 = 06
4 ] 20 mgL SO-4
Confirmed empirically in case studies of water utilities (Dodrill and Edwards 1995 Edwards and Triantafyllidou 2007) - High CSMR gt 05-06 more corrosive- Low CSMR lt 05-06 less corrosive
28
Verified in laboratory experiments with fresh galvanic Pb junctions (eg Wang et al 2013 Triantafyllidou et al 2011 Nguyen et al 2010 2011)
Tria
ntaf
yllid
ou e
t al
201
1
Chloride and sulfate
29
CSMR considered contributing factor to the Flint water Pb crisis (Pieper et al 2017)
30
Chloride and sulfate
bull Absolute concentrations of chloride and sulfate important chloride considered particularly detrimental to galvanic Pb corrosion
bull Increasing chloride and sulfate concentrations at a constant CSMR of 1 increased total lead concentrations in galvanic corrosion experiments (Ng amp Lin 2016)
bull High CSMR ratio proven empirically important in aggravating galvanic lead corrosion (verified by experiments and modeling)
bull Water source or treatment changes that affect the CSMR must be evaluated prior to change
bull Absolute concentrations also important particularly for chloride
31
Manganese
Two distinct pipe scale categories in 5 LSLs (Lead 1-5 Pipes) excavated from same city
31
Schock et al 2014 JAWWAbull The outer pipe scale is in direct contact with the flowing water affecting
water lead contamination the mostbull Plattnerite an insoluble Pb(+4) scale that protects water from lead
contamination was inhibited by Mn (and Fe) presence
Manganese
In this case study (Schock et al JAWWA 2014)bull Where Mn was present in more abundance
- Pb scales had less PbO2- Pb associated with Mn sorbing onto Mn+Fe-rich surface deposit
bull Erratic Pb persisted after full LSLR in areas rich with Mn deposits (and to lesser extent Fe)
bull Galvanized interior plumbing also accumulated dissolved Pb while LSL present
32
33
Manganese
Other studiesbull Dong et al 2000 Mn acted as vector for particulate Pb movementbull Trueman et al 2019 Increasing Mn increased Pb in experimentbull McNeill and Edwards 2004 Mn sequestration by polyphosphate destabilized
Pb scales and caused Pb release
bull Pan et al 2019 Possible benefit by promoting formation of Pb(+4)
Reference Analysis Mn Impact on PbSchock et al 2014 Excavated LSLs
amp historical water Pb data in a US city
bull Promoted amorphous Pb scalebull Inhibited insoluble Pb(+4) scalebull Erratic Pb persisted after full LSLR in areas rich
with Mn deposits
Tully et al 2019 Excavated LSLs amp water Pb data in several Midwest case studies
bull Mn was a component of amorphous pipe scales in several systems
bull Amorphous scales not amenable to determining their adherence to pipe walls their responses to changes in various water chemistry variables or quantitatively calculating their lead release
Itrsquos not just manganese
34
Broad variety of coatings on LSLs analyzed at EPA
Al Mn Fe P and Ca-rich coatings may interfere withorthophosphate pH adjustment or other CCT
35
More water parameters affect metal release
bull pH amp AlkalinityDICbull Concentration and type of
oxidants (disinfectantsdissolved oxygen)
bull ORPcorrosion potentialbull Corrosion Inhibitorsbull Chloridebull Sulfatebull Manganese
bull Temperaturebull Sorptive surfaces downstream of
LSLs (ie galvanized interior pipe)bull Iron (deposition and corrosion)bull Calciumbull Aluminumbull NOM (type amount)bull Mixing from water treatment plants
or sourcesbull Ammoniabull Hydrogen Sulfidebull Microbial activity (nitrification amp
other)
Corrosion Control Treatment is intertwined with all treatments affecting water chemistry
Water Quality Parameters(WQP)
bull pH alkalinity calcium orthophosphate or silica othersbull WQP measurements are important for
bull Process control (measurement frequency needs to be HIGH)bull Keeping treatment from being turned offbull Keeping inhibitor dosages amp pH control from being cut to save $$$bull Keeping required repairs to treatment on tight schedulebull Preventing frequent variations and alternations of water sources
bull Effectiveness of Pb treatment only determinable from tap sampling by combination of different assessment approaches for different lead plumbing sources to which people can be exposed
36
bull OWQP measurements important for process control look at yourdata to understand what they mean and avoid problems
bull OWQP do not predict lead release There are no substitutes fordirectly monitoring lead release
37
M+
M+
M+
M+
Corrosion barrier film hypothesis
Inert barrier film (eg CaCO3)
Pipe Wall
Barrier film would allow tobull Protect ALL pipe materialsbull Monitor corrosion based on one chemical reaction
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
Chloride and sulfateCSMR a factor controlling galvanic corrosion (Oliphant 1983 Gregory 1986)- Chloride detrimental- Sulfate beneficial if sufficient to overcome chloride thus- CSMR ratio was created
[Cl- ] 12 mgL Cl-Chloride to Sulfate Mass Ratio (CSMR) = =[SO -2 2 = 06
4 ] 20 mgL SO-4
Confirmed empirically in case studies of water utilities (Dodrill and Edwards 1995 Edwards and Triantafyllidou 2007) - High CSMR gt 05-06 more corrosive- Low CSMR lt 05-06 less corrosive
28
