water quality risk assessment of the fpaa101d specification fire … · 2019. 3. 13. · water...

Water quality risk assessment of theFPAA101D Specification fire sprinklersystem

Reviewofhealthandaestheticwaterqualityrisksassociatedwiththeintroductionofcombinationdomesticfiresprinklersystemsthatdirectly

connecttopotablewaterplumbingsystemsinresidentialbuildingsoflessthan25mmetresinheight.Documentdated17thOctober2018.

Water quality risk assessment of the FPAA101D Specification fire sprinkler system.17thOctober2018.ByDanDeere. Page2of33

TableofContents1 Documentinformation...........................................................................................................................4

2 Acronyms......................................................................................................................................................5 3 Executivesummary..................................................................................................................................6

3.1 Purpose...............................................................................................................................................6

3.2 Context................................................................................................................................................6 3.3 Approach............................................................................................................................................6

3.4 Stakeholders.....................................................................................................................................7

3.5 Summaryofprincipalfindings.................................................................................................7 3.5.1 Riskstohumanhealthfromentericpathogens...........................................................7

3.5.2 Riskstohumanhealthfromopportunisticpathogens..............................................7 3.5.3 Riskstohumanhealthfromtoxicants..............................................................................8

3.5.4 Riskstohumanperceptionfrompooraestheticwaterquality.............................8

3.5.5 Conclusions..................................................................................................................................8 3.6 Furtherconsiderationofrisksassociatedwithopportunisticpathogens.............9

3.6.1 Thestandardofbackflowpreventionrequired...........................................................9 3.6.2 Cross-connectiontothesprinklerline.............................................................................9

3.6.3 Toiletflushing..........................................................................................................................10

3.7 Summaryofprincipalrecommendations.........................................................................10 4 Introduction.............................................................................................................................................11

4.1 Purpose............................................................................................................................................11

4.2 Context.............................................................................................................................................11 4.3 Scope.................................................................................................................................................11

4.4 Approach.........................................................................................................................................11 5 Systemdescription................................................................................................................................12

6 Riskassessment......................................................................................................................................14

6.1 Riskidentification.......................................................................................................................14 6.2 Riskassessmentmethodology...............................................................................................14

6.3 Riskratingcriteria......................................................................................................................14

6.4 Resultsoftheriskassessment...............................................................................................14 6.5 Summaryofriskassessmentfindings................................................................................15

6.5.1 Identifiedriskpathways......................................................................................................15 6.5.2 Specificcontrolsofsignificance.......................................................................................16


6.5.3 Summaryofrisksrelatingtoentericpathogens.......................................................16 6.5.4 Summaryofrisksrelatingtoopportunisticpathogens.........................................17

6.5.5 Summaryofrisksrelatingtotoxigenicchemicals....................................................19

6.5.6 Summaryofrisksrelatingtoaesthetics.......................................................................20 7 Discussionoffindingsoftheriskassessment...........................................................................22

7.1 Prioritiesandlimitinghazards..............................................................................................22

7.1.1 Lowconsequencehazards..................................................................................................22 7.1.2 Moderateconsequencehazards.......................................................................................22

7.1.3 Highconsequence,limitinghazards...............................................................................22 7.2 Mostsignificantandlimitinghazardousevents............................................................23

7.2.1 Currentlyacceptedrisks......................................................................................................23

7.3 Relativeriskassessmentrelatingtoopportunisticpathogens...............................24 7.4 Professionalassessmentoftheriskfromopportunisticpathogens.....................24

7.4.1 Lengthofdeadlegs.................................................................................................................25 7.4.2 Thestandardofbackflowpreventionrequired........................................................25

7.4.3 Cross-connectiontothesprinklerline..........................................................................26

7.4.4 Toiletflushing..........................................................................................................................27 7.5 Quantitativeriskassessment.................................................................................................27

7.5.1 Toxicologicalriskassessment...........................................................................................27

7.5.2 Quantitativemicrobialriskassessment.......................................................................27 8 Conclusion.................................................................................................................................................29

9 References.................................................................................................................................................30


1 Documenthistory

Documentname Versionnumber

Documentauthors

Documentreviewer

Date ofversion

Water quality riskassessment of theFPAA101D Specificationfiresprinklersystem

Version1 DanDeere MarkWhybro 6thOctober2018

Version2 DanDeere Mark Whybroandstakeholders

7thOctober2018


16thOctober2018


17thOctober2018


2 Acronyms

Abbreviation Meaning

ABCB AustralianBuildingCodesBoard

ADWG AustralianDrinkingWaterGuidelines

AFAC AustralasianFireandEmergencyServiceAuthoritiesCouncil

AHPPC AustralianHealthProtectionPrincipalCommittee

BPAA BackflowPreventionAssociationofAustralia

BETG BuiltEnvironmentTechnicalGroup

EnHealth EnvironmentalHealthStandingCommittee

FPAAustralia FireProtectionAssociationAustralia

FRNSW FireandRescueNSW

NCC NationalConstructionCode

PCC PlumbingCodesCommittee

PFC ProposalForChange

QMRA Quantitativemicrobialriskassessment

TRA Toxicologicalriskassessment


3 Executivesummary

3.1 PurposeThe purpose of this review was to provide an independent, evidence-based andreferencedreportthatassessestheproposednewresidentialsprinklerdesignassetoutintheFireProtectionAssociation(FPA)Australia’sTechnicalSpecifications“FPAA101DAutomaticFireSprinklerSystemDesignandInstallation–DomesticWaterSupply”(draftatthetimeofwriting).Thereviewwasundertakenbyanindependentdrinkingwaterquality specialist experienced in assessing risks arising in buildings from plumbingsystemswithmore than one plumbing system (e.g. plumbing for recycled, irrigation,process,flushinganddeadlegwaters).

3.2 ContextTheFPAA101DSpecification responds to a jointProposal forChange (PFC) fromFPAAustralia,FireandRescueNSW(FRNSW)andAustralasianFireandEmergencyServiceAuthoritiesCouncil(AFAC)tothe2019NationalConstructionCode(NCC),whichseeksto mandate sprinklers in Class 2 and 3 Sole Occupancy Units and common areas inresidentialaccommodationinbuildingslessthan25mmetresinheight.

Theproposedchangewillallow“CombinationDomesticFireSprinklerSystems”forsuchbuildings to have their sprinklers directly connected to the domestic water supplyplumbingratherthanbeingfedbyafullyseparatefirefightingplumbingsystem.ForsuchCombinationDomesticFireSprinklerSystems,thePFCproponentsconsiderthatitwillnot be necessary to install testable backflow prevention devices on thewater supplyplumbingdirectlyfeedingthesprinklers.Thesprinklerswillbefedfromplumbingsupplylinesthatfeedotherpartsofthepropertyandwilloftenbeveryshortand,assuch,theFPAA101DSpecificationrequiresnon-testabledualcheckvalvebackflowprotectionbutwill not require testable backflow prevention devices on the feeds to the sprinklerplumbingforeachindividualsprinklerfeedline.Duetothisdifferencebetweentestableandnon-testabledevices,backflowriskshavebeensingledoutfordetailedanalysis inthisreport.

3.3 ApproachThe assessment identifies and considers relevant contaminants (e.g. pathogens andtoxicants) and assesses events thatmay present risks to drinkingwater quality (e.g.backflow,cross-connection,diffusionorinhalation)fromimplementingtheFPAA101DSpecification. The assessment compared risks with benchmarks of water qualityprotectionassetoutinAustralia’sexistingplumbing,buildinganddrinkingwaterqualityguidelines and standards as they relate to public health, aesthetic quality and publicsafety. The assessment commentedon relative risks. For instance,manyof thewaterqualityrisksrelatetorelativelylowriskaestheticconsiderationswhereastheobjectiveof a fire suppression system is toprotect from the foreseeablehigh riskofdeathandserious injury from fire. In addition, risks relating to the proposed FPAA101DSpecificationwerecomparedtorisksassociatedwithcurrentplumbingstandardsandextantplumbingsystems.


3.4 StakeholdersParties interested in the findings of this assessment include the above-mentionedproponentsoftheFPAA101DSpecification(AFAC,FPAAustraliaandFRNSW)aswellastheAustralianBuilding CodesBoard (ABCB), Plumbing Codes Committee (PCC), BuiltEnvironment Technical Group (BETG), Backflow Prevention Association of Australia(BPAA), Queensland Health, WA Building and Energy, Plumbing Technical ReferenceGroup,andtheEnvironmentalHealthStandingCommittee(enHealth)oftheAustralianHealthProtectionPrincipalCommittee (AHPPC). Inaddition,waterutilities,plumbingregulators and inspectorsmay have an interest. The assessmentwas presentedwiththoseinformedexpertpartiesinmind,notforthegeneralpublicorlaypersons.

3.5 SummaryofprincipalfindingsFortheconvenienceofreaders, theprincipal findingsarisingfromtheassessmentaresummarisedasfollows.Furtherdetailsdiscussingthesefindingsaregiveninthebodyofthereport.

3.5.1 RiskstohumanhealthfromentericpathogensImplementingtheFPAA101DSpecificationwasnotconsideredtopresentelevatedrisksfromentericpathogens1whencomparedtoexistingplumbingstandards.Significantriskswere not identified due to the proposed FPAA101D Specification and riskswere notconsideredtobeelevatedrelativetorisksassociatedwithexistingplumbingstandards.The principal reason for this conclusionwas that deadlegs are not sources of entericpathogens,whicharefaecal-oralintheirtransmissionpathways.

