API 521 SRU RF WHB Evaluation for Overpressure due to …pm-engr.com/wp-content/uploads/2016/03/2015-Brin... · Brimstone Sulfur Symposium API 521 SRU RF WHB Evaluation for Overpressure

Brimstone Sulfur Symposium API 521 SRU RF WHB Evaluation for Overpressure due to Tube Rupture

© Black & Veatch Holding Company 2015. All Rights Reserved. The Black & Veatch name and logo are registered trademarks of Black & Veatch Holding Company.

Porter McGuffie, Inc. and Black & Veatch 1 September 2015

API 521 SRU REACTION FURNACE WASTE HEAT BOILER EVALUATION FOR OVERPRESSURE DUE TO TUBE RUPTURE

DennisH.MartensPorterMcGuffie,Inc.

AlanD.Mosher

Black&VeatchCorporation

BrimstoneSulfurSymposiumVail,Colorado

September14‐18,2015Abstract

American Petroleum Institute (API) Standard 521, Pressure‐relieving and DepressuringSystems (current edition: 2014) is being revised to include information regarding SulfurRecoveryUnit (SRU)ReactionFurnaceWasteHeatBoiler (WHB) tube rupturedevelopedprocess side overpressure andother updates. Theproposedmodifications regarding theSRU Reaction Furnace WHB were first balloted in the spring of 2015. Once all of thechangesareballotedandapproved,aneweditionofAPI521willbeissuedin2019attheearliest.ThemodificationstoAPI521thatwereballotedstatethattheusershouldevaluatethe system to determine if the system can relieve a rate equivalent to double the crosssectional area of a single tube in the WHB without exceeding the corrected hydrotestpressure of the reaction furnace and other low‐pressure side equipment. The proposedmodifications suggest that Steady State analysis be completed first. If the steady stateanalysispredictsthatthecorrectedhydrotestpressurewillbeexceededitissuggestedthatothermethods such asDynamic analysis can be completed. The proposedmodificationsalsostatethatanalternativetodynamicanalysiswouldbetheapplicationofASMESectionVIII, Division 1 UG‐140 Overpressure Protection by System Design. The proposedmodifications do not provide any real guidance on how the analysis should be done orsuggestedscenariosandsequencesforthevariousanalysesthatshouldbecompleted.Thispaper focuseson the authors’ understandingof how theoverpressure analysis shouldbecompletedandprovidesaflowchartwithasuggestedsequencefortheanalysis.Thebulkofthis paper was originally developed and submitted to API for possible inclusion in themodifications to API 521 to provide guidance for engineers that are responsible forevaluation of tube ruptures in new and existing SRU reaction furnace WHBs. Afterconsideration,theinformationwasrejected,bytheAPI521TaskGroup, frominclusioninAPI521becauseitwasriskbaseddirection,utilizingASMECodeCase2211addressingUG140 Overpressure Protection by System Design and WRC Bulletin 498 Guidance on theApplicationofCodeCase2211‐OverpressureProtectionbySystemsDesign.ThispaperprovidesguidanceforevaluatingoverpressurescenariosduetoaWHBdoubleendedtubefailureincludingasuggestedsequenceforanalysisflowchartandexamples.