Verified in laboratory experiments with fresh galvanic Pb junctions (eg Wang et al 2013 Triantafyllidou et al 2011 Nguyen et al 2010 2011)
Tria
ntaf
yllid
ou e
t al
201
1
Chloride and sulfate
29
CSMR considered contributing factor to the Flint water Pb crisis (Pieper et al 2017)
30
Chloride and sulfate
bull Absolute concentrations of chloride and sulfate important chloride considered particularly detrimental to galvanic Pb corrosion
bull Increasing chloride and sulfate concentrations at a constant CSMR of 1 increased total lead concentrations in galvanic corrosion experiments (Ng amp Lin 2016)
bull High CSMR ratio proven empirically important in aggravating galvanic lead corrosion (verified by experiments and modeling)
bull Water source or treatment changes that affect the CSMR must be evaluated prior to change
bull Absolute concentrations also important particularly for chloride
31
Manganese
Two distinct pipe scale categories in 5 LSLs (Lead 1-5 Pipes) excavated from same city
31
Schock et al 2014 JAWWAbull The outer pipe scale is in direct contact with the flowing water affecting
water lead contamination the mostbull Plattnerite an insoluble Pb(+4) scale that protects water from lead
contamination was inhibited by Mn (and Fe) presence
Manganese
In this case study (Schock et al JAWWA 2014)bull Where Mn was present in more abundance
- Pb scales had less PbO2- Pb associated with Mn sorbing onto Mn+Fe-rich surface deposit
bull Erratic Pb persisted after full LSLR in areas rich with Mn deposits (and to lesser extent Fe)
bull Galvanized interior plumbing also accumulated dissolved Pb while LSL present
32
33
Manganese
Other studiesbull Dong et al 2000 Mn acted as vector for particulate Pb movementbull Trueman et al 2019 Increasing Mn increased Pb in experimentbull McNeill and Edwards 2004 Mn sequestration by polyphosphate destabilized
Pb scales and caused Pb release
bull Pan et al 2019 Possible benefit by promoting formation of Pb(+4)
Reference Analysis Mn Impact on PbSchock et al 2014 Excavated LSLs
amp historical water Pb data in a US city
bull Promoted amorphous Pb scalebull Inhibited insoluble Pb(+4) scalebull Erratic Pb persisted after full LSLR in areas rich
with Mn deposits
Tully et al 2019 Excavated LSLs amp water Pb data in several Midwest case studies
bull Mn was a component of amorphous pipe scales in several systems
bull Amorphous scales not amenable to determining their adherence to pipe walls their responses to changes in various water chemistry variables or quantitatively calculating their lead release
Itrsquos not just manganese
34
Broad variety of coatings on LSLs analyzed at EPA
Al Mn Fe P and Ca-rich coatings may interfere withorthophosphate pH adjustment or other CCT
35
More water parameters affect metal release
bull pH amp AlkalinityDICbull Concentration and type of
oxidants (disinfectantsdissolved oxygen)
bull ORPcorrosion potentialbull Corrosion Inhibitorsbull Chloridebull Sulfatebull Manganese
bull Temperaturebull Sorptive surfaces downstream of
LSLs (ie galvanized interior pipe)bull Iron (deposition and corrosion)bull Calciumbull Aluminumbull NOM (type amount)bull Mixing from water treatment plants
or sourcesbull Ammoniabull Hydrogen Sulfidebull Microbial activity (nitrification amp
other)
Corrosion Control Treatment is intertwined with all treatments affecting water chemistry
Water Quality Parameters(WQP)
bull pH alkalinity calcium orthophosphate or silica othersbull WQP measurements are important for
bull Process control (measurement frequency needs to be HIGH)bull Keeping treatment from being turned offbull Keeping inhibitor dosages amp pH control from being cut to save $$$bull Keeping required repairs to treatment on tight schedulebull Preventing frequent variations and alternations of water sources
bull Effectiveness of Pb treatment only determinable from tap sampling by combination of different assessment approaches for different lead plumbing sources to which people can be exposed
36
bull OWQP measurements important for process control look at yourdata to understand what they mean and avoid problems
bull OWQP do not predict lead release There are no substitutes fordirectly monitoring lead release
37
M+
M+
M+
M+
Corrosion barrier film hypothesis
Inert barrier film (eg CaCO3)
Pipe Wall
Barrier film would allow tobull Protect ALL pipe materialsbull Monitor corrosion based on one chemical reaction
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
Chloride and sulfate
29
CSMR considered contributing factor to the Flint water Pb crisis (Pieper et al 2017)
30
Chloride and sulfate
bull Absolute concentrations of chloride and sulfate important chloride considered particularly detrimental to galvanic Pb corrosion
bull Increasing chloride and sulfate concentrations at a constant CSMR of 1 increased total lead concentrations in galvanic corrosion experiments (Ng amp Lin 2016)
bull High CSMR ratio proven empirically important in aggravating galvanic lead corrosion (verified by experiments and modeling)
bull Water source or treatment changes that affect the CSMR must be evaluated prior to change
bull Absolute concentrations also important particularly for chloride
31
Manganese
Two distinct pipe scale categories in 5 LSLs (Lead 1-5 Pipes) excavated from same city