3.5.2 RiskstohumanhealthfromopportunisticpathogensAlthough implementing the FPAA101D Specification was considered to present apotential risk fromopportunisticpathogens2, importantly,whencompared toexistingplumbing standards, the risk was similar to, or at worst only marginally andincrementallyelevated,relativetoexistingstandards.TheFPAA101DSpecificationdidnotpresentastepchange increase inrisksincenonewbroadcategoriesorclassesofhazards or risks were introduced. The potentially incremental increase in risk wasrelated to the presence of fire sprinkler system deadlegs protected from the potablewatersupplybyanon-testabledualcheckvalve.Thisriskcomparestoplumbingsystemsthat either lacked fire sprinkler systems and the associated deadlegs, or that hadconventionalfiresprinklersystemsincludingdeadlegschargedbyaseparatefireservicefromwhichthepotablesystemisprotectedbyatestablebackflowpreventiondevice.Inthelattercaseitwasnotedthatconventionalfiresprinklersystemswouldbeexpectedtopresentagreatertotalvolumeofdeadlegduetothegreatertotalvolumeofwaterinthefire service plumbing system as a whole. It was noted that deadlegs are alreadycommonplaceinbuildingsduetoseldomusedorabandonedtapsandfittingsandmost

1 In this context, enteric pathogens are microorganisms that cause gastrointestinal illness, such asdiarrhoeaandvomiting,asapartoftheirnormallifecycle,e.g.SalmonellaandGiardia.Wheretheybecomewaterbornetheyarepassedduetofaecalcontaminationofwaterthatistheningested.2Inthiscontext,opportunisticpathogensaremicroorganismsthatdonotnormallycauseillnessandlivenaturallyinwateratlowlevelsbutthatcancauseillnessundercertaincircumstancesiftheyreachhighenoughlevelsinstagnantwaterlackinginadisinfectantresidual,e.g.LegionellaandNaegleria.


havenoformofbackflowprotectionwithinthebuilding.TheseriskswereconsideredinmoredepthwithinthebodyofthisreportandarediscussedagainwithinthisExecutiveSummaryatsection3.6.

3.5.3 RiskstohumanhealthfromtoxicantsAlthough implementing the FPAA101D Specification was considered to present apotential risk associated with chemical toxicants3, importantly, when compared toexistingplumbing standards, the riskwas similar to, or atworst onlymarginally andincrementallyelevated,relativetoexistingstandards.TheFPAA101DSpecificationdidnotpresentastepchange increase inrisksincenonewbroadcategoriesorclassesofriskswereintroduced.Anypotentiallyincrementallyincreasedriskwasconsideredtobeinsignificantonthegroundsthattobehazardoustohealthchemicalseitherneedtobeacutely toxic at levels that can be ingested without being discernible, or need to beconsumed over long periods. Reasonably foreseeable scenarios by which chemicaltoxicantsthathadbuiltuptohazardousconcentrationsindeadlegswouldreachpotablewater fittings would be expected to be rare, sporadic, short-lived and typically self-evidentandself-limitingduetothediscerniblypooraestheticqualityofstagnantwater.Furthermore, unlike conventional fire sprinkler systems, the FPAA101D SpecificationrequiredWaterMarkpotablewatergradeplumbinglinesandfittingstobeused.

3.5.4 RiskstohumanperceptionfrompooraestheticwaterqualityAlthough implementing the FPAA101D Specification was considered to present apotentialriskassociatedwithpooraestheticwaterquality4,importantly,whencomparedtoexistingplumbingstandards,theriskwassimilarto,oratworstonlymarginallyandincrementallyelevated,relativetoexistingstandards.TheFPAA101DSpecificationdidnotpresentastepchange increase inrisksincenonewbroadcategoriesorclassesofriskswereintroduced.Anypotentiallyincrementallyincreasedriskwasconsideredtobeinsignificant becausepoor aestheticwater quality arisingunder rare, sporadic, short-livedscenarioswouldbelikelytobeconsideredacceptablegiventheoverwhelmingandself-evidentbenefitsofhavingafiresprinklersystemwhereonemightotherwisenotbe.Inaddition,sucheventswouldbereadilymitigatedbysimplyflushingthewaterforavery short period given the limitations placed on the lengths of deadlegs within theFPAA101DSpecification. Furthermore, unlike conventional fire sprinkler systems, theFPAA101DSpecificationrequiredWaterMarkpotablewatergradeplumbing linesandfittingstobeused.

3.5.5 ConclusionsThe FPAA101D Specification presents no increase in risk associated with entericpathogensrelativetoexistingplumbingstandards.

TheFPAA101DSpecificationpresentsatworstan incremental,butnota stepchange,increaseinrisksassociatedwithopportunisticpathogens,toxicantsandaestheticwaterqualityrelativetoexistingplumbingstandards.

3 In this context, toxicantsarechemical substances that can lead toadversehealthconsequencesuponingestion,e.g.plumbingmaterialsthatrisetoelevatedconcentrationsaboveguidelinevalues.4Inthiscontext,aestheticwaterqualityreferstoperceptibleaspectsofwater,e.g.colour,odourortaste.


The potentially increased risks associated with toxicants and aesthetic water qualitypresented by the FPAA101D Specification are sufficiently low and insignificant thatfurther study and analysis is not warranted and no modification to the FPAA101DSpecificationneedstobeconsidered.

ThepotentiallyincreasedrisksassociatedwithopportunisticpathogenswereconsideredpotentiallysignificantenoughtowarrantfurtherreviewoftheFPAA101DSpecification.Theseriskswereconsideredfurther.

3.6 FurtherconsiderationofrisksassociatedwithopportunisticpathogensItcanbeassumedthatopportunisticpathogenswillpersistwithinthedeadlegsfeedingsprinklers. Exposure to persons from those pathogens could occur through threeprincipalmechanisms:

• Backflowofwaterfromthesprinklerplumbingtopotableplumbingsystem.• Cross-connectionbetweenthesprinklerplumbingandpotableplumbingsystem.• Aerosolsfromtoiletflushingwithwaterfedviathetoiletflushingfeedwater.

3.6.1 ThestandardofbackflowpreventionrequiredOnthebasisofthebackflowriskbeinglowerthanthatacceptedfromexistingdeadlegs,(bothdue to thepresenceof someexistingdeadlegsbeing longer than3mand theirlackingbackflowprotection),therecommendationintheFPAA101DSpecification,thatdeadlegsfeedingsprinklerslimitedtoa3mmaximumlineallengthbeprotectedfromthepotablenetworkbybothatoiletplumbingfeedlineandanon-testabledualcheckvalvebackflow device, is considered sufficient risk mitigation. Conventional fire sprinklersystems require a testable backflowprevention device but such systemswould oftencarryagreatervolumeofdeadlegwaterthantheproposedFPAA101DSpecificationandwouldlackprotectionssuchaslimitationsonlineallengthandtheregularlyturnedovertoiletplumbingfeedline.Existingdeadlegsinconventionalplumbingsystemscarrynospecialprotections.

3.6.2 Cross-connectiontothesprinklerlineOn thebasis of the cross-connection riskbeing similar to that accepted fromexistingdeadlegs,andlowerthantheriskassociatedwithcross-connectionstoexistingdeadlegsonconventionalfiresprinklersystemsthatcouldbemuchlongerthan3m,noadditionalmodifications to the FPAA101D Specification are considered necessary. However, asrecommendedbyQueenslandHealth,somefurtheremphasisonplumbing inspectionsandsignageiswarrantedforpropertiesinstallingcombinedsprinklersystemsundertheFPAA101DSpecification.ItisunderstoodthattheFPAA101DSpecificationrequiresafullsysteminspectionatcommissioning.Inaddition,itisunderstoodthatoneveryfloortheFPAA101Dsprinklerisolationvalvewillhaveasigntoidentifythevalveandstatethatthe sprinkler system is connected to the toiletson the floor. Itwouldbenecessary toisolate this valve before tapping into the FPAA101D system, and tappings should beundertakenbycompetentlicencedindividuals.Asabenchmark,thisisconsistentwiththe observation that plumbing regulators often require higher priority and higherfrequencyinspectionsandenhancedsignagefordualreticulationplumbingsystemsthatcarrynon-potable,e.g.recycled,water.


3.6.3 ToiletflushingAdditional risks fromexposure to toilet flushingwater fedvia thesprinklerplumbinglineswereconsideredtobeverylow.Toiletsarewidelyunderstoodtocontainpotentiallycontaminatedwaterandtoiletsare,bydesign,notsourcesofsmallsize-classaerosols.Asaresult,toiletsareaverylowlikelihoodsourceofopportunisticrespiratoryinfections.Therefore, no special controls or requirements are considered necessary in theFPAA101DSpecification.

3.7 SummaryofprincipalrecommendationsTheprincipalconsideration inagreeingthe finedetailsof theFPAA101DSpecificationshould be whether the potentially incrementally increased risk associated withopportunistic pathogens relative to existing plumbing standards is acceptable.Whilstotherrisksarenoted,therisksfromopportunisticpathogensarethemostsignificantandcanbe considered tobe the limiting factor fordecision-makers. If controlswithin theFPAA101DSpecificationaresufficient toadequatelymitigate these limitingrisks fromopportunisticpathogensthenthosecontrolsshouldmorethanadequatelymitigateotherrisks.

Based on the risk assessment presented here, the controls proposed within theFPAA101D Specification would reduce the risks posed by opportunistic pathogensarisingfromthefiresprinklersystemdeadlegstoalevelthatisinadditionto,butlowerthan, the existing risks posedby other deadlegswithin buildings. The reason for thisconclusion is that existing deadlegs lack the three key controls within the proposedFPAA101DSpecification,being:

i) thehydraulicbufferfromtheregularlyturnedoverwaterfeedingthetoilet;ii) thenon-testabledualcheckvalvebackflowdevice;andiii) thelimitationondeadlegpipelengths.

Therefore,were a building to have plumbing of a nature, andwater of a quality andtemperature,thatitwouldsupporthazardouslevelsofopportunisticpathogensarising,theFPAA101DSpecificationwouldpresentariskfromfiresprinklerdeadlegsthatwasinadditionto,but inandof itself lowerthan, thatposedbyotherdeadlegswithinthesamebuilding.

Giventhatthefiresprinklersystemhasaself-evidentbenefit(Duranceauetal,1999),whereastheotherdeadlegsdonot, theproposedFPAA101DSpecificationpresentsanacceptable practicable solution since it encourages the installation of fire sprinklerswheretheymightotherwisebeabsentwithoutpresentingmorethanatworstamarginal,incrementalincreaseinriskfromopportunisticpathogens.


4 Introduction

4.1 PurposeThe purpose of this review was to provide an independent, evidence-based andreferencedreportthatassessestheproposednewresidentialsprinklerdesignassetoutintheFireProtectionAssociation(FPA)Australia’sTechnicalSpecifications“FPAA101DAutomaticFireSprinklerSystemDesignandInstallation–DomesticWaterSupply”(draftatthetimeofwriting).Thereviewwasundertakenbyanindependentdrinkingwaterquality specialist experienced in assessing risks arising in buildings from plumbingsystemswithmore than one plumbing system (e.g. plumbing for recycled, irrigation,process,flushinganddeadlegwaters).