1.0 Introduction

API 521, Pressure‐relieving and Depressuring Systems (current edition: 2014)(1) is beingrevised to include information regarding Sulfur Recovery Unit (SRU) Reaction FurnaceWaste Heat Boiler (WHB) tube rupture developed process side overpressure and otherupdates.TheproposedmodificationsregardingtheSRUReactionFurnaceWHBwerefirstballoted in thespringof2015. Onceallof thechangesareballotedandapproved,aneweditionofAPI521willbeissuedin2019attheearliest.ThemodificationstoAPI521(1)thatwereballotedstatethattheusershouldevaluatethesystemtodetermineifthesystemcanrelieve a rate equivalent to double the cross sectional area of a single tube in theWHBwithoutexceedingthecorrectedhydrotestpressureofthereactionfurnaceandotherlow‐pressuresideequipment.TheproposedmodificationssuggestthatSteadyStateanalysisbecompletedfirst. Ifthesteadystateanalysispredictsthatthecorrectedhydrotestpressurewill be exceeded it is suggested that other methods such as Dynamic analysis can becompleted. Theproposedmodificationsalsostate thatanalternative todynamicanalysiswouldbetheapplicationof2013ASMEBPVCSectionVIIIRulesforConstructionofPressureVessels, Division 1UG‐140(4) “Overpressure Protection by SystemDesign”. The proposedmodificationsdonotprovide any real guidanceonhow the analysis shouldbedoneor asuggestedsequenceforthevariousanalysesthatshouldbecompleted.Martens&SternpresentedapaperattheBrimstoneSulfurSymposium–Vailin2014titled“DesigningtoproposedAPIWHBtubefailuredocument”(2). Thatpaperincludedtheresultsofa2001and2014SRUWasteHeatBoilerSafetySurveyofWHBtuberupturesresultinginloss of containment in the SRU system and a proposed methodology for overpressureanalysisofanSRUbasedonaWHBtuberupturewherethemaximumallowablestresstominimumtensilestressprovidesthesamesafetymarginasASME(4)acceptsfordeflagrationdesign. The paper also addressed Quantitative Risk Analysis and Layers Of ProtectionAnalysis(LOPA)referencingWeldingResearchCouncil(WRC)Bulletin498GuidanceontheApplicationofCodeCase2211–OverpressureProtectionbySystemDesign(5) andLayersofProtectionAnalysisSimplifiedProcessRiskAssessment,CenterforChemicalProcessSafety(7).Mosher&OggpresentedapaperattheBrimstoneSulfurSymposium–Vail in2014titled“SRU Overpressure ScenariosWasteHeat Exchanger Failureswith Different Sulfur SealingTechnologies”(3).ThatpaperincludeddetailedanalysisofthreeSRUexamplesforapplicablescenarios based onAPI 521(1) guidance for over pressure evaluationwith a tube ruptureequivalenttodoublethecrosssectionalareaofasingletube.Thispaperisafollow‐uptothese2014papersandfocusesontheauthors’understandingofhow the overpressure analysis should be completed and provides a flowchart with asuggested sequence for the analysis. This paper was developed to provide guidance forengineers who are responsible for evaluation of tube ruptures in new and existing SRUreactionfurnaceWHBs.




2.0 Risk and Probability

“In thisworld,nothing iscertainexceptdeathand taxes” isanoftencitedquotation fromBenjamin Franklin. This quote came from a letter from Benjamin Franklin to Frenchscientist Jean‐BaptisteLeroyonNovember13,1789. Thequotewasactually referring tothenewUSConstitutionandhishopethatitwouldbedurable.Itcanbeinferredfromthisquotethateverythingelseinvolvessomelevelofrisk.Inourdailyliveswemakepracticallyevery decision based on our interpretation of the risk of the action. For example,manypeopleintheUSchooseonadailybasistoeatredmeat.Therearecertainlybothshorttermand longtermrisksassociatedwith thisaction. In theshort term, there isarisk that themeat couldbe contaminatedwithbacteria causing foodpoisoning that can lead to illnessandpossibly death. However,manypeople in theUS trust that the inspectionprocessesthroughout the foodsupplychainand the final foodpreparation isadequate to lower theriskofdeathtoatolerablelevel.Thosesameindividualsmaynotbeabletomakethesamedecisioninotherpartsoftheworldbecausetheriskofillnessordeathishigher.Inthelongterm,thereisariskthateatingredmeatcancausehealthissues(e.g.,pastresearchhastiedred meat to increased risks of diabetes, cardiovascular disease and certain cancers).However,thereisasegmentofthepopulationthatcontinuestoevaluatetheriskonadailybasisanddeterminethattheriskislowenoughorperhapsthattheriskisfarenoughinthefuturethatconsumptionofredmeatisacceptable.Weevaluateriskseverydaywhetherweareconsciousoftheactionornot.A risk is an adverse event that "may" occur. The probability of it occurring can rangeanywhere from justabove0percent to justbelow100percent. (Note: It can'tbeexactly100 percent, because then it would be a certainty, not a risk. And it can't be exactly 0percent,or itwouldn'tbearisk.) Arisk,byitsverynature,alwayshasanegativeimpact(consequence). Themagnitude of the impact (consequence) varies in terms of cost andimpactonhealth,humanlife,environmental,orsomeothercriticalfactor.Onemethodthatengineersusetosemi‐quantifyrisk(socomparisonscanbemadebetweenalternatives and toanacceptable risk level) is to complete a LayerofProtectionAnalysis(LOPA) based on qualifying analysis such as HAZOP. LOPA utilizes the following steps:selectionofascenarioandestablishingaconsequence(impact)andexpressionofrisklevel,identificationofinitiatingeventFrequency(F),identificationoftheIndependentProtectionlayer(IPL),estimationoftheProbabilityofFailureonDemand(PFD),andcombiningthesetocalculateanestimatedquantitativerisk(R).Initssimplestterms,Risk(R)=Frequencyofeventoccurring(F)timesProbabilityofFailureonDemand(PFD)ofthelayerofrelatedprotectionbeingevaluated(7).Whenlookingattheriskitisimportanttodistinguishbetweenprobabilityoffailureofanindividual itemandtheproductofthatprobabilityresulting inaspecificconsequence. Inevaluatingprocessplantswetypicallyconsiderriskintermsoflossoflife,personnelinjury,environmentalconsequence,andlossofproduction.




