Subsurface Flow and TransportModeling Research: Incorporating

Biologically Mediated Processes

Annual NABIR PI MeetingWarrenton, Virginia

Steve Yabusaki

Yilin Fang

Phil Long

Pacific Northwest National LaboratoryPacific Northwest National Laboratory

April 19, 2005

2

Predicting Coupled Process Behavior inPredicting Coupled Process Behavior inField-Scale SystemsField-Scale Systems

Recent Workshops“Integrating Numerical Models of Reactive Flow and Transport into FundamentalGeoscience Research,” DOE-BES, June 2003

“A Science-Based Case for Large-Scale Simulation,” DOE-SC, June 2003

“Conceptual Model Development for Subsurface Reactive Transport Modeling,”Multiagency Working Group on Subsurface Reactive Solute Transport, April 2004

Reliable prediction of field scale behavior is a scientific challengeMany field-scale issues are difficult to address at the lab scale

Many processes and properties are difficult to monitor in the field

Reactive transport models integrate fundamental earth science research andfocus on complex natural environments where individual time and space-dependent processes are linked

Need for multidisciplinary research teams dedicated to developing aquantitatively mechanistic understanding of behaviors at a particular field siteto address the range of scales and multiple interacting processes

Build field-scale process models on a framework of understanding fromfundamental experiments and characterization studies at complementarylength scales in the field

3

Conceptual Model: Old RifleConceptual Model: Old RifleBiostimulationBiostimulation Experiments Experiments

Bulk of uranium and sulfate in the aquifer originated asleachate from mill tailings

Uranium transported as U(VI) with the bulk adsorbed tothe sediments under background geochemical conditions

Acetate stimulates the growth of microbial populations thatremove aqueous U(VI) from solution via homogeneousreduction reactions that form uraninite

Initial bioreduction of aqueous U(VI) is 75 to 85 percentefficient and is attributed to iron reducing bacteria that useFe(III) minerals as a terminal electron acceptor

Once bioavailable iron is depleted, the iron reducingbacteria are succeeded by sulfate reducing bacteria,which are less efficient at U(VI) removal from groundwater

4

How do spatial and temporal variations inHow do spatial and temporal variations inhydrogeology and chemistry affect uraniumhydrogeology and chemistry affect uranium

behavior?behavior?Heterogeneous materials

Permeability

Iron and uranium

Depth-dependent U(VI) and DOHighest DO and U(VI) near the water table

IssuesOxygen diffusion through water table

Background utilization of DO

5

How do seasonal and episodic hydrologicHow do seasonal and episodic hydrologicevents affect uranium behavior?events affect uranium behavior?

Seasonal and event-driven changes

Velocity field

Oxidation of zonesaffected by water tablefluctuations

Issues

Rapid oxidation of zonesaffected by water tablefluctuations

Highest Uconcentrationsbypassing treatmentzones

6

What controls the post-amendmentWhat controls the post-amendmenturanium behavior?uranium behavior?

Residual enzymatic reductive capacity of biomass

Uranium surface complexation

Fe(II) adsorption / desorption

Mineral precipitation and dissolution [Fe(III)oxides/hydroxides, uraninite, FeS(am), siderite, calcite]

Coprecipitation

Alteration in surface reactivity

Alteration of access to reactive surfaces

7

Goal and ObjectivesGoal and Objectives

Goal

Systematic and quantitatively predictiveunderstanding of the mechanistic contribution byindividual subsurface processes to the observeduranium behavior at the Old Rifle UMTRA field site.

Objectives

Determine the interplay between microbial and abiotic reactionsgoverning field-scale bioremediation in the context of site-specifichydrologic and geochemical conditions

Determine the impact of biostimulation on the geochemical controls(Eh, carbonate speciation and complexation, pCO2, mineralsolubility, adsorption, pH) governing the mobility and long-term fateof uranium

8

ApproachApproach

Build field-scale conceptual process modelsFlow and transport

Biogeochemistry of biostimulation

Uranium surface complexation

Systematically integrate process models into acomprehensive field-scale flow and biogeochemicalreactive transport simulation capability

TEAPs and abiotic consequences of biostimulation affectinguranium mobility

Reoxidation and uranium mobility

Latent capacity for removal of aqueous U(VI)

Evolving surface chemistry (mineral precipitation/dissolution,adsorption/desorption)