4.2 ContextTheFPAA101DSpecificationwasadraftatthetimeofwritingwiththemostrecentdraft,as seen when completing this review, updated and dated to 21st August 2018. TheFPAA101DSpecificationwasdevelopedfortheFPAAustralia,FRNSWandAFACProposalforChange(PFC)tothe2019NationalConstructionCode(NCC),whichseekstomandatesprinklers in Class 2 and 3 Sole Occupancy Units and common areas in residentialaccommodationinbuildingslessthan25mmetresinheight.Theproposedchangewillprovidefor“CombinationDomesticFireSprinklerSystems”forsuchbuildingstohavetheirsprinklersdirectlyconnectedtothedomesticwatersupplyplumbingratherthanbeing fedbyaseparate firefightingplumbingsystem.ForsuchCombinationDomesticFireSprinklerSystems,thePFCproponentsconsiderthatitwillnotbenecessarytoinstallbackflow prevention devices on the water supply plumbing directly feeding thesprinklers.Thesprinklerswillbefedfromplumbingsupplylinesthatfeedotherpartsofthepropertyandwilloftenbeveryshortand,assuch,theFPAA101DSpecificationwillrequirenon-testabledualcheckvalvebackflowpreventionratherthantestablebackflowpreventiondevicesonthefeedstothesprinklerplumbingforeachindividualsprinklerfeedline.

4.3 ScopeThedocumentreportsonhealthandaestheticriskstopersonsresidinginbuildingsthathavefiresuppressionsystemsmeetingtheproposedFPAA101DSpecification.

4.4 ApproachThephysicalsystemunderconsiderationwasfirstdescribedinsummaryformbasedonwhat isdocumentedintheFPAA101DSpecification.Riskswerethenassessedforthatsystem,asdescribed.TheriskassessmentfocusedonwhatisdifferentaboutaplumbingsysteminabuildingthatmeetstheproposedFPAA101DSpecificationasdistinctfromother more conventional plumbing systems. Risks inherent in the existing, moreconventionalsystemsarenotassessedinthisdocumentsincetheyareconsideredtobeexisting,acceptedrisks.However,directcomparisonsaremadebetweenrisksassociatedwiththeproposedFPAA101DSpecificationandthoseexistingwithinbuildingsthatmeetexistingplumbingstandardsorareknowntotypicallyexistwithincurrentbuildings.Thiscomparisonwithexistingsystemsisimportantasitprovidesabasisfordefiningalevelofacceptable,oratleastcurrentlyaccepted,risk.


5 SystemdescriptionThesystemconsidered in thisriskassessment isdescribed indetail in theFPAA101DSpecification (FPA Australia, 2018). A concise summary is illustrated as a schematicplumbingdiagram(Figure5-1)andprocessflowdiagram(Figure5-2).Bothdiagramsareillustrative with actual configurations being building-specific. The purpose of thediagramsistoconceptuallyillustratetheprincipalwaterflowsandprocesssteps.Forthepurposesoftheriskassessmenttheprocessflowdiagramisusedastheprincipalpointofreference.Mostoftheprocessisconsistentwithaconventionaldrinkingwatersupplyplumbingsystem.Departuresfromsuchconventionsare:

• A dedicated dual check valve non-testable backflow device and plumbing linesupplyingthetoiletfacility.

• Asetofdedicateddeadlegplumbinglines,meetingthepotablewaterplumbing‘WaterMark’ standards, endingwithwet sprinklers connected to the plumbinglinesupplyingthetoilet.

Figure 5-1. Illustrative schematic plumbing system diagram of the proposedFPAA101DSpecificationsprinklersystem(FPAAustralia,2018).


Figure5-2.ExampleprocessflowdiagramforaFPAA101DSpecificationsprinklersystem.

Drinking water reticulation main

Property service connection

Dual check valve

Water meter

(Pump)

Riser

Floor isolation valve

(Water meter)

Domestic plumbing

Dual check valve

Toilet plumbing line

Toilet

Deadlegplumbing line

Sprinkler


6 Riskassessment

6.1 RiskidentificationPotential risks that have been identified are described and summarised in Table 4,appendedtothisreport.Therisksincludedwereidentifiedlargelybaseduponfeedbackfollowing expert technical review by stakeholders to the FPA Australia SpecificationFPAA101DthatformthePCC.Thestakeholdersidentifyingspecificrisksnotedalongsidethe summary of risks, below. This report has taken those risks, as identified, andpresentedtheminastructuredriskassessmentformat.Notethattheassessmentcanbemodifiedinresponsetofeedbackandincludessomequalitativeriskassessmentrankingsthatarebasedonprofessionaljudgement.

6.2 RiskassessmentmethodologyFor the purposes of presentation, the risk assessment grouped and ordered risksaccordingtoprocessstepsidentifiedwithintheprocessflowdiagram(Figure5-1),withrisksbeingassessedforanumberofprocesssteps.

TheriskassessmentfocusedonwhatwasdifferentabouttheFPAA101DSpecificationasdistinctfromothermoreconventionalplumbingsystems.Risksinherentintheexisting,moreconventionalplumbingsystemsarenotassessedinthisdocumentsincetheyareconsideredtobeexisting,acceptedrisks.

For each risk, potential contaminants (‘hazards’) that might be hazardous to humanhealth or aesthetic water quality, and events by which the contaminants could beintroducedand/orexacerbated(‘hazardousevents’),werenoted.

RiskswereratedasiftherewerenospecificcontrolsgivenintheFPAA101DSpecification(as ‘maximum’ risks) and then again with consideration being given to the specificcontrolsgivenintheFPAA101DSpecification(as‘residual’risks).

6.3 RiskratingcriteriaRiskswereratedusingtheAustralianDrinkingWaterGuidelines(ADWG,2011)exampleriskratingcriteria,(whichisbroadlyconsistentwiththeapproachtoriskassessmentsetoutinAS/NZSISO31000:2009RiskManagement–PrinciplesandGuidelines).

For convenience, the ADWG risk rating criteria are reproduced here in Table 1(likelihood),Table2(consequence),andTable3(overallrating).

Inaddition,arecordwasretainedofthebasisoftheriskratingandacommentwasmadeontheuncertainty,orlevelofconfidence,inthatrating.TheratedrisksaresummarisedinTable4,appendedtothisreport.

6.4 ResultsoftheriskassessmentTheresultsoftheriskassessmentaregiveninanattachedsummarytable(Table4)withthelive,originalfilebeingretainedasanExcelworkbookforeaseoffutureupdating.

Table1.Risklikelihoodratingcriteria(imageextractedfromtheADWG,2011).


Table2.Riskconsequenceratingcriteria(imageextractedfromtheADWG,2011).

Table3.Riskratingcriteria(imageextractedfromtheADWG,2011).

6.5 Summaryofriskassessmentfindings

6.5.1 IdentifiedriskpathwaysAnumberofriskfactorswereidentifiedthatmadeabuildingplumbedinaccordancewiththeFPAA101DSpecificationsomewhatdifferent toonemoreconventionallyplumbed.Water supply distribution and plumbing systems are increasingly being identified aslocations in which water supply contamination arises which can in turn result inwaterborne disease outbreaks (Craun and Calderon, 2001). Based on these historicalreviews of waterborne disease outbreaks, the major credible and established riskpathwaysofrelevancetotheFPAA101DSpecificationincludethefollowing:

• Water hammer, pressure transients and changes of pressure, including loss ofpressure, could lead to the physical hydraulic drawing ofwater fromdeadlegsboth back into the potable water supply system and forward into the toiletflushingwater.

• Diffusioncouldleadtothemovementofcontaminantsfromdeadlegsbothbackintothepotablewatersupplysystemandforwardintothetoiletflushingwater.


• Cross-connections,mis-connectionsandtap-insthatconnectplumbinglinesnotintendedtocarrypotablewatertothepotablewaterplumbinglinescouldtransfercontaminantsintothepotablewater.

• Inallofthethreeaboveexamples,therewouldneedtobeanexposurepathwayfor adverse consequences to arise. In relation to failure modes involving thetransfer of contaminants into potablewater, exposure pathways could includeaerosolinhalation, ingestionandskin,noseoreyecontact.Inrelationtofailuremodesinvolvingtransferofcontaminantsintothetoiletflushingwater,importantexposurepathwaysaremostlikelylimitedtoaerosolinhalation.

6.5.2 SpecificcontrolsofsignificanceAnumberofspecificcontrolsgivenintheFPAA101DSpecificationwereofsignificanceinthattheysignificantlyreducedmaximumrisksandtherebyreducedtheratinggiventotheresidualrisks.Thesesignificantcontrolsincludedthefollowing:

• “2.5.3.2.1 Sprinkler pipework deadlegs/stagnant water. The sprinkler systempipework shall be designed and installed so that no part of any pipeworkcontainingstagnantwaterismorethan3minlineallength.”

o Thiscontrolsignificantlylimitsboththepotentialsurfaceareaandvolumethatcouldgiverisetorisksassociatedwithcorrosionproductsandbiofilmformationandtostagnantwaterandleachingwithinthosedeadlegs.

• “2.5.3.2.3Backflowprevention.AbackflowpreventiondeviceinaccordancewithClause5.3.2shallbeinstalledonthesprinklerpipework...”

o Thiscontrolreduces the likelihoodofwater flowing fromdeadlegsbackinto themainpotablewater systemusingnon-testabledual checkvalvedevices.

• “2.5.3.2.4Connectiontotoilets.Alltoiletsshallbeconnectedandsupplieddirectlyfromthesprinklersystem...”

o This control provides a means to keep water turning over in themainfeederlinetothedeadlegsaswellasprovidingahydraulicbarrierwithavolumeof regularly turnoveroverwater sittingasabufferbetween thedeadlegsandthemainpotablesystem.Thehydraulicbufferwon’tpreventbackflowintheeventoftotalpressurelossbutwillhelpmitigatetheeffectsof diffusion and minor hydraulic transfers due to water hammer andpressuretransients.