In the following sections of the paper (Section 3Rationale and Section 4 Flowchart) twoterms are introduced (credible and non‐credible). In this context, non‐credible does notmeanthattheeventwillneverhappen.Instead,non‐crediblemeansthattheeventhasanannual probability of occurrence less than the acceptable risk benchmark probabilityestablished by the User and the appropriate jurisdiction, if applicable. For example, thebenchmarkusedbysomeusersisanannualprobabilitythattheeventwilloccur(F)islessthan 0.0001 (can be written as < 1x10‐4 or < 1 in 10,000) as the maximum acceptableprobabilityforanoverpressurescenariotobeconsiderednon‐credible(5).To help put this value in perspective, the authors found references indicating theprobabilityoffailureofsomecommonlyuseditems;

Theprobabilityof failureondemandofbrakes failingonacar isonce in10years(1x10‐1) per Layers of Protection Analysis Simplified Process Risk Assessment(7).Another reference, The Business of Risk(9), lists the probability of brake failurewithoutwarningasoncein200driver‐years(1/200=0.5x10‐2annually).Evenwiththis probability of failure rate, most of us drive our vehicles every day withoutgivingthebrakesmuchthought.

IntheGuidelinesforProcessEquipmentReliabilityData(10)bytheCenterforChemicalProcess Safety the mean Probability of Failure on Demand (PFD) to Open for aspringoperatedpressurereliefdevice is2.12x10‐4yearsandforapilotoperatedpressure relief device is 4.15 x 10‐3 years in a clean or mild service. APIRecommended Practice (RP) 581 Risk‐Based Inspection Technology(11) providesguidance on how to adjust the probability of failure on demand (PFD) based onservice,dischargedestination,andotherparameters. Basedonthe information inAPI RP 581(11) a conventional spring operated pressure relief device inmoderateservice (most refinery applications) that discharges to the flare (closed system)wouldhave aprobability of failure ondemand (PFD) to openof 5.97 x10‐4 years(oncein1,675years).Thisprobabilityoffailureondemand(PFD)is8to167timeslesslikelytooccurthanthecarbrakefailurePFDdescribedinthefirstexample.

The National Safety Council listed the lifetime odds of death in a motor vehicleincidentat1in112(13). GivenaUSaveragelifespanof79yearsthatmeansinanyyeartheoddsare1in8,848(79x112)ofdyinginamotorvehicleincident. WhichmeanstheRisk(R)ofdeathbymotorvehicleincidentis1.13x10‐4years.Thisisthecombination of probability of the event occurring (F) and the consequence of theeventbeingdeath.Atthisrateagainmostofusaccepttheserisksandeitherdriveorrideinmotorvehiclesonadailybasis.Thepreviouslydescribeduserbenchmarkannual probability <1 x 10‐4 is equivalent to saying that the greatest risk anemployeefacesisdrivingtoandfromwork.