PhilosophyStart simple to isolate major behaviors

Systematically add process complexity and detail

Use modeling to gain insight and target knowledge gaps

9

2002 Flow Field2002 Flow Field

2002 M-07 Water Level

7

7.2

7.4

7.6

7.8

8

8.2

8.4

Jun-02 Jul-02 Aug-02 Sep-02 Oct-02 Nov-02 Dec-02

Days From Injection

Wat

er T

able

Ele

vatio

n

Gradient Direction and Relative Magnitude

N

2002 field experiment

No prior augmentation

Steady flow field

Injection June 22 – Oct 23

Modeling assumptions

1-D domain

Constant velocity anddispersivity based onbromide transport

Bulk tank release ratedistributed uniformly overinjection gallery

Injection averaged oversaturated thickness

10

Initial Flow and Transport ModelingInitial Flow and Transport Modeling

10 mM Bromide injected

Breakthrough at monitoringwells

General trends reproduced withconstant velocity anddispersivity

Row 2 has highestconcentrations and maximumvariability missed by averageinjection

Preferential flow paths

Release not uniform or fullymixed

0

200

400

600

800

1000

0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300

Time (d)

Br

(uM

)

M-01

M-02

M-03

M-04

M-05

Model

25d 54d 89d 123d

0

200

400

600

800

1000

0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300

Time (d)

Br

(uM

)

M-06

M-07

M-08

M-09

M-10

Model

25d 54d 89d 123d

0

200

400

600

800

1000

0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300

Time (d)

Br

(uM

)

M-11

M-12

M-13

M-14

M-15

Model

25d 54d 89d 123d

Row 1

3.66 m downgradient

Row 3

14.63 m downgradient

Row 2

7.32 m downgradient

M-06 M-11M-01

Old Rifle Test Plot

Monitoring wells Control

Wells Acetate

GW flow

6.1m

16m

24m

Injection Gallery

M-15M-05 M-10

11

Biologically Mediated ReactionsBiologically Mediated Reactions

3 energetics-based TEAPreactions

Stoichiometry fromRittman/McCarty 2001

Includes biomass yield

Uranium reduction onlyactive during iron reduction

Dual Monod Rate LawHalf-saturation constantsfrom Wang/Jaffe et al. 2003

Calibrate rate and thresholdconcentration for utilization

( )

( )

acceptorelectronalminterfortcoefficiensaturationhalfK

substrateorganicfortcoefficiensaturationhalfK

acceptorelectronalminterforrateoxidationsubstrateorganic

nutilizatioacceptorelectronalminterfortcoefficienindicator

ionconcentratacceptorelectronalminterC

acetateeiionconcentratsubstrateorganicdissolvedC

acceptorselectronalminterofnumberN

where

sitessorptionfreewithtermlastreplacesTEAPIIIFe

CK

C

CK

CR

Cs

Cs

eAm

eA

eA

C

eA

eAeAs

eA

N

eA CCs

CeAmeA

bio

C

eA

=

=

=

=

=

=

=

++=

,

,

,

,,

,

.,.

)(

µ

µ

0.125CH3COO- + 0.6FeOOH(s) + 1.155H+ + 0.02NH4+ = 0.02BM_iron + 0.6Fe++ + 0.96H2O + 0.15HCO3

-

0.125CH3COO- + 0.0057H+ + 0.0038NH4+ + 0.1155SO4

-- = 0.0038BM_sulfate + 0.0114H2O + 0.231HCO3- + 0.1155HS-

0.1250CH3COO- + 0.3538H2O + 0.0113NH4+ + 0.775UO2

++ = 0.0113BM_iron + 0.855H+ + 0.1938HCO3- + 0.775UO2(s)

12

Sulfate / Sulfide BehaviorSulfate / Sulfide Behavior

3-4 mM/L sulfide generation

Typical field aqueoussulfide measurement 3uM/L

Implies sulfide associatedwith sediment

2003 AVS: equivalent to~15 mM/L, 5 mM/L, 4mM/L, 1.5 mM/L

Acid Volatile Sulfide (AVS)

Distance of Collected Sediment Core from Injection Gallery

-20 -10 0 10 20 30

0.0

0.5

1.0

1.5

2.0

2.5

Depth-13' BLS

Depth-15' BLS

Depth-17' BLS

Upgradient Downgradient

P-11Upgradient corecollected at -14.75ftfrom gallery.No AVS detected

P-12

P-13P-14

P-15

uM

/g

13

Insights: Need for additional Fe++Insights: Need for additional Fe++

FeS formation on surfaceinferred

Need 3-4 mM/L Fe++ toreact sulfide

Maximum aqueous Fe++measurement is 196 uM/L

Additional Fe++ wouldhave to be associated withsolid phases

Fe++ adsorption

Consistent with otherinvestigators

Multilevel sampling cell endsJanuary 2004

B-02 M-03 M-08 M-13

GroundwaterFlow Direction

Injection Gallery

Aquifer top

Aquifer bottom

PNC-CAT x-ray microprobe

Blackened sediment

Fe and Sulfur

No XRD for sulfide minerals

evidence for FeS(am)