• “2.5.3.3Connectiontoothersystems.OtherthandetailedinthisSpecification,nopart of the sprinkler system shall be connected to any other service includingdrenchersystems,wallwettingsystems,hosereelsorthelike.”

o Thiscontrolhelpstominimisetheextenttowhichthemainfeederlinetothe deadlegs and the deadlegs themselvesmight be connected to othercomponentsoftheplumbingsystemwhichmightinturnleadtoelevatedexposuretowaterinfluencedbystagnantwaterfromthosedeadlegs.

6.5.3 SummaryofrisksrelatingtoentericpathogensEnteric(gastrointestinal)pathogensaremicroorganismsthatinfectthegastrointestinaltractandthatcanbetransmittedbypotablewaterincludebacteria(e.g.Salmonellaspp.and Campylobacter spp.), parasites (e.g. Cryptosporidium spp.) and viruses (e.g.


norovirus) (ADWG).They typically causediarrhoeaandvomitingandcancausemorecomplexsymptoms,andsometimesdeath.Ingeneral,inrelationtowaterbornedisease,entericpathogensarethemosthazardouscontaminantthathashistoricallycausedthemost harm to health from contaminated drinking water. The primary purpose ofproviding safewater is toprevent transmissionof entericpathogens.Therefore, risksassociatedwithentericpathogenswereconsideredfirst.

Enteric pathogens are typically only present in water if that water has becomecontaminatedwithfaecalmatterorputrescingfoodwaste.Theyarenomorelikelytobepresent in deadleg fire sprinkler lines than in any other plumbing since they do nottypicallypersistorgrowinpotablewater,butrathertheygraduallydieoffovertime.

Itwasconcludedthatentericpathogensareprimarilyderivedfromfreshfaecalmatterandarenotspecificallyofconcernindeadlegsorpipework.Therelativelylownutrientsfound inpotablewaterdonot typicallysupport theproliferationofentericpathogensevenintheabsenceofadisinfectantresidual,e.g.indeadlegs.

Significant concentrations of enteric pathogens capable of resulting in significantconsequencesarenotposedbytheproposedFPAA101DSpecification,evenwithoutanydedicatedcontrols,resultinginbotha lowmaximumandlowresidualrisk.Therefore,risksposedbyentericpathogenswerenotconsideredfurther.

6.5.4 SummaryofrisksrelatingtoopportunisticpathogensOpportunisticpathogensaremicroorganismsthatarenaturallypresentinwaterandthataren’tordinarily,(aspartoftheirordinaryecologicalnicheor lifecycle),pathogenictohumans. However, these microorganisms can potentially cause infections in humans‘opportunistically’iftheopportunitytoinfectapersonarises.Potablewatercanprovideameanstotransmitsuchpathogensifexposurepathwaysarise(Benthametal.,2007).Infectionscanoccurviaskincontact,eyes,ingestion,inhalationandnasalcontact.

Thehighertheconcentrationofsuchmicroorganismsinwater,thehighertherisk.Therelevance to the FPAA101D Specification is that such opportunistic pathogens canproliferatetolevelsofconcerninstagnantwater.Therisksrisewherethepathogenscanformaspartofbiofilms(e.g.onpipesurfaces), inwaterlackingadisinfectantresidual(e.g.indeadlegswherewaterisnotturnedovertobringinfresh,effectivedisinfectant),andinwarmwater(e.g.inthe20to50˚Ctemperaturerange,andparticularlyinwaterthatiscontinuallyabove25˚C,orseasonallyabove30˚Cformonthsatatime).

OfmostsignificancewithintheAustraliancontextthisgroupofmicroorganismsincludesbacteria,suchasLegionellaspp.(thatinfectstherespiratorysystem)andamoebae,suchasNaegleriaspp.(thatinfectsthebrainviathenose)andAcanthamoebaspp.(thatinfectstheeyes).

Theriskassessmentfoundthatwithoutaddinganyspecialcontrols,addingwetsprinklersystemstothepotablebuildingplumbingnetworkwouldpresentariskofopportunisticpathogensbackflowinganddiffusingintothepotableplumbingsystem,beingtransferredforward intothetoilet flushingwaterandbeing fed intothepotablewatersystemviacross-connections.Connectingdeadlegsprinklersupplylinesprovidesthepotentialforproliferationofopportunisticpathogenswithinthosedeadlegsresultinginapotentiallyveryhighriskifuncontrolled.


Torespondtothatrisk,threecontrolswerenotedintheFPAA101DSpecificationthatreducedtheriskrelativetowhatitwouldhavebeenwithnosuchcontrols.Thesecontrolsare summarised in section 6.5.1, above. Briefly, the proposed controlsmean that thelengthofdeadlegpipeworkisminimised,theriskofbackflowisreduced,andthewaterrunningtothetoiletwillprovideahydraulicbufferofwaterthatisturnedoverbetweenthe deadleg pipework and the potable plumbing system.Only the first of these threecontrolswasrelevanttothetoiletflushingwaterbutexposurestotoiletflushingwaterwereconsideredtobemuchmorelimitedintermsofboththeirpathwayandpotentialextent (limited to low likelihood of sporadic aerosol inhalation). It is noted thatconventional fire sprinkler systems with a second plumbing system have only oneprotectionagainstthisrisk,beingthetestablebackflowpreventiondevice.

Theproposedcontrolswillreduce,butnoteliminate,theriskofopportunisticpathogensdiffusingorbeingdrawnbackinto,orcross-connectedinto,thepotableplumbingsystem,orofbeinginhaledfollowingtoiletflushing.However,itwasnotedthat,particularlyforrisks related to backflow or cross-connection, the risk is further mitigated by thelikelihood that stagnant water would often be associated with elevated colour,particulates,tasteorodour(discussedinsection6.5.6)andwould,therefore,typicallybediscernible.

It should be noted that the risk of opportunistic pathogens being present atconcentrationsofconcernisheavilytemperature-dependent,andthatriskstohealtharerelatedtotheageofexposedpersons.Withincreasedglobalaveragetemperatures,andincreasedlifeexpectancyandhencemoreagedpersons,therisksposedbyopportunisticpathogens is increasing globally. Much of Australia is in climatic zones where watertemperatures routinely reside above the 25˚C range atwhich opportunistic pathogengrowth ismost problematic. For instance, ambientwater temperatures can routinelyexceed30˚C inmuchofnorthernAustralia and indoor shade air temperatures, in theabsenceofairconditioning,canoftenrangebetweenthemid-twentiestohighthirtiesof˚Cthroughoutmuchoftheyear.TheimplicationsarethatallplumbinginAustraliawilloftenbe,andwillincreasinglybe,inthetemperaturerangethatwillsupportthegrowthofopportunisticpathogenstolevelsofconcernintheabsenceofdisinfectantresiduals,e.g.indeadlegs.Ontheotherhand,thescaleofbuildingsthatareintendedtohousethefiresprinklersystemsdescribedintheFPAA101DSpecificationarelikelytoberelatively-wellprotectedfromextremesofheatandare likelytobe insulatedandcooled,unlikesmall,isolateddwellings.

Itisimportanttonotethattheincidenceofopportunisticpathogeninfectionsattributedtodrinkingwatersuppliesandplumbingsystems,(asdistinctfromestablishedhighrisksources,suchascoolingtowers),hasbeenincreasinginrecentdecades.Someofthemoststriking evidence of this comes from the US Centres for Disease Control (CDC) whoundertake systematic long-term reporting on waterborne disease for the US. NoequivalentreportingisavailableforAustralia,whichiswhyUSdataispresentedinthisreport.RecentUSreportsshow,forinstance,thatduringthemostrecentreportingyear(2013–2014),ofthe42drinkingwater–associatedoutbreaks,theopportunisticpathogenLegionellawasassociatedwith57%oftheoutbreaksandallofthereported13deaths(Benedict et al., 2017). The evidence shows this to be in linewith trends and not anisolatedevent,withfiguresshowingthatinthepastdecadeopportunisticpathogenshave


overtakenotheragentsasthemajorcauseofwaterbornediseaseoutbreaksassociatedwithdrinkingwaterintheUS(Figure6-1).

Figure 6-1. Aetiology of 885 drinking water–associated outbreaks, by year —UnitedStates,1971–2012.USCentresforDiseaseControl(Beeretal,2015).

Insummary,foropportunisticpathogens,theriskassessmentratedtheresidualriskasrare in likelihood but captured the fact that, being fatal, opportunistic pathogens canpresentmajorlocalisedconsequences.Thisriskneedscarefulevaluationasisdiscussedbelowtodecidewhetherthecontrolsareadequatesuchthattheriskisacceptable.

Itisnotedthatshortdeadlegsalreadyexistwithinexistingplumbingsystems,andmanydon'thaveanybackflowpreventionorlengthlimitation,asproposedwiththeFPAA101DSpecification.Itisprobablethat,onbalancetheriskisnotsignificantlyhigherthantherisk associated with conventional plumbing systems. The risk associated withconventional plumbing systems is currently accepted and plumbing systems are notoperated as zero or near zero risk systems.Nonetheless, given the high consequencerating for this risk, the it requires special deliberation in evaluating the FPAA101DSpecification.Theseriskswereconsideredinmoredepthinsection7.1.3ofthisreport.

6.5.5 SummaryofrisksrelatingtotoxigenicchemicalsToxigenicchemicalhazardscanarisefromplumbingfittings,(e.g.copper,lead,cadmium,chromiumorplastics),leachedand/orcorrodedfromplumbingdeadlegs.Overtimetheconcentrationsofchemicalsubstancesthatmayleachfrompipes,orformincorrosionproducts and then leach into thewater, could readily reach levels that exceedADWGguideline values in deadlegs.More likely thannot one ormore chemicalswill exceedADWGguidelinevalues in stagnantwater residing inplumbingpipesand fittings thatcomplywiththeWaterMarkcertificationschemeandAS/NZS4020:2005.Australian/NewZealand Standard. Testing of products for use in contact with drinking water. Thesestandardsandcompliancesystemsreducetheriskoftoxigenicchemicalsbeingpresentinplumbingbutonlyprovideahighlevelofassuranceofsuchprotectionforpipesandfittingsthatareadequatelyturnedover.Thatis,WaterMark-compliantplumbingwillbelesssubjecttocorrosionandleachingthanotherplumbing,butdeadlegscanstillpermit


contamination to arise to levels abovehealth-relatedguideline values, albeit to lowerlevelsthanwouldarisefromplumbingthatwasnotWaterMarkcompliant.