AClassIIIairplane(multipleturbineengineswithgreaterthan6,000poundsofgrosstakeoffweight,i.e.Boeing737orMD‐80)hasprobabilityoffailurecausingaseriousorfatalinjurytoanoccupantof<1x10‐7flighthoursaccordingtoUSDepartmentofTransportationFederalAviationAdministration(12).Convertingthistoflightyearstheprobabilityis8.76x10‐4.Thisisthecombinationofprobabilityofthefailureevent(F)andtheconsequenceofseriousorfatalinjury.Again,manyofusflyoncommercialairplaneswithoutexcessiveconcernoversafety.

InMartens& Stern(2) the resultswere presented for a 2014 survey of Claus ThermalReactorWasteHeatBoilersafetyhistory.Inthatsurveytherewasasinglereportedlossofcontainment(nopersonnelinjuries)ofalowpressure‐ratedSRUduetoaWHBtuberuptureresultinginanoverpressureevent.Thesurveycoveredagrandtotalof20,734SRUoperatingyears.Theonelossofcontainmenteventresultsintherisk(R=F*PFD)ofalossofcontainmentduetoaWHBTuberuptureeventof1/20,734=4.8x10‐5annually.Thevalueislessthanthebenchmarkthatsomeusersuseforthecutoffondefiningwhatisacrediblevsnon‐credibleevent. TherationalepresentedinSection3andshownintheFlowchart inSection4 indicates thateven though thereportedevent frequency islessthanwhatsomeusersdefineasthebenchmarktodeterminecredibleversusnon‐credible, the authors believe it is prudent to add another layer of protection. Thatadditional layer of protection for a maximum double ended tube failure scenario isverifying that the developed overpressure does not exceed the ASME Section VIII(4)criteria and accepted design method presented in NFPA 69 Standard on ExplosionPrevention Systems(6) allowable maximum deflagration pressure methodology fordeformationbutnotrupture.

3.0 Rationale for Sequence of Calculations for SRU Reaction Furnace WHB Tube Failure Overpressure Analysis

Thebasis for this analysis is derived from theAmerican Society ofMechanical Engineers(ASME)BoilerPressureVesselCode(BPVC),anInternationalCode.The2013ASMEBPVCSectionVIIIRulesforConstructionofPressureVesselsDiv1,(4)UG‐21DesignPressure,UG‐22Loadings, and further discussed in Overpressure Protection UG‐125 General requires“overpressureprotectioninaccordancewiththerequirementsofUG‐125throughUG‐138,orwithoverpressureprotectionbysystemdesigninaccordancewiththerequirementsofUG‐140,oracombinationofthetwo”.UG‐126throughUG‐138describestherequirementsforvariouspressurereliefdevices.ForanSRUReactionFurnaceWHBthefollowingpointsneedtobeconsidered;

The“CausesofOverpressure”tobeevaluatedincludetheAPI521Pressure‐relievingand Depressuring Systems(1) events for tube failures, including leaks and a tubefailureequivalenttotwicethecrosssectionalareaofthetube,resultinginpossibleoverpressureoftheprocessequipment.Forsimplicityfortherestofthispaper,thistypeofbreakwillbereferredtoasadoubleendedtubefailure.

The SRU industry experience is a pressure relief device is not suitable for theprocess side overpressure protection due to the process conditions, including thesolidificationofsulfur,resultinginpluggingofsuchdevicesrenderingthemuseless.




Therefore theapplicationofUG‐140OverpressureProtectionbySystemDesign isnecessary.

o UG‐140OverpressureProtectionbySystemDesignincludestheapplicationofCodeCases2211,2211‐1andassociatedWeldingResearchCouncil(WRC)Bulletin498Guidanceon theApplicationofCodeCase2211–OverpressureProtectionbySystemDesign(5).