31
Schock et al 2014 JAWWAbull The outer pipe scale is in direct contact with the flowing water affecting
water lead contamination the mostbull Plattnerite an insoluble Pb(+4) scale that protects water from lead
contamination was inhibited by Mn (and Fe) presence
Manganese
In this case study (Schock et al JAWWA 2014)bull Where Mn was present in more abundance
- Pb scales had less PbO2- Pb associated with Mn sorbing onto Mn+Fe-rich surface deposit
bull Erratic Pb persisted after full LSLR in areas rich with Mn deposits (and to lesser extent Fe)
bull Galvanized interior plumbing also accumulated dissolved Pb while LSL present
32
33
Manganese
Other studiesbull Dong et al 2000 Mn acted as vector for particulate Pb movementbull Trueman et al 2019 Increasing Mn increased Pb in experimentbull McNeill and Edwards 2004 Mn sequestration by polyphosphate destabilized
Pb scales and caused Pb release
bull Pan et al 2019 Possible benefit by promoting formation of Pb(+4)
Reference Analysis Mn Impact on PbSchock et al 2014 Excavated LSLs
amp historical water Pb data in a US city
bull Promoted amorphous Pb scalebull Inhibited insoluble Pb(+4) scalebull Erratic Pb persisted after full LSLR in areas rich
with Mn deposits
Tully et al 2019 Excavated LSLs amp water Pb data in several Midwest case studies
bull Mn was a component of amorphous pipe scales in several systems
bull Amorphous scales not amenable to determining their adherence to pipe walls their responses to changes in various water chemistry variables or quantitatively calculating their lead release
Itrsquos not just manganese
34
Broad variety of coatings on LSLs analyzed at EPA
Al Mn Fe P and Ca-rich coatings may interfere withorthophosphate pH adjustment or other CCT
35
More water parameters affect metal release
bull pH amp AlkalinityDICbull Concentration and type of
oxidants (disinfectantsdissolved oxygen)
bull ORPcorrosion potentialbull Corrosion Inhibitorsbull Chloridebull Sulfatebull Manganese
bull Temperaturebull Sorptive surfaces downstream of
LSLs (ie galvanized interior pipe)bull Iron (deposition and corrosion)bull Calciumbull Aluminumbull NOM (type amount)bull Mixing from water treatment plants
or sourcesbull Ammoniabull Hydrogen Sulfidebull Microbial activity (nitrification amp
other)
Corrosion Control Treatment is intertwined with all treatments affecting water chemistry
Water Quality Parameters(WQP)
bull pH alkalinity calcium orthophosphate or silica othersbull WQP measurements are important for
bull Process control (measurement frequency needs to be HIGH)bull Keeping treatment from being turned offbull Keeping inhibitor dosages amp pH control from being cut to save $$$bull Keeping required repairs to treatment on tight schedulebull Preventing frequent variations and alternations of water sources
bull Effectiveness of Pb treatment only determinable from tap sampling by combination of different assessment approaches for different lead plumbing sources to which people can be exposed
36
bull OWQP measurements important for process control look at yourdata to understand what they mean and avoid problems
bull OWQP do not predict lead release There are no substitutes fordirectly monitoring lead release
37
M+
M+
M+
M+
Corrosion barrier film hypothesis
Inert barrier film (eg CaCO3)
Pipe Wall
Barrier film would allow tobull Protect ALL pipe materialsbull Monitor corrosion based on one chemical reaction
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
30
Chloride and sulfate
bull Absolute concentrations of chloride and sulfate important chloride considered particularly detrimental to galvanic Pb corrosion
bull Increasing chloride and sulfate concentrations at a constant CSMR of 1 increased total lead concentrations in galvanic corrosion experiments (Ng amp Lin 2016)
bull High CSMR ratio proven empirically important in aggravating galvanic lead corrosion (verified by experiments and modeling)
bull Water source or treatment changes that affect the CSMR must be evaluated prior to change
bull Absolute concentrations also important particularly for chloride
31
Manganese
Two distinct pipe scale categories in 5 LSLs (Lead 1-5 Pipes) excavated from same city
31
Schock et al 2014 JAWWAbull The outer pipe scale is in direct contact with the flowing water affecting
water lead contamination the mostbull Plattnerite an insoluble Pb(+4) scale that protects water from lead
contamination was inhibited by Mn (and Fe) presence
Manganese
In this case study (Schock et al JAWWA 2014)bull Where Mn was present in more abundance
- Pb scales had less PbO2- Pb associated with Mn sorbing onto Mn+Fe-rich surface deposit
bull Erratic Pb persisted after full LSLR in areas rich with Mn deposits (and to lesser extent Fe)
bull Galvanized interior plumbing also accumulated dissolved Pb while LSL present
32
33
Manganese
Other studiesbull Dong et al 2000 Mn acted as vector for particulate Pb movementbull Trueman et al 2019 Increasing Mn increased Pb in experimentbull McNeill and Edwards 2004 Mn sequestration by polyphosphate destabilized
Pb scales and caused Pb release
bull Pan et al 2019 Possible benefit by promoting formation of Pb(+4)