ItwasconsideredthattheproposedFPAA101DSpecificationwillprovidethepotentialfor chemicals to reach concentrations of concern within deadlegs resulting in apotentiallyhighrisk.

ThecontrolsproposedwithintheFPAA101DSpecification(summarisedinsection6.5.1)meanthatthelengthofdeadlegpipeworkisminimsed,theriskofbackflowisreduced,andthewaterrunningtothetoiletwillprovideahydraulicbufferofwaterthatisturnedoverbetweenthedeadlegpipeworkandthepotableplumbingsystem.Thecontrolswillreduce, but not eliminate, the risk of chemicals reaching levels that exceed guidelinevaluesbeingingested.However,withsuchshortdeadlegs,verylimitedexposureswouldoccur, for isolated events, and typically with water that is evidently contaminatedthrough colour, taste, particulates, or odour, and unlikely to be consumed in largequantities, and not foreseeably consumed for prolonged periods. The latter point isimportant since it renders the risks relating to toxigenic chemicals self-limiting. Forinstance,elevatedmetalsinstagnantwaterwouldtypicallybeevidentasbothcolourortaste.

In practice, whilst short term exceedances of chemical guideline values might arise,exposureperiodswouldbe too short, exposures too infrequentandadverseaestheticpropertiestooself-limitingtopresentsignificanthealthconsequences.Itwasconcludedthat the risk is acceptable and is similar to, and at worst only incrementally andmarginally higher than, the risk associated with conventional plumbing systems.Therefore,risksposedbytoxicantswerenotconsideredfurther.

6.5.6 SummaryofrisksrelatingtoaestheticsAesthetic hazards in water can reach levels that cause perceptible taste, odour,particulatesandcolour.Overtimetheconcentrationsofsubstancesthatmayleach,forminbiofilms,orariseincorrosionproductscouldreachlevelsthataredetectableascolour,particles,tastesorodours.Asnotedinsection6.5.5,theWaterMarkcertificationschemeandAS/NZS4020:2005.Australian/NewZealandStandard.Testingofproductsforuseincontactwith drinkingwater only assure full protection for pipes and fittings that areadequatelyturnedover.Thatis,WaterMark-compliantplumbingwillbelesssubjecttocorrosionandleachingthanotherplumbing,butdeadlegscanstillpermitcontaminationtoarisetolevelsaboveaestheticguidelinevalues,albeittolowerlevelsthanwouldarisefromplumbingthatwasnotWaterMarkcompliant

ItwasconsideredthattheproposedFPAA101DSpecificationwillprovidethepotentialfor substances to reach discernible concentrations within deadlegs resulting in apotentiallymoderaterisk.

ThecontrolsproposedwithintheFPAA101DSpecification(summarisedinsection6.5.1)meanthatthelengthofdeadlegpipeworkisminimsed,theriskofbackflowisreduced,andthewaterrunningtothetoiletwillprovideahydraulicbufferofwaterthatisturnedoverbetweenthedeadlegpipeworkandthepotableplumbingsystem.Thecontrolswillreduce, but not eliminate, the risk of substances reaching levels that are discernible.However,with such short deadlegs, very limited exposureswould occur, for isolatedevents,andnotforeseeablyforprolongedperiods.


In practice, whilst short term exceedances of aesthetic guideline values might arise,exposureperiodswouldbetooshort,andinstancestooinfrequent,topresentsignificantconsequences.Furthermore, therisk isacceptableand issimilar to,andatworstonlyincrementally and marginally higher than, the risk associated with conventionalplumbing systems. Therefore, risks posed by aesthetic hazards were not consideredfurther.


7 Discussionoffindingsoftheriskassessment

7.1 PrioritiesandlimitinghazardsThere were low, moderate and high consequence hazards identified in the riskassessment.However,inpractice,thecontrolsrequiredtoreducetherisksfromhigherconsequencehazardstoanacceptablylowlevelarelikelytomorethanadequatelyreducetherisksposedbylessconsequentialhazards.Therefore,thefocusofanassessmentoftheFPA101DSpecificationshouldbeonensuringthatthecontrolsaimedatmanagingthehigherconsequencehazardsareadequate.Theprinciple is that ifhigher, limiting,risks are adequately controlled, then the lower risks will, by default, be adequatelymitigated, provided themitigationmeasures identified are relevant toboth groupsofrisks.SincetherecommendedcontrolsproposedundertheFPAA101DSpecification(assummarised in section 6.5.1) are largely physical and hydraulic in nature, they willmitigatethevarioushazardsinsimilarways,whichsupportsrelyinguponmanagementofthelimitingandmostsignificanthazardandriskasbeingjustifiablythefocusofthisassessment.

7.1.1 LowconsequencehazardsThelowconsequencehazardsincludematerialsthatmaycauseaesthetic,butnothealth-related,consequences.Thesemaypresentascolour, taste,particulatesorodour.Suchriskscouldarguablybeacceptedonthebasisoftheirlow‘nuisance’ratherthanmoderateorhigh‘health’consequence.Inaddition,asnotedabove,hydraulicmitigationofhigher,limiting, risks should serve to mitigate these lower risks. Therefore, risks posed byaesthetic hazards were not considered further. Similarly, enteric pathogens are notdiscussedfurthersincetheywereconsideredtopresentnoadditionalriskfromadoptingtheFPAA101DSpecification.

7.1.2 ModerateconsequencehazardsModerate consequence hazards include chemical toxicants that will potentially bepresent above guideline values, but only for brief periods and infrequently, andmostlikely inassociationwithcolour,particulates, tasteorodour, limitingthe likelihoodoftheir ingestion.Therefore, suchrisks couldarguablybeacceptedon thebasisof theirmoderate consequence. In addition, as noted above, hydraulic mitigation of higher,limiting,risksshouldservetomitigatethesemoremoderaterisks.Therefore,risksposedbytoxicanthazardswerenotconsideredfurther.

7.1.3 Highconsequence,limitinghazardsTheonlyhighconsequencehazardpresentedrelatestoopportunisticpathogens.ThekeytoensuringthattherisksassociatedwiththeFPAA101DSpecificationareacceptableistoreducetheriskposedbyopportunisticpathogenstoanacceptablylowlikelihood.Thecontrols required to reduce this opportunistic pathogen risk to an acceptably lowlikelihood are likely to more than adequately reduce the risk posed by the lessconsequentialhazards,notedabove.


7.2 MostsignificantandlimitinghazardouseventsTherisksposedbybackflowwereidentifiedasthemostsignificantofthefailuresmodesandhazardouseventsassessed.Therefore,akeyquestionrelatingtotheacceptabilityoftheproposedFPAA101DSpecificationrelatestotheacceptabilityof thebackflowrisk.From a design, installation, inspection and maintenance perspective, the focus infinalisingtheFPAA101DSpecificationshouldbeonspecifyingthecontrolsrequiredtoadequatelymitigatethatbackflowrisk.

7.2.1 CurrentlyacceptedrisksAkeyfindingwhenconsideringthebackflowriskwasthatthereareroutinelymanylowturnover and zero turnover deadleg pipes and fittings in existing plumbing systems.Therefore,theFPAA101DSpecificationpresentsriskssimilarinnatureandmagnitudetothose associated with conventional plumbing systems. Potentially the FPAA101DSpecificationpresentslowerrisksthanconventionalplumbingsystems.Thereasonforthis difference is that conventional plumbing systemsoften include low turnover anddeadlegpipesandfittingswithoutthebackflowpreventionorlengthlimitationcontrolsgiven in the proposed FPAA101D Specification (as summarised in section 6.5.1).Examplesofsuchexistingpipesandfittingsmayinclude:

• Entirebuildingplumbingsystemsininfrequentlyusedproperties,e.g.short-termlease properties or holiday units thatmay go unused for someweeks or evenmonthsbetweenuses.

• Entirebuildingplumbingsystemsinsitesthatclosedownforprolongedperiods,e.g. seasonal holiday buildings, educational establishments, commercial orindustrialfacilities,thatmaynotbeinuseformanyweeksatatime.

• Infrequently used plumbing lines, taps and fittings supplying points outsidebuildings,suchashosefittingstotapsthatareseldomused,orseasonallyusedtapsandfittings,suchasoutdoorshowers,irrigationlinesorpoolfillingpoints.Thesemaybeconnected to longhosereelsand irrigation linesof considerablelength.

• Infrequentlyusedplumbinglines,tapsandfittingsinsidebuildings,suchasroomswithensuitesthatarenotinuse,abandonedshowersorbathsthatareleakingandnotinuse,ortapssupplyingappliancesthatarenolonger,orseldom,used.

• Infrequently plumbing lines and safety fittings, such as eye wash stations oremergencyshowers.

• Abandonedandpluggedpipelengthsarisingwhenpropertiesaremodifiedandpipes are left in place and remain plumbed into the water supply plumbing,formingadeadlegwithnoturnover.

Theuseofanon-testabledualcheckvalve,asproposedintheFPAA101DSpecification,reducestheriskofcontaminationflowingbackintothepotablenetworktolessthantheriskassociatedwithexistingdeadlegs,suchasthoselistedabove.However,astowhethertherisk is lowenough tobeacceptablecomesdowntoprofessional judgement in theabsenceofobjectiveevidencefromfieldexperience.SincetheFPAA101DSpecificationisnew,thereisnotalongstandingfieldsitetousefordirectassessment.

Withinexistingplumbingarrangements, larger fire systemsmustbeprotectedwithatestabledevice.Therefore,ifaspecialcaseismadeforthesmallerfiresystemstoonlybe


protectedbyanon-testabledualcheckvalve,thensomecriterianeedtobeplacedaroundthatinordertojustifyforegoingthetestabledevice.Thosecriteriacaninclude:

• Limitingthesizeandscaleofthebuilding(25minheightisproposedinthecaseoftheFPAA101DSpecification);

• setting a maximum length on deadlegs (3 m is proposed in the case of theFPAA101DSpecification);and

• the need for the line feeding the deadlegs that lead to the sprinklers to beconnectedtoasystemthatdoesflow(e.g.atoiletisproposedinthecaseoftheFPAA101DSpecification).

Thefocusofthisassessmentcomesdowntodecidingwhethertheseadditionalcontrolsareadequateand,ifnot,whatcontrolswouldbesufficient.Assessmentoftheadequacyofthosecontrolscanbefocusedonconsiderationofthemostsignificant,orlimiting,risks,inthiscase,therisksposedbyopportunisticpathogens,asdiscussedinthefollowingsub-section.