14

Expansion of the BiogeochemicalExpansion of the BiogeochemicalReaction NetworkReaction Network

42 total reactions

Aqueous

Ca++, Fe++, K+, Mg++, Na+, H+, NH4+, Cl-, CO3

-, HS-, SO4--

Mineral

CaCO3, FeOOH, FeCO3, FeS, UO2

Sorption

Fe++

Biologically-mediated (by acetate)

Fe(III), U(VI), sulfate TEAPs

15

AcetateAcetate

Acetate injected for 123 days

In Row 1, acetate concentrationsdiminish significantly after 40days

Decrease in sulfate

Decrease in Fe++

Model: acetate peak diminisheswith travel distance

Acetate concentrations arespatially and temporally variable

Multiple peaks with later arrivalsin Row 1

Highest concentration in Row 2

Row 3 has much lower acetatethan predicted

0

500

1000

1500

2000

2500

3000

3500

4000

0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300

time (d)

Aceta

te (

uM

)

M-01

M-02

M-03

M-04

M-05

Model

25d 54d 89d 123d

0

500

1000

1500

2000

2500

3000

3500

4000

0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300

time (d)

Aceta

te (

uM

)

M-06

M-07

M-08

M-09

M-10

Model

25d 54d 89d 123d

0

500

1000

1500

2000

2500

3000

3500

4000

0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300

time (d)

Aceta

te (

uM

)

M-11

M-12

M-13

M-14

M-15

Model

25d 54d 89d 123d

Row 1

Row 2

Row 3

16

Uranium Uranium BioreductionBioreduction

Initial aqueous U(VI)spatially variable

Initial timing ofaqueous U(VI)removal reproducedby model

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300

time (d)

U(V

I) u

M

M-01

M-02

M-03

M-04

M-05

Model

25d 54d 89d 123d

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

0 20 40 60 80 100 120 140 160 180 200 220 240 260 280

time (d)

U(V

I) u

M

M-06

M-07

M-08

M-09

M-10

Model

25d 54d 89d 123d

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300

time (d)

U(V

I) u

M

M-11

M-12

M-13

M-14

M-15

Model

25d 54d 89d 123d

Row 1

Row 2

Row 3

17

Iron Iron BioreductionBioreduction, Dissolution, and Sorption, Dissolution, and Sorption

Initial Fe++ increasefollowed by decreaseafter onset of sulfatereduction

Model predictscumulative peak whichis not obvious in welldata

Effect is observed in2004 experiment

Issues

Fe++ adsorption

Fe++ production may bemuch higher

0

100

200

300

400

500

0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300

time (d)

Fe

(II)

(u

M)

M-01

M-02

M-03

M-04

M-05

Model

25d 54d 89d 123d

0

100

200

300

400

500

0 20 40 60 80 100 120 140 160 180 200 220 240 260 280

time (d)

Fe

(II)

(u

M)

M-06

M-07

M-08

M-09

M-10

Model

25d 54d 89d 123d

0

100

200

300

400

500

0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300

time (d)

Fe

(II)

(u

M)

M-11

M-12

M-13

M-14

M-15

Model

25d 54d 89d 123d

Row 1

Row 2

Row 3

18

Sulfate Sulfate BioreductionBioreduction

Sulfate reduction beginsafter 40 days of ironreduction

Rapid and completeutilization of acetate

Sulfate reduction limitedby acetate supply

3 - 4 mM sulfate removedunderestimated by model

Row 1

Row 2

Row 3

0

1

2

3

4

5

6

7

8

9

10

0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300

time (d)

Su

lfate

(m

M)

M-01

M-02

M-03

M-04

M-05

Model

25d 54d 89d 123d

0

1

2

3

4

5

6

7

8

9

10

0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300

time (d)

Su

lfa

te (

uM

)

M-11

M-12

M-13

M-14

M-15

Model

25d 54d 89d 123d

0

1

2

3

4

5

6

7

8

9

10

0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300

time (d)

Su

lfate

(m

M)

M-06

M-07

M-08

M-09

M-10

Model

25d 54d 89d 123d

19

Uranium Experiments on Rifle SedimentsUranium Experiments on Rifle Sediments

Rifle Aquifer Background Sample(RABS)

Sediment sample collectedDecember 2004

10 gallons of < 2 mm prepared

Experiments

Grand Junction

Princeton (Jaffe)

Issues

High background U

Break off point between labile andnonlabile

Kinetics

530504Elapsed time, h

1.120.208Labile U(VI), ug/g

4.2565.15Surface area, m2/g

< 2< 3Particle size, mm

RABSNABS

Set of batch equilibriumadsorption experiments

RABS+AGW-3

10-8 to 10-5 M/L U(VI)