Reference Analysis Mn Impact on PbSchock et al 2014 Excavated LSLs
amp historical water Pb data in a US city
bull Promoted amorphous Pb scalebull Inhibited insoluble Pb(+4) scalebull Erratic Pb persisted after full LSLR in areas rich
with Mn deposits
Tully et al 2019 Excavated LSLs amp water Pb data in several Midwest case studies
bull Mn was a component of amorphous pipe scales in several systems
bull Amorphous scales not amenable to determining their adherence to pipe walls their responses to changes in various water chemistry variables or quantitatively calculating their lead release
Itrsquos not just manganese
34
Broad variety of coatings on LSLs analyzed at EPA
Al Mn Fe P and Ca-rich coatings may interfere withorthophosphate pH adjustment or other CCT
35
More water parameters affect metal release
bull pH amp AlkalinityDICbull Concentration and type of
oxidants (disinfectantsdissolved oxygen)
bull ORPcorrosion potentialbull Corrosion Inhibitorsbull Chloridebull Sulfatebull Manganese
bull Temperaturebull Sorptive surfaces downstream of
LSLs (ie galvanized interior pipe)bull Iron (deposition and corrosion)bull Calciumbull Aluminumbull NOM (type amount)bull Mixing from water treatment plants
or sourcesbull Ammoniabull Hydrogen Sulfidebull Microbial activity (nitrification amp
other)
Corrosion Control Treatment is intertwined with all treatments affecting water chemistry
Water Quality Parameters(WQP)
bull pH alkalinity calcium orthophosphate or silica othersbull WQP measurements are important for
bull Process control (measurement frequency needs to be HIGH)bull Keeping treatment from being turned offbull Keeping inhibitor dosages amp pH control from being cut to save $$$bull Keeping required repairs to treatment on tight schedulebull Preventing frequent variations and alternations of water sources
bull Effectiveness of Pb treatment only determinable from tap sampling by combination of different assessment approaches for different lead plumbing sources to which people can be exposed
36
bull OWQP measurements important for process control look at yourdata to understand what they mean and avoid problems
bull OWQP do not predict lead release There are no substitutes fordirectly monitoring lead release
37
M+
M+
M+
M+
Corrosion barrier film hypothesis
Inert barrier film (eg CaCO3)
Pipe Wall
Barrier film would allow tobull Protect ALL pipe materialsbull Monitor corrosion based on one chemical reaction
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
31
Manganese
Two distinct pipe scale categories in 5 LSLs (Lead 1-5 Pipes) excavated from same city
31
Schock et al 2014 JAWWAbull The outer pipe scale is in direct contact with the flowing water affecting
water lead contamination the mostbull Plattnerite an insoluble Pb(+4) scale that protects water from lead
contamination was inhibited by Mn (and Fe) presence
Manganese
In this case study (Schock et al JAWWA 2014)bull Where Mn was present in more abundance
- Pb scales had less PbO2- Pb associated with Mn sorbing onto Mn+Fe-rich surface deposit
bull Erratic Pb persisted after full LSLR in areas rich with Mn deposits (and to lesser extent Fe)
bull Galvanized interior plumbing also accumulated dissolved Pb while LSL present
32
33
Manganese
Other studiesbull Dong et al 2000 Mn acted as vector for particulate Pb movementbull Trueman et al 2019 Increasing Mn increased Pb in experimentbull McNeill and Edwards 2004 Mn sequestration by polyphosphate destabilized
Pb scales and caused Pb release
bull Pan et al 2019 Possible benefit by promoting formation of Pb(+4)
Reference Analysis Mn Impact on PbSchock et al 2014 Excavated LSLs
amp historical water Pb data in a US city
bull Promoted amorphous Pb scalebull Inhibited insoluble Pb(+4) scalebull Erratic Pb persisted after full LSLR in areas rich
with Mn deposits
Tully et al 2019 Excavated LSLs amp water Pb data in several Midwest case studies
bull Mn was a component of amorphous pipe scales in several systems
bull Amorphous scales not amenable to determining their adherence to pipe walls their responses to changes in various water chemistry variables or quantitatively calculating their lead release
Itrsquos not just manganese
34
Broad variety of coatings on LSLs analyzed at EPA
Al Mn Fe P and Ca-rich coatings may interfere withorthophosphate pH adjustment or other CCT
35
More water parameters affect metal release
bull pH amp AlkalinityDICbull Concentration and type of
oxidants (disinfectantsdissolved oxygen)
bull ORPcorrosion potentialbull Corrosion Inhibitorsbull Chloridebull Sulfatebull Manganese
bull Temperaturebull Sorptive surfaces downstream of
LSLs (ie galvanized interior pipe)bull Iron (deposition and corrosion)bull Calciumbull Aluminumbull NOM (type amount)bull Mixing from water treatment plants
or sourcesbull Ammoniabull Hydrogen Sulfidebull Microbial activity (nitrification amp
other)
Corrosion Control Treatment is intertwined with all treatments affecting water chemistry
Water Quality Parameters(WQP)
bull pH alkalinity calcium orthophosphate or silica othersbull WQP measurements are important for