7.3 RelativeriskassessmentrelatingtoopportunisticpathogensAsnotedabove(section7.2.1),currentlyacceptedrisksincludedsomelow,orevenno,turnovertapsandfittingswithinbuildings.

Installingadditionalnoturnoverfittings,inthiscasefiresprinklerfeeddeadlegs,withinbuildings,asrecommendedundertheFPAA101DSpecification,unquestionablyincreasestherisksassociatedwithopportunisticpathogens.Theriskisincreasedbyvirtueoftheincreasedsurfaceareapresentedandincreasedlengthofpipeinstalledthatisintendedtobesubjecttonosignificantturnoverrelativetothenormalsituation.

However, from a relative risk perspective, the FPAA101D Specification as proposedpresentsanincrementalincreaseinriskratherthanastepchangeinriskoranewrisk.Thatis,theextentoflowandnoturnoverplumbingpipesandfittingswillincreaseabovecurrentlevelsundertheFPAA101DSpecification,butnocompletelynewandcurrentlynon-existentcategoryofriskisintroduced.

Furthermore,asnotedinsection6.5.1,additionalcontrolsareproposedintheFPAA101DSpecification for thewet sprinkler systems: a dual check valve non-testable backflowdevice;turnoveroftheplumbinglinefeedingthedeadlegsusingthetoiletflushingsupplyline; and limiting deadlegs to at most 3 m linear length. These controls significantlyreducetheextentoftheincrementalincreaseinriskfrombackflow.Thekeyquestiontobe addressed is: are those controls sufficient, and is the incremental increase in riskacceptable?Thisquestionisconsideredinthefollowingdiscussion.

7.4 ProfessionalassessmentoftheriskfromopportunisticpathogensAquantitativeassessmentofriskisdiscussedbutwasnotundertakenforthereasonsexplainedbelow(7.5).Therefore,theexerciseofprofessionaljudgement,basedontheobjectiveevidence,hasbeenusedtoinformaprofessionalopinionontheacceptabilityandadequacyofthecontrolsproposedundertheFPAA101DSpecification.Inordertosupportsuchjudgement,existingevidencewasassembledandreviewed.


7.4.1 LengthofdeadlegsOne importantcontrol identified in theFPAA101DSpecification is todrawa linewithrespecttohowlongdeadlegscanbe.Whilstdrawingsuchlinesbasedonevidenceandtheoreticalriskcanbedifficult,andtheremaybevaryingviewsonwherethatlineshouldbedrawn,thereislikelytobeconsensusthatsomeworkingupperlimitonthelengthofdeadlegsshouldbeset.

Fortunately,theproposedFPAA101DSpecificationhasbeenabletodrawdirectlyfromevidencefromresearchcarriedoutinNZ(Soja,2006)insettingalimitonthelengthofdeadlegs.Inthatrespect,theFPAA101DSpecificationcanbeconsideredtobeevidence-based. That research studied microbial and chemical water quality in domestic firesprinklersystemsoveroneyear.Thereportconcluded:

“Itwasfoundthatthemicrobialqualitywouldnotbehazardoustohealthwhererangepipedeadlegswereupto4.5metres longforawatersupplyofequalorbetterqualitytothatusedinthisresearch.Fromthisitisrecommendedthatdeadlegs up to 3 metres could be used in Combination Domestic Fire SprinklerSystems.”

ItisnotedthatoneofthecontrolsrecommendedintheFPAA101DSpecificationistolimitlinearlengthsofdeadlegsto3m(section6.5.1).Thiscontrolappearstobeimportantanddrawing the line based on the Soja (2006) study seems appropriate given thatmorerecentevidencehasnotbeenidentifiedthatsuggeststhatthisrecommendationshouldbe modified. Therefore, it is considered that the recommendation in the FPAA101DSpecificationthatdeadlegsbelimitedinlengthto3mcanbeaccepted.

7.4.2 ThestandardofbackflowpreventionrequiredTheBackflowPreventionAssociationofAustralia (BPAA),QueenslandHealthandWABuildingandEnergy,PlumbingTechnicalReferenceGrouphaveallstronglyquestionedthe idea that sprinkler systems connected to the potable water supply under theFPAA101D Specification can avoid the need to protect those systems using testablebackflow prevention devices. Therefore, a key decision in finalising the FPAA101DSpecification relates to whether the backflow risks arising under the FPAA101DSpecificationarrangementsareofamagnitudethatthebackflowprotectioncanjustbebasedonadualcheckvalvenon-testablebackflowdeviceascurrentlyproposed(section6.5.1),orwhetheranaccessibleandtestabledeviceisrequired.

Asimilarquestionwasconsidered inastudy fromtheUSbyNFPA(2009).Thestudyconcludedthatitwasnotuncommonforresidentialcombinationfiresprinklersystemstobesuppliedwithouttestablebackflowpreventiondevices,includinginwarmclimatesouthern US states. Problems with backflow from such residential combination firesprinkler systemswerenot commonly reportedunder routine, field conditions in theresearch, and backflow prevention was not considered necessary in many of thejurisdictionssurveyed.

AnearlierUSstudy(Duranceauetal,1999)foundthat,undersimulatedconditionsusingasinglecheckvalve,unrestricteddeadleglengthandnotoiletconnectionflowthrough,backflow from sprinkler system deadlegs was observed, with a range of hazardouschemicals and microbial indicators of stagnant water being detected. However, theauthorsconcludedthattherisksposedbysucheventsweretoolowtojustifyretrofitting


backflowpreventiontoexistingresidentialcombinationfiresprinklersystemsgiventherelativelylowrisksimpliedcomparedtheveryhighrisksassociatedwithfires.

Ausefulconsiderationinmakingadecisiononthelevelofbackflowpreventionrequiredin the FPAA101D Specification is ‘relative risk’. Existing plumbing systems containdeadlegs that are not routinely protected by any backflow prevention system. TheFPAA101D Specification has not proposed to use a testable backflow device, such aswouldberequiredforamediumhazardconnection.ButtheFPAA101DSpecificationhasproposedtouseanon-testabledualcheckvalvebackflowdevice.Thispositionappearsto be a pragmatic solution that reduces the risk of backflow associatedwith the firesprinklersystemtolowerthanthatassociatedwithotherdeadlegs.Inaddition,thewatersupplyingthetoiletlineprovidesahydraulicbufferofwaterthatislikelytoberegularlyturnedoversuchthatonlyquitesignificantbackfloweventswoulddrawwaterintothepotableplumbingsystemifthebackflowdevicefailed.

Onthebasisoftheriskbeinglowerthanthatacceptedfromexistingdeadlegs,(bothdueto the presence of some existing deadlegs being longer than 3 m and their lackingbackflowprotection),therecommendationintheFPAA101DSpecification,thatdeadlegsfeedingsprinklersbeprotectedfromthepotablenetworkbybothatoiletplumbingfeedlineandanon-testabledual checkvalvebackflowdevice, is considered sufficient riskmitigation.

7.4.3 Cross-connectiontothesprinklerlineQueensland Health have raised concerns about cross-connections. Cross-connections,mis-connectionsandinadvertenttap-inscouldoccurandthenconnecttheplumbinglinesfeedingthetoiletandsprinklerplumbingdeadlegstothepotablewater.Theconnectedtaps and fittings could be influenced by the deadlegs supplying the sprinklers. Suchconnectionsarelikelytooccurwithoutanybackflowdeviceinplaceandcouldaffectboththeimmediateconnectioninthecaseofsingletap-insaswellasotherconnectionsinthecase of full cross-connections. During building commissioning, existing full cross-connectionswouldbe readily identifiedwhendomesticor sprinkler systemplumbingwereattemptedtobe isolatedandtested,but ifsucheventsoccurredovertimewhenplumbingworkswerecarriedoutsuchconnectionsmightbemissed.

Therelevanthazardsandrisksassociatedwithcross-connectionsarelargelythesameasthose associatedwith backflowwith the exception of the fact that cross-connectionscouldbecontinuous,whereasbackfloweventsarelikelytobesporadic.

ItisnotedthatoneofthecontrolsrecommendedintheFPAA101DSpecificationistolimitlinearlengthsofdeadlegsto3mandtominimiseconnectionstoothersystems(section6.5.1).Whilst a cross-connection of some sort is reasonably foreseeable, (with cross-connectionsreportedmanytimesperyeararoundAustraliainrecycledwaternetworksfor instance),with such a limited extent for the deadlegs, the risks arising from suchcross-connectionswouldberelativelylow,andnohigherthanthatassociatedwiththeriskofexistingdeadlegsthatarepresentinplumbingsystems.Suchcross-connectionscanalreadyoccurbetweenthepotableplumbingandconventionalfiresprinklersystems,recycledwater,irrigationwater,roof-harvestedrainwaterandothernon-potablewaterplumbingsystems.


On thebasis of the cross-connection riskbeing similar to that accepted fromexistingdeadlegs,(andlowerthantheriskassociatedwithcross-connectionstoexistingdeadlegslonger than 3 m), no additional modifications to the FPAA101D Specification areconsiderednecessary.However,asrecommendedbyQueenslandHealth,somefurtheremphasis on plumbing inspections and signage iswarranted for properties installingcombined sprinkler systems under the FPAA101D Specification. As a benchmark,plumbing regulators often require higher priority and frequency inspections andenhancedsignagefordualreticulationplumbingsystems.

7.4.4 ToiletflushingItispossiblethatoccasionalsplashesordropletsoftoiletflushingwatercouldbeinhaledleading to exposure to opportunistic pathogenswithin those droplets. However, it iswidelyunderstoodinthecommunitythattoiletscontainhazardouswaterduetotheirprimarygoalofcapturingfaecalmatter,vomitandblood.Bydesign,toiletsaredesignedtocapture,andnotsprayorspreadwater.Thecommunityhasgoodawarenessof theneednottoingestorcontacttoiletwater.Inaddition,toiletsarenotsourcesofsignificantquantitiesofaerosolcontaminationandtheyareinparticularnotsourcesofsmallsize-classaerosols.Therefore,whilsttracesofopportunisticpathogensmightfindtheirwayfromfiresprinklerfeeddeadlegstotoiletsviathetoiletflushingline,therisksfromsuchtransfers to health would be minimal due to the absence of a significant exposurepathway.