Comparison of experimentalresults with Naturita surfacecomplexation model

RABS appears to be moresorptive

Adsorption of lower U(VI)concentrations sensitive tolabile U(VI)

Solid to

Solution

TIC pCO2 pH

ADS-1 25 g/L 8.25 mg/L 0.10% 7.62

ADS-2 820 108.35 2.39 7.33

ADS-3 125 92.7 2.23 7.28

ADS-4 125 100.2 3.62 7.06

ADS-5 250 238.4 22.8 6.52

ADS-6 125 100.8 5.23 6.90

ADS-7 250 213.6 20.8 6.52

20

Uranium Surface Uranium Surface ComplexationComplexation Modeling Modeling

Generalized composite surface complexation model

Preliminary fit of formation constants for 2 surface reactions, 3 sitetypes

1.92 µM/m2 site density (bidentate)

23 aqueous uranium complexation reactions

Includes Ca-UO2-CO3 ternary complexes

U(VI) Surface Reaction Log Kf

SSOH + UO22+

= SSOUO2+ + H

+ 12.28

SOH + UO22+ = SOUO2

+ + H+ 6.95

WOH + UO22+ = WOUO2

+ + H+ 2.74

SSOH + UO22+

+ H2O = SSOUOOH + 2 H+ 0.033

SOH + UO22+

+ H2O = SOUOOH + 2 H+ -2.12

SOH + UO22+ + H2O = SOUOOH + 2 H + -5.01

Total site concentration of 1.92 _moles/m 2

SSOH denoting very strong binding sites: 0.01% of total sites

SOH denoting strong binding sites: 0.1% of total sites WOH denoting weak binding sites: 99.89% of total sites

-0.50

0.00

0.50

1.00

1.50

2.00

2.50

3.00

-8.5 -8.0 -7.5 -7.0 -6.5 -6.0 -5.5 -5.0 -4.5

Log Dissolved U(VI) (moles/L)

Lo

g K

d (

mL

/g)

ADS-1

ADS-2

ADS-3

ADS-4

ADS-5

ADS-6

ADS-7

model ADS-1

model ADS-2

model ADS-3

model ADS-4

model ADS-5

model ADS-6

model ADS-7

21

Dissolved Oxygen Stratification:Dissolved Oxygen Stratification:Framework for 2-D Reactive TransportFramework for 2-D Reactive Transport

~20 cm layer of higher DO near watertable

Transported in from upgradient conditions

Diffusion through water table

OO22 influent with GW influent with GW

+ O+ O22 diffusion at WT diffusion at WT

+ O+ O22 microbial TEAP microbial TEAP

Flow

22

Next StepsNext Steps

2-D simulations to investigate depth-dependent issuesIncorporate new information on permeability distribution in the verticalIncorporate new information from geophysics

Refine uranium surface complexation modelAssess labile uranium and adsorption kineticsConsider other combinations of uranium surface complexation reactionsand site types in calibrationIdentify additional experiments for model refinement

Incorporate new process models into comprehensive reactive transportsimulator

Datasets2003 and 2004 field experimentsLaboratory experiments

Impacts of biostimulation on iron chemistrydissolution of uranium-bearing Fe(III) mineralsgeneration of Fe(II) and effect of adsorbed Fe(II) on bioavailability of Fe(III)mineralseffect of precipitation and dissolution on the accessibility of Fe(III) and U(IV)mineral surfaces to aqueous components

Post-amendment uranium mobilityconditions for residual capacity to remove U(VI) from groundwaterconditions where ambient geochemistry is being re-established

23

AcknowledgmentsAcknowledgments

University of Massachusetts

University of Tennessee

Dick Dayvault and Stan Morrison, SM StollerCorporation

This research was funded by the Natural andAccelerated Bioremediation Research (NABIR)Program, Biological and Environmental Research(BER), Office of Science, U.S. Department ofEnergy. Pacific Northwest National Laboratory isoperated by Battelle for the United StatesDepartment of Energy under Contract DE-AC06-76RLO 1830.

24

Backup SlidesBackup Slides

25

AGW-3

in air at 0.8547 atmpH 7.79pCO2 0.06%

Adsorption ExperimentsAdsorption Experiments

0.5

1.0

1.5

2.0

0 50 100 150 200 250 300

Experiments with AGW-3Approach based on Davis et al 2004

Assess applicability of GC-SCM approach

Compare with Naturita results

Adsorption Kinetics

25 g/L

1 uM/L U(VI) added

26

Depth-Dependent TransportDepth-Dependent Transport