bull Process control (measurement frequency needs to be HIGH)bull Keeping treatment from being turned offbull Keeping inhibitor dosages amp pH control from being cut to save $$$bull Keeping required repairs to treatment on tight schedulebull Preventing frequent variations and alternations of water sources
bull Effectiveness of Pb treatment only determinable from tap sampling by combination of different assessment approaches for different lead plumbing sources to which people can be exposed
36
bull OWQP measurements important for process control look at yourdata to understand what they mean and avoid problems
bull OWQP do not predict lead release There are no substitutes fordirectly monitoring lead release
37
M+
M+
M+
M+
Corrosion barrier film hypothesis
Inert barrier film (eg CaCO3)
Pipe Wall
Barrier film would allow tobull Protect ALL pipe materialsbull Monitor corrosion based on one chemical reaction
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
Manganese
In this case study (Schock et al JAWWA 2014)bull Where Mn was present in more abundance
- Pb scales had less PbO2- Pb associated with Mn sorbing onto Mn+Fe-rich surface deposit
bull Erratic Pb persisted after full LSLR in areas rich with Mn deposits (and to lesser extent Fe)
bull Galvanized interior plumbing also accumulated dissolved Pb while LSL present
32
33
Manganese
Other studiesbull Dong et al 2000 Mn acted as vector for particulate Pb movementbull Trueman et al 2019 Increasing Mn increased Pb in experimentbull McNeill and Edwards 2004 Mn sequestration by polyphosphate destabilized
Pb scales and caused Pb release
bull Pan et al 2019 Possible benefit by promoting formation of Pb(+4)
Reference Analysis Mn Impact on PbSchock et al 2014 Excavated LSLs
amp historical water Pb data in a US city
bull Promoted amorphous Pb scalebull Inhibited insoluble Pb(+4) scalebull Erratic Pb persisted after full LSLR in areas rich
with Mn deposits
Tully et al 2019 Excavated LSLs amp water Pb data in several Midwest case studies
bull Mn was a component of amorphous pipe scales in several systems
bull Amorphous scales not amenable to determining their adherence to pipe walls their responses to changes in various water chemistry variables or quantitatively calculating their lead release
Itrsquos not just manganese
34
Broad variety of coatings on LSLs analyzed at EPA
Al Mn Fe P and Ca-rich coatings may interfere withorthophosphate pH adjustment or other CCT
35
More water parameters affect metal release
bull pH amp AlkalinityDICbull Concentration and type of
oxidants (disinfectantsdissolved oxygen)
bull ORPcorrosion potentialbull Corrosion Inhibitorsbull Chloridebull Sulfatebull Manganese
bull Temperaturebull Sorptive surfaces downstream of
LSLs (ie galvanized interior pipe)bull Iron (deposition and corrosion)bull Calciumbull Aluminumbull NOM (type amount)bull Mixing from water treatment plants
or sourcesbull Ammoniabull Hydrogen Sulfidebull Microbial activity (nitrification amp
other)
Corrosion Control Treatment is intertwined with all treatments affecting water chemistry
Water Quality Parameters(WQP)
bull pH alkalinity calcium orthophosphate or silica othersbull WQP measurements are important for
bull Process control (measurement frequency needs to be HIGH)bull Keeping treatment from being turned offbull Keeping inhibitor dosages amp pH control from being cut to save $$$bull Keeping required repairs to treatment on tight schedulebull Preventing frequent variations and alternations of water sources
bull Effectiveness of Pb treatment only determinable from tap sampling by combination of different assessment approaches for different lead plumbing sources to which people can be exposed
36
bull OWQP measurements important for process control look at yourdata to understand what they mean and avoid problems
bull OWQP do not predict lead release There are no substitutes fordirectly monitoring lead release
37
M+
M+
M+
M+
Corrosion barrier film hypothesis
Inert barrier film (eg CaCO3)
Pipe Wall
Barrier film would allow tobull Protect ALL pipe materialsbull Monitor corrosion based on one chemical reaction
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
33
Manganese
Other studiesbull Dong et al 2000 Mn acted as vector for particulate Pb movementbull Trueman et al 2019 Increasing Mn increased Pb in experimentbull McNeill and Edwards 2004 Mn sequestration by polyphosphate destabilized
Pb scales and caused Pb release
bull Pan et al 2019 Possible benefit by promoting formation of Pb(+4)
Reference Analysis Mn Impact on PbSchock et al 2014 Excavated LSLs
amp historical water Pb data in a US city
bull Promoted amorphous Pb scalebull Inhibited insoluble Pb(+4) scalebull Erratic Pb persisted after full LSLR in areas rich
with Mn deposits
Tully et al 2019 Excavated LSLs amp water Pb data in several Midwest case studies
bull Mn was a component of amorphous pipe scales in several systems
bull Amorphous scales not amenable to determining their adherence to pipe walls their responses to changes in various water chemistry variables or quantitatively calculating their lead release
Itrsquos not just manganese
34
Broad variety of coatings on LSLs analyzed at EPA
Al Mn Fe P and Ca-rich coatings may interfere withorthophosphate pH adjustment or other CCT
35
More water parameters affect metal release