Insummary,exposurestotoilet flushingwaterwouldbeverylowduetotoiletsbeingwidely understood to contain potentially contaminatedwater. In addition, by design,toilets are not sources of small size-class aerosols meaning that they are not a highlikelihoodsourceofopportunisticrespiratoryinfections.Therefore,nospecialcontrolsorrequirementsareconsiderednecessaryintheFPAA101DSpecification.

7.5 Quantitativeriskassessment

7.5.1 ToxicologicalriskassessmentA toxicological risk assessment (TRA) can be used to set quantitative limits on risksassociatedwithchemicalsagainstdefinedtargets.Inthiscase,aTRAwasnotundertakentosupportthisassessment.Itwasconcludedfromthequalitativeriskassessmentthatrisks from chemicals pathogens were not significant given that any exposures tohazardous chemicals arising from plumbing leaching and corrosion would be rare,sporadic, short indurationand typicallyassociatedwithelevatedcolour,particulates,tasteorodour.Noadditionalassessmentwasrequiredsincetheresultingresidualriskwasconsideredacceptable.

7.5.2 QuantitativemicrobialriskassessmentAquantitativemicrobialriskassessment(QMRA)canbeusedtosetquantitativelimitsonrisksassociatedwithpathogenagainstdefinedtargets.Inthiscase,aQMRAwasnotundertakentosupportthisassessment.

QMRAisoftenappliedtoassessingrisksfromentericpathogensandwouldhavebeencompletedwereentericpathogensconsideredpotentiallyofsignificantrisk.However,thequalitative risk assessment concluded that risks fromentericpathogenswerenot


significantandwerenohigherthanexistingrisksfromconventionalplumbingsystems.Noadditionalassessment,includingaQMRA,wasconsideredrequiredsincetheresultingresidualriskwasconsideredacceptable.

QMRA can be applied to assessing risks from opportunistic pathogens. However, atpresent,suchtechniquesareofverylimitedvalueinassessingrisksfromopportunisticpathogens(BenthamandWhiley,2018).ThemajorrisksofconcerninthereviewoftheFPAA101DSpecificationrelatedtoopportunisticpathogensandnotentericpathogens.Therefore,theabovequalitativeriskassessmentwasusedasthebasistoconsidertheacceptabilityofrisksandrequirementsofcontrols,ratherthanaQMRAthatwouldbetoouncertaintoprovideahelpfulcontributiontodecision-making.


8 ConclusionImplementingtheFPAA101DSpecificationdoesnotpresentnewclassesorcategoriesofrisks from enteric pathogens, opportunistic pathogens, toxicants or aesthetic hazardsthat arenot alreadypresent in existingpotableplumbing systems.All suchplumbingsystemshavethepotentialtocontainsomedeadlegswithverylowornoturnoverthatcaninturninfluencethepotablewatersystemviabackfloworcross-connection.

Relative to a building with no fire sprinkler system, implementing the FPAA101DSpecification adds the potential for some additional deadlegs and, hence, anincrementallyincreasedriskfromsuchhazards.However,theFPAA101DSpecificationincludesanumberofcontrolstoreducethoserisksforthefiresprinklersystemstolowerthanthelevelsassociatedwithotherdeadlegs,namely:

• limitingdeadlegstomaximum3mlineallengths;• providingbackflowpreventionviaanon-testabledualcheckvalve;and• providingforahydraulicbarrierwiththeplumbinglinefeedingthesprinklerlines

beingturnedoverthroughconnectionviathetoilet.

Thepresenceofsuchcontrolsmeansthat incremental increases inexistingrisksfromhaving theFPAA101DSpecification fire sprinkler systems inplace is small relative toexistingriskspresentedbyexistingplumbingdeadlegsthathavenosuchprotection.

Relativetoabuildingwithaconventionalseparatefiresprinklersystem,implementingthe FPAA101D Specification is likely to be present lower deadwater volumes in lessdeadlegs than those presented by conventional separate fire sprinkler systems. Theconventional separate fire sprinkler systems are separated from the potable watersystembyatestablebackflowpreventiondevicewhereastheFPAA101DSpecificationincludesnon-testabledualcheckvalveprotection.However,theFPAA101DSpecificationincludesadditionalcontrolsnotpresentinconventionalseparatefiresprinklersystems,namely:

• limitingdeadlegstomaximum3mlineallengths;• providingforahydraulicbarrierwiththeplumbinglinefeedingthesprinklerlines

beingturnedoverthroughconnectionviathetoilet;and• allcomponentsbeingWaterMarkcompliantandso lesspronetocorrosionand

leachingoftoxicantsandaestheticallyobjectionablehazards.

Thepresenceofsuchcontrolsmeansthatifthereisanincreaseinriskitisverysmallrelative to theexistingriskpresentedbyconventional separate firesprinklersystemsthatareonlyprotectedbyasinglebackflowprotectionbarrier,albeitatestabledevice.

In summary, whilst there are risks arising from implementing the FPAA101DSpecification, those risksarecomparable toacceptedrisks found inpotableplumbingsystems,includingthosewithnofiresprinklersystemsbuthavingsomedeadlegs,andthosewithconventionalseparatefiresprinklersystems.Thefiresafetyrisksfromhavingno automatic fire sprinkler systems arewell established. Therefore, if the FPAA101DSpecificationencouragestheuptakeoffiresprinklersystemswheretheymightotherwisenotbeimplemented,e.g.whereinstallingconventionalseparatefiresprinklersystemsisconsideredimpracticalorcostprohibitive,theFPAA101DSpecification,withitsmultiplecontrols toprotectpotablewaterquality,presentsanacceptable,practical solution toincreasetheupdateoffiresprinklersystems.


9 ReferencesAS/NZS 4020:2018. Australian/NewZealand Standard. Testing of products for use incontact with drinking water. Standards Australia and Standards New Zealand, lastupdated29June2018.

BeerKD,GarganoJW,RobertsVA,etal.SurveillanceforWaterborneDiseaseOutbreaksAssociatedwithDrinkingWater—UnitedStates,2011–2012.MMWRMorbMortalWklyRep2016;64:842-848.

Benedict KM, ResesH, VigarM, et al. Surveillance forWaterborneDiseaseOutbreaksAssociatedwithDrinkingWater—UnitedStates,2013–2014.MMWRMorbMortalWklyRep2017;66:1216–1221.

BenthamR,Surman-LeeS,LeeJV,BriandE.andVandeKooijD.,2007.Potablewaterandin-building distribution systems. Chapter 4. Pp57-68 Legionella and the presence oflegionellosis.EdsBartramJ,ChartierY,LeeJV,PondK.andSurman-Lee,S.WorldHealthOrganizationISBN9241562974.

Bentham R. and Whiley H., 2018. Quantitative Microbial Risk Assessment andOpportunistWaterborne Infections – Are There TooManyGaps to Fill? InternationalJournalofEnvironmentalResearchandPublicHealth,15(1150):1-11.

CraunGFandCalderonRL,2001.Waterbornediseaseoutbreakscausedbydistributionsystemdeficiencies.JournalAWWA93(9):64-75.

Duranceau S., Pool J. and Foster J., 1999.Wet-pipe fire sprinklers andwater quality.Journal-AmericanWaterWorksAssociation,91(7):78-90.

FPAAustralia,2018.FPAA101DAutomaticFireSprinklerSystemDesignandInstallation–DomesticWaterSupply,Publicationdraft,FireProtectionAssociationAustralia,1May2018,asamendedasaworkingdraft,21August2018.

NFPA,2009.IntegrationofResidentialSprinklerswithWaterSupplySystemsASurveyof TwentyU.S. Communities, Newport Partners, LLC for TheNational Fire ProtectionAssociation,September2009.

NHMRC/NRMMC(2011)AustralianDrinkingWaterGuidelines,NationalWaterQualityManagementStrategy,Version3.4,NationalHealthandMedicalResearchCouncil andNaturalResourceManagementMinisterialCouncil,lastupdatedOctober2017.

SojaE.,2006.DomesticFireSprinklerSystems–ReportOnWaterQuality,ReliabilityandApplicationtoOtherProperty.BRANZReport,Judgeford,NewZealand.

Table4.Riskassessmentsummarytable

Process Step(from flow diagram)

Contaminants('hazards')

How can the hazard be introduced or exacerbated as a result of the proposed FPAA101D standard?('hazardous events') Li

kelih

ood

(A-E

)Se

verit

y(1

to 5

)

Maximum Risk (no specific controls)

Controls ('preventive measures') that are proposed in FPAA101D

Resid

ual

Like

lihoo

d(A

-E)

Resid

ual

Cons

eque

nce

(1 to

5)

Residual Risk (with proposed controls)

Notes on basis of the scoring Uncertainty Acceptability of risk Follow up actions

Deadlegs feeding sprinklers

Enteric pathogens including bacteria (e.g. Salmonella spp. and Campylobacter spp.), parasites (e.g. Cryptosporidium spp.) and viruses (e.g. norovirus).

These pathogens are only present in water contaminated with faecal matter or putrescing food waste. They are no more likely to be present in deadleg fire sprinkler lines than in any other plumbing since they do not typically persist or grow in potable water, but rather they gradually die off over time.

If the water backflows into the potable system persons may be exposed via ingestion.

E 1 Low E 1 Low Enteric pathogens are primarily derived from fresh faecal matter and are not specifically of concern in deadlegs or pipework. The relatively low nutrients found in potable water do not typically support the proliferation of enteric pathogens even in the absence of a disinfectant residual, e.g. in deadlegs. Significant concentrations of enteric pathogens capable of resulting in significant consequences are not presented by the proposed standard, even without any dedicated controls, resulting in a low maximum and residual risk.

High level of confidence. The source, fate and transport of enteric pathogens in water is well understood.

The risk is acceptable and is no higher than the risk associated with conventional plumbing systems that already include low turnover plumbing lines and fittings, often without backflow prevention or length limitations on deadlegs.

None.


Opportunistic pathogens including bacteria such as Legionella spp. and amoebae such as Naegleria spp. and Acanthamoeba spp.

These pathogens are present naturally in water. They proliferate to levels of concern in stagnant water, particularly where they can form biofilms (e.g. on pipe surfaces), in water lacking a disinfectant residual (e.g. in deadleg pipes) and in warm water (e.g. in the 20 to 50˚C temperature range).