bull pH amp AlkalinityDICbull Concentration and type of
oxidants (disinfectantsdissolved oxygen)
bull ORPcorrosion potentialbull Corrosion Inhibitorsbull Chloridebull Sulfatebull Manganese
bull Temperaturebull Sorptive surfaces downstream of
LSLs (ie galvanized interior pipe)bull Iron (deposition and corrosion)bull Calciumbull Aluminumbull NOM (type amount)bull Mixing from water treatment plants
or sourcesbull Ammoniabull Hydrogen Sulfidebull Microbial activity (nitrification amp
other)
Corrosion Control Treatment is intertwined with all treatments affecting water chemistry
Water Quality Parameters(WQP)
bull pH alkalinity calcium orthophosphate or silica othersbull WQP measurements are important for
bull Process control (measurement frequency needs to be HIGH)bull Keeping treatment from being turned offbull Keeping inhibitor dosages amp pH control from being cut to save $$$bull Keeping required repairs to treatment on tight schedulebull Preventing frequent variations and alternations of water sources
bull Effectiveness of Pb treatment only determinable from tap sampling by combination of different assessment approaches for different lead plumbing sources to which people can be exposed
36
bull OWQP measurements important for process control look at yourdata to understand what they mean and avoid problems
bull OWQP do not predict lead release There are no substitutes fordirectly monitoring lead release
37
M+
M+
M+
M+
Corrosion barrier film hypothesis
Inert barrier film (eg CaCO3)
Pipe Wall
Barrier film would allow tobull Protect ALL pipe materialsbull Monitor corrosion based on one chemical reaction
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
Itrsquos not just manganese
34
Broad variety of coatings on LSLs analyzed at EPA
Al Mn Fe P and Ca-rich coatings may interfere withorthophosphate pH adjustment or other CCT
35
More water parameters affect metal release
bull pH amp AlkalinityDICbull Concentration and type of
oxidants (disinfectantsdissolved oxygen)
bull ORPcorrosion potentialbull Corrosion Inhibitorsbull Chloridebull Sulfatebull Manganese
bull Temperaturebull Sorptive surfaces downstream of
LSLs (ie galvanized interior pipe)bull Iron (deposition and corrosion)bull Calciumbull Aluminumbull NOM (type amount)bull Mixing from water treatment plants
or sourcesbull Ammoniabull Hydrogen Sulfidebull Microbial activity (nitrification amp
other)
Corrosion Control Treatment is intertwined with all treatments affecting water chemistry
Water Quality Parameters(WQP)
bull pH alkalinity calcium orthophosphate or silica othersbull WQP measurements are important for
bull Process control (measurement frequency needs to be HIGH)bull Keeping treatment from being turned offbull Keeping inhibitor dosages amp pH control from being cut to save $$$bull Keeping required repairs to treatment on tight schedulebull Preventing frequent variations and alternations of water sources
bull Effectiveness of Pb treatment only determinable from tap sampling by combination of different assessment approaches for different lead plumbing sources to which people can be exposed
36
bull OWQP measurements important for process control look at yourdata to understand what they mean and avoid problems
bull OWQP do not predict lead release There are no substitutes fordirectly monitoring lead release
37
M+
M+
M+
M+
Corrosion barrier film hypothesis
Inert barrier film (eg CaCO3)
Pipe Wall
Barrier film would allow tobull Protect ALL pipe materialsbull Monitor corrosion based on one chemical reaction
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
35
More water parameters affect metal release
bull pH amp AlkalinityDICbull Concentration and type of
oxidants (disinfectantsdissolved oxygen)
bull ORPcorrosion potentialbull Corrosion Inhibitorsbull Chloridebull Sulfatebull Manganese
bull Temperaturebull Sorptive surfaces downstream of
LSLs (ie galvanized interior pipe)bull Iron (deposition and corrosion)bull Calciumbull Aluminumbull NOM (type amount)bull Mixing from water treatment plants
or sourcesbull Ammoniabull Hydrogen Sulfidebull Microbial activity (nitrification amp
other)
Corrosion Control Treatment is intertwined with all treatments affecting water chemistry
Water Quality Parameters(WQP)
bull pH alkalinity calcium orthophosphate or silica othersbull WQP measurements are important for
bull Process control (measurement frequency needs to be HIGH)bull Keeping treatment from being turned offbull Keeping inhibitor dosages amp pH control from being cut to save $$$bull Keeping required repairs to treatment on tight schedulebull Preventing frequent variations and alternations of water sources
bull Effectiveness of Pb treatment only determinable from tap sampling by combination of different assessment approaches for different lead plumbing sources to which people can be exposed
36
bull OWQP measurements important for process control look at yourdata to understand what they mean and avoid problems
bull OWQP do not predict lead release There are no substitutes fordirectly monitoring lead release
37
M+
M+
M+
M+
Corrosion barrier film hypothesis
Inert barrier film (eg CaCO3)
Pipe Wall
Barrier film would allow tobull Protect ALL pipe materialsbull Monitor corrosion based on one chemical reaction