If the water backflows into the potable system persons may be exposed, e.g. via inhalation, eye, nose or skin contact.

C 4 Very High 2.5.3.2.1 Sprinkler pipework deadlegs/stagnant water. The sprinkler system pipework shall be designed and installed so that no part of any pipework containing stagnant water is more than 3 m in lineal length.

2.5.3.2.3 Backflow prevention. A backflow prevention device in accordance with Clause 5.3.2 shall be installed on the sprinkler pipework...

2.5.3.2.4 Connection to toilets. All toilets shall be connected and supplied directly from the sprinkler system...

E 4 High The proposed standard will increase the potential for proliferation of opportunistic pathogens by providing increased surface area and deadlegs resulting in a potentially very high risk.

The controls proposed mean that the length of deadleg pipework is minimsed, the risk of backflow is reduced, and the water running to the toilet will provide a buffer of water that is turned over between the deadleg pipework and the potable plumbing system. The controls will reduce, but not eliminate, the risk of opportunistic pathogens diffusing or being drawn back into the potable plumbing system. The risk is further mitigated by the likelihood that backflow of stagnant water would be associated with elevated colour, particulates, taste or odour. The resulting residual risk remains rare in likelihood but could have a major localised impact if the controls fail.

Moderate level of confidence. The presence of deadlegs similar to those proposed is not uncommon in existing plumbing systems so the risk is expected to be rare. However, opportunistic pathogen infections from plumbing systems do occur at times.

The risk needs discussion to decide whether it is acceptable. It is noted that short deadlegs already exist within existing plumbing systems, and many don't have any backflow prevention as proposed with the FPAA101D standard. It is probable that on balance the risk is not significantly higher than the risk associated with conventional plumbing systems. Nonetheless, given the high consequence rating for this risk, the risk requires special deliberation.

Single out this risk for special consideration.


Toxigenic chemical hazards from plumbing fittings, e.g. copper, lead, cadmium, chromium or plastics leached from plumbing deadlegs.

Over time the concentrations of toxicants that may leach from pipes, or form in corrosion products, could reach levels that exceed guideline values in deadlegs. The AS/NZS 4020:2018 requirements only assure protection for pipes that are adequately turned over.

If the water backflows into the potable system persons may be exposed via ingestion.

B 3 High 2.5.3.2.1 Sprinkler pipework deadlegs/stagnant water. The sprinkler system pipework shall be designed and installed so that no part of any pipework containing stagnant water is more than 3 m in lineal length.



D 2 Low The proposed standard will increase the potential for toxicant to reach concentrations of concern by providing increased surface area and deadlegs resulting in a potentially high risk.

The controls proposed mean that the length of deadleg pipework is minimsed, the risk of backflow is reduced, and the water running to the toilet will provide a buffer of water that is turned over between the deadleg pipework and the potable plumbing system. The controls will reduce, but not eliminate, the risk of toxicants reaching levels that exceed guideline values being ingested. With such short deadlegs, very limited exposures would occur, for isolated events, and typically with water that is evidently contaminated through colour, taste, particulates, or odour, and unlikely to be consumed in large quantities, and not foreseeably consumed for prolonged periods. In practice, whilst short term exceedances of chemical guideline values might arise, exposure periods would be too short, and exposures too infrequent, to present significant health consequences.

High level of confidence. The leaching and corrosion of substances from plumbing fittings is well understood.

The risk is acceptable and is at worst only marginally higher than the risk associated with conventional plumbing systems that already include low turnover plumbing lines and fittings, often without backflow prevention or length limitations on deadlegs.

None.


Aesthetic hazards including elevated colour, particulates, taste or odour, leached from, or arising in, plumbing deadlegs.

Over time the concentrations of substances that may leach, form in biofilms, or arise in corrosion products could reach levels that are detectable as colour, particles, tastes or odours. The AS/NZS 4020:2018 requirements only assure protection for pipes that are adequately turned over.

If the water backflows into the potable system persons may notice.

C 2 Moderate 2.5.3.2.1 Sprinkler pipework deadlegs/stagnant water. The sprinkler system pipework shall be designed and installed so that no part of any pipework containing stagnant water is more than 3 m in lineal length.



D 2 Low The proposed standard will increase the potential for substances to reach discernible concentrations by providing increased surface area and deadlegs resulting in a potentially moderate risk.

The controls proposed mean that the length of deadleg pipework is minimsed, the risk of backflow is reduced, and the water running to the toilet will provide a buffer of water that is turned over between the deadleg pipework and the potable plumbing system. The controls will reduce, but not eliminate, the risk of substances reaching levels that are discernible. With such short deadlegs, very limited exposures would occur, for isolated events, and not foreseeably for prolonged periods.

High level of confidence. The leaching and corrosion of substances from plumbing fittings is well understood.

The risk is acceptable and is only marginally higher than the risk associated with conventional plumbing systems that already include low turnover plumbing lines and fittings, often without backflow prevention or length limitations on deadlegs.

None.

WaterqualityriskassessmentoftheFPAA101DSpecificationfiresprinklersystem.17thOctober2018.ByDanDeere. Page32of33

Process Step(from flow diagram)

Contaminants('hazards')

How can the hazard be introduced or exacerbated as a result of the proposed FPAA101D standard?('hazardous events') Li

kelih

ood

(A-E

)Se

verit

y(1

to 5

)

Maximum Risk (no specific controls)

Controls ('preventive measures') that are proposed in FPAA101D

Resid

ual

Like

lihoo

d(A

-E)

Resid

ual

Cons

eque

nce

(1 to

5)

Residual Risk (with proposed controls)

Notes on basis of the scoring Uncertainty Acceptability of risk Follow up actions

Domestic plumbing line

Enteric pathogens including bacteria (e.g. Salmonella spp. and Campylobacter spp.), parasites (e.g. Cryptosporidium spp.) and viruses (e.g. norovirus).

These pathogens are only present in water contaminated with faecal matter or putrescing food waste. They are no more likely to be present in deadleg fire sprinkler lines than in any other plumbing since they do not typically persist or grow in potable water, but rather they gradually die off over time.

If the water is cross-connected into the potable system persons may be exposed via ingestion.

E 1 Low 2.5.3.2.1 Sprinkler pipework deadlegs/stagnant water. The sprinkler system pipework shall be designed and installed so that no part of any pipework containing stagnant water is more than 3 m in lineal length.

2.5.3.3 Connection to other systems. Other than detailed in this Specification, no part of the sprinkler system shall be connected to any other service including drencher systems, wall wetting systems, hose reels or the like.

E 1 Low Enteric pathogens are primarily derived from fresh faecal matter and are not specifically of concern in deadlegs or pipework. The relatively low nutrients found in potable water do not typically support the proliferation of enteric pathogens even in the absence of a disinfectant residual, e.g. in deadlegs. Significant concentrations of enteric pathogens capable of resulting in significant consequences are not presented by the proposed standard, even without any dedicated controls, resulting in a low maximum and residual risk.

High level of confidence. The source, fate and transport of enteric pathogens in water is well understood.

The risk is acceptable and is no higher than the risk associated with conventional plumbing systems that already include low turnover plumbing lines and fittings, often without special cross-connection controls or length limitations on deadlegs.

None.


Opportunistic pathogens including bacteria such as Legionella spp. and amoebae such as Naegleria spp. and Acanthamoeba spp.

These pathogens are present naturally in water. They proliferate to levels of concern in stagnant water, particularly where they can form biofilms (e.g. on pipe surfaces), in water lacking a disinfectant residual (e.g. in deadleg pipes) and in warm water (e.g. in the 20 to 50˚C temperature range).

If the water is cross-connected into the potable system persons may be exposed, e.g. via inhalation, eye, nose or skin contact.

E 4 High 2.5.3.2.1 Sprinkler pipework deadlegs/stagnant water. The sprinkler system pipework shall be designed and installed so that no part of any pipework containing stagnant water is more than 3 m in lineal length.


E 4 High The proposed standard will increase the potential for proliferation of opportunistic pathogens by providing increased surface area and deadlegs resulting in a potentially very high risk.

The controls proposed mean that the length of deadleg pipework is minimsed and the extent of pipework that could be connected to the potable plumbing system is minimised. The controls will reduce, but not eliminate, the risk of opportunistic pathogens diffusing or being drawn back into the potable plumbing system if a cross-connection occurs. The risk is further mitigated by the likelihood that the presence of stagnant water would be associated with elevated colour, particulates, taste or odour. The resulting residual risk remains rare in likelihood but could have a major localised impact if the controls fail.

Moderate level of confidence. The presence of deadlegs similar to those proposed is not uncommon in existing plumbing systems so the risk is expected to be rare. However, opportunistic pathogen infections from plumbing systems do occur at times.

The risk needs discussion to decide whether it is acceptable. It is noted that short deadlegs already exist within existing plumbing systems, and many don't have any cross connection prevention. It is probable that the risk is only marginally incrementally higher than the risk associated with conventional plumbing systems. Nonetheless, given the high consequence rating for this risk, the risk requires special deliberation.

Single out this risk for special consideration.


Toxigenic chemical hazards from plumbing fittings, e.g. copper, lead, cadmium, chromium or plastics leached from plumbing deadlegs.

Over time the concentrations of toxicants that may leach from pipes, or form in corrosion products, could reach levels that exceed guideline values in deadlegs. The AS/NZS 4020:2018 requirements only assure protection for pipes that are adequately turned over.

If the water is cross-connected into the potable system persons may be exposed via ingestion.

E 3 Moderate 2.5.3.2.1 Sprinkler pipework deadlegs/stagnant water. The sprinkler system pipework shall be designed and installed so that no part of any pipework containing stagnant water is more than 3 m in lineal length.


E 2 Low The proposed standard will increase the potential for toxicants to reach concentrations of concern by providing increased surface area and deadlegs resulting in a potentially high risk.

The controls proposed mean that the length of deadleg pipework is minimsed and the extent of pipework that could be connected to the potable plumbing system is minimised. The controls will reduce, but not eliminate, the risk of toxicants diffusing or being drawn back into the potable plumbing system if a cross-connection occurs. The risk is further mitigated by the likelihood that the presence of stagnant water would be associated with elevated colour, particulates, taste or odour. With such sho

water quality risk assessment of the fpaa101d specification fire … · 2019. 3. 13. · water...

Documents