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
Water Quality Parameters(WQP)
bull pH alkalinity calcium orthophosphate or silica othersbull WQP measurements are important for
bull Process control (measurement frequency needs to be HIGH)bull Keeping treatment from being turned offbull Keeping inhibitor dosages amp pH control from being cut to save $$$bull Keeping required repairs to treatment on tight schedulebull Preventing frequent variations and alternations of water sources
bull Effectiveness of Pb treatment only determinable from tap sampling by combination of different assessment approaches for different lead plumbing sources to which people can be exposed
36
bull OWQP measurements important for process control look at yourdata to understand what they mean and avoid problems
bull OWQP do not predict lead release There are no substitutes fordirectly monitoring lead release
37
M+
M+
M+
M+
Corrosion barrier film hypothesis
Inert barrier film (eg CaCO3)
Pipe Wall
Barrier film would allow tobull Protect ALL pipe materialsbull Monitor corrosion based on one chemical reaction
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
37
M+
M+
M+
M+
Corrosion barrier film hypothesis
Inert barrier film (eg CaCO3)
Pipe Wall
Barrier film would allow tobull Protect ALL pipe materialsbull Monitor corrosion based on one chemical reaction
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
38
In excavated pipe samples that the EPA analyzed
Inert barrier film (eg CaCO3) is a complex scale
+ M+M
M+
M+ Scale- Heterogeneous- Several layers- Amorphous amp crystalline
Pipe WallPb and non-Pb compounds
Although barrier film would be convenient for important reasonsthe CaCO3 hypothesis has not been proven and scale is overall
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
Are these ldquocorrosionrdquo indicesIndex Equation Range Water Condition
Langelier Saturation Index
LSI = pHactual - pHs LSIgt0
LSIlt0
bull Supersaturated tends to precipitate CaCO3
bull Undersaturated tends to dissolve CaCO3
Ryznar Stability Index RSI = 2pHs-pHactual RSIgt6
RSIlt6
bull Undersaturated tends to dissolve CaCO3
bull Supersaturated tends to precipitate CaCO3
Aggressive Index AI=pH+ log(AH) AIlt1010ltAIlt12AIgt12
bull Highly Aggressivebull Moderately Aggressivebull Non-aggressive
Larson Scold Index Ls=[Cl-+SO4-2] Alkalinity Lslt08
08ltLslt12
Lsgt12
bull Chloride amp Sulfate will not likely interfere with natural film formation
bull chloride amp sulfate may interfere with natural film formation
bull Tendency toward higher localized corrosion rates
39
bull Most indices were not developed or named for corrosion (eg saturation stability etc)bull The LSI and subsequent variations predict CaCO3 SCALING not corrosion or metal
releasebull Calcium carbonate was not the active inhibition mechanism in the hundreds lead and
copper pipes analyzed by EPA
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
ldquoCorrosionrdquo indices
bull When used outside their limitations these indices become unreliable
bull Different objectives for different materialsbull THERE IS NO UNIVERSAL ldquoINDEXrdquo FOR
MULTIPLE MATERIALS
40
Relying on the LSI as an indicator of a waterrsquos corrosivityshould be abandoned (AWWARF 1996 Internal Corrosion ofDistribution Systems)
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
Summary
41
bull Materials inventory is critical to know where and what lead sourcesstill exist
bull To understand corrosion and lead release it is important to keep upwith pH the carbonate system (ie dissolved inorganic carbon andalkalinity) and disinfectants (ie oxidation-reduction potential)
bull Corrosion inhibitor efficiency depends on type and sufficientdoseconcentration for the background pH and other reactive waterquality constituents
bull There are many types of scale on lead pipe with different solubilitiesPb(+4) solubility ltlt Pb(+2)
bull Optimum Water Quality Parameters (OWQP) important for processcontrol look at your data to understand what they mean and avoidproblems
bull Setting OWQP or calculating ldquocorrosionrdquo indices does not predictlead release
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
How we communicate our research (recent examples)
Peer-review journal papersbull Doreacute E Lytle DA Wasserstrom L Swertfeger J Triantafyllidou S Field Analyzers for
Lead Quantification in Drinking Water Samples Critical Reviews in Environmental Science and Technology 2020 httpsdoiorg1010801064338920201782654
bull Lytle D A Schock M R Wait K Cahalan K Bosscher V Porter A Del Toral M Sequential Drinking Water Sampling as a Tool for Evaluating Lead in Flint Michigan Water Res 2019 157 40-54
bull Tully J DeSantis M K Schock M R Water QualityndashPipe Deposit Relationships in Midwestern Lead Pipes AWWA Water Science 2019 1 (2) e1127
bull DeSantis M K Triantafyllidou S Schock M R Lytle D A Mineralogical Evidence of Galvanic Corrosion in Drinking Water Lead Pipe Joints Environ Sci Technol 2018 52 (6) 3365-3374
Conference presentationsbull Triantafyllidou S Hensley K Lytle D James R Lal V Tools to Identify Lead Service
Lines AWWA Water Quality Technology Conference Dallas 2019bull Schock M Triantafyllidou S Tully J DeSantis M and Lytle D Diagnostic Sampling
Tools for Lead in Drinking Water AWWA Annual Conference and Exposition Denver 2019bull Tully J DeSantis M and Schock M Actual vs Predicted Lead Scale Observations from
the Field AWWA WQTC Toronto 2018
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
How we communicate our research (recent examples)
Magazine Articlesbull Lytle D A Schock M R Triantafyllidou S Identify Lead Plumbing Sources to Protect
Public Health AWWA Opflow 2018 44 (3) 16-20EPA Science Matters Newslettersbull Revealing the Complicated Nature of Tap Water Lead Contamination A Madison Wisconsin
Case Study July 30 2018bull Identifying the Best Lead Sampling Techniques to Protect Public Health October 22 2018bull Scaling Back EPA Researchers Help Communities Protect Drinking Water Systems from Lead
April 8 2019bull Search httpswwwepagovsciencemattersFact Sheets on Leadbull httpswwwepagovwater-researchwater-research-fact-sheetsDWHow to Identify Lead Free Certification Marks for Drinking Water System and Plumbing ProductsConsumer Tool for Identifying POU Drinking Water Filters Certified to Reduce LeadStandard Developmentbull Participate on the workgroups for NSF Drinking Water Treatment Units (point-of-use filters)
NSF 6061 (Health Effects for Drinking Water Treatment Chemicals and Health Effects for Drinking Water System Components) and ASHRAE Standard 514 (Building water quality)
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
How we communicate our research (recent examples)
Workshopsbull ORD-SSWROWASDWA 17th Small Drinking Water Systems Annual Workshop in Cincinnati next
one September 1-3 2020 Information at httpswwwepagovwater-research17th-annual-epa-drinking-water-workshop
bull Contributions to the 16th Annual EPA Drinking Water Workshop Cincinnati OH September 24-26 2019
Presentation Pipe Scale Analysis for Solving Lead Problems DeSantis et alBreak-out sessions on Corrosion Premise Plumbing and Lead in Schools (3Ts)Training Session Setting up Bench and Pilot Scale Studies Cost -Effectively
Webinarsbull ORDOW Small Systems Monthly Webinar Series schedule and recordings at
httpswwwepagovwater-researchsmall-systems-monthly-webinar-seriesLead Management in Homes and Buildings DeSantis Tully and Latham March 26 2019Lead and Copper Sampling and Water Quality Challenges Viveiros and Lytle July 26 2016Corrosion Control for Drinking Water Systems Schock and DrsquoAmico June 30 2015Reduction of Lead in Drinking Water Schock and Latham December 15 2015
bull 2018 AWWA Corrosion 101 WebinarSchock M Lead Corrosion Control 101 New Understanding of Best PracticesLytle D Copper Corrosion 101 Principles and Guidance
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
DisclaimerThe information in this presentation has been reviewed andapproved for public dissemination in accordance with USEnvironmental Protection Agency (EPA) The views expressedin this presentation are those of the author(s) and do notnecessarily represent the views or policies of the AgencyAny mention of trade names or commercial products does notconstitute EPA endorsement or recommendation for use
45
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
Corrosion is Oxidation-Reduction
AWWARF 1990 Lead Control Strategies AWWA Research Foundation and AWWA Denver CO
Millions of anodecathode sites across interior fresh lead pipe surface
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
What is Buffer Intensity (β)
Buffer Intensity is a measure of the resistance of a water to change pH upon addition of [H+] or [OH-] at constant DIC concentration
β = (partCApartpH)DIC for added acid concentration CA
Alkalinity is a measure of the capacity of a water to neutralize [H+] to a fixed point normally [HCO3
-] = [H2CO3]
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-
Difference between Alkalinity and BufferIntensity for Different pH
mg CaCO3LAlkalinity
20202020
pH mg CL Buffer IntensityDIC (β10000)
60 1483 62670 580 16080 486 02790 443 072
Computations for 25ordmC I=0005
- Slide Number 1
- Slide Number 2
- Slide Number 3
- Slide Number 4
- Slide Number 5
- Slide Number 6
- Slide Number 7
- Corrosion and scale formation
- Slide Number 9
- Slide Number 10
- Slide Number 11
- pH is master variable
- pH is master variable
- Slide Number 14
- Slide Number 15
- Slide Number 16
- ldquoClassicrdquo divalent Pb+2 solubility
- Lead ldquocorrosion Inhibitorsrdquo
- Slide Number 19
- Ortho-P Treatment for Pb+2
- Effect of pH and ortho-P on Pb release
- Ortho-P at pH 90 (DIC 6 mgL)
- Ortho-Ppoint of diminishing returns
- Sodium silicate
- Slide Number 25
- Slide Number 26
- Example profile of Pb(+4) (ie PbO2) scale house
- Chloride and sulfate
- Chloride and sulfate
- Chloride and sulfate
- Manganese
- Manganese
- Manganese
- Itrsquos not just manganese
- More water parameters affect metal release
- Water Quality Parameters (WQP)
- Corrosion barrier film hypothesis
- In excavated pipe samples that the EPA analyzed
- Are these ldquocorrosionrdquo indices
- ldquoCorrosionrdquo indices
- Summary
- Slide Number 42
- Slide Number 43
- Slide Number 44
- Slide Number 45
- Slide Number 46
- What is Buffer Intensity (b)
- Slide Number 48
-