WRC498(5)statesthatCodeCase2211‐1grewoutofrecognitionthatoverthe years, attempts to implement literally the requirements of UG‐125 onpressurevesselsinsomeprocessingsystemscreatedsituationsinwhichtheoverpressure protection devicewas not dependable, the effectivenesswasnot predictable or the use of overpressure protection devices introducedother dangers due to the need for frequent maintenance or increasedemissions.Ineffect,theliteralimplementationoftheCoderequirementsdidnotprovidethelevelofprotectionintendedbytheCode.Thisinconsistencywasmostpronouncedinprocessingsystemsthatplugtheinletsoroutletsofpressurereliefdeviceswithdepositsineitherthenon‐relievingorrelievingmode,whichisasignificantconcerninatypicalSulfurRecoveryUnit(SRU).

o UG‐140paragraph(b)states“Ifthepressureisnotself‐limiting,apressurevessel may be protected from overpressure by system design or by acombinationofoverpressurebysystemdesignandpressurereliefdevices,ifthefollowingconditionsaremet.TherulesbelowarenotintendedtoallowfornormaloperationabovetheMAWPatthecoincidenttemperature.”

o UG‐140paragraph(b)(2)states“Thedecisiontolimittheoverpressurebysystemdesign is theresponsibilityof theuser. Theusershallrequestthatthe Manufacturer’s data report state that overpressure protection isprovided by system design per (b) if no pressure relief device compliantwithUG‐125throughUG‐138istobeinstalled.Ifnopressurereliefdeviceistobeinstalled,acceptanceofthejurisdictionmayberequired.”

ThesuggestedfirststepapproachtoconfirmOverpressureProtectionbySystemDesignis to analyze thepossibledouble ended tube failure scenarios as a steady state eventanalysis. Mosher & Ogg(3) presented a discussion on this at Brimstone SulfurSymposium–Vail2014.

Iftheprocessequipmentoverpressureislessthan116%oftheprocessequipmentcorrectedMAWP, the systemdesign is considered to be appropriate for any tubefailurescenarioincludingadoubleendedtubefailurescenario.

Iftheprocessequipmentoverpressureexceedsthe116%ofcorrectedMAWPorthecorrectedhydrotestpressure,furtherevaluationofthesystemdesignisnecessary.




The use of Dynamic Analysis Simulation of the double ended tube failure scenario toconfirmthedevelopedoverpressureincludingpossiblesystemmodificationsbyASME(4)and API(1) methods to reduce the severity of the overpressure; An example systemmodificationwouldbetheeliminationoftheinfiniteexternalsteamsourceforthe100%steamscenariobyuseofanadditionaldissimilarcheckvalve(s) to limit thebackflowofsteamintotheWHBfromtheexternalsteamsource.

If the dynamic analysis confirms that the process equipment overpressure(including system modifications if identified) is less than 116% of the processequipmentadjustedMAWP, thesystemdesign is considered tobeappropriate foranytubefailurescenarioincludingadoubleendedtubefailurescenario.

Ifthedynamicanalysisconfirmsthattheprocessequipmentoverpressureisgreaterthan 116% of the process equipment adjustedMAWP (per ASME BPVC(4)) or thecorrectedhydrotestpressure(perAPI521(1))butlessthantheNFPA69StandardonExplosionPreventionSystems(6)allowabledeflagrationpressurefordeformationbutnotrupturecriteria[Martens&Stern(2)describedthisanalysisatBrimstoneSulfurSymposium–Vail in2014]orexceeds theallowablesystemdeflagrationpressurethe use ofQuantitativeRiskAnalysis (perASMEBPVC(4) andWRC498(5))may beutilized to determine if the system design provides acceptable risk for theoverpressureconditions.

TheuseofQuantitativeRiskAnalysis(QRA),suchasLayersofProtectionAnalysis(LOPA)(7),forthedoubleendedtubefailureandothertubeleakscenariosmaybeusedtoconfirmifthe overpressure scenarios are credible (quantified risk evaluation) or non‐credible andlowriskasqualifiedusingWRC498guidance[Martens&Stern(2)].Itiscautionedthateachowner/operatormust establish the necessary quantitative risk analysis criteria includingvalues for occurrence frequency, probability of failure on demand (loss of containment)frequency, acceptable risk, and risk matrix. The authors used the example risk matrixpresentedinFigure4AofWRC498(5)(SeeAppendixB).

The ASME BPVC(4) accepted guidance to industry good engineering practice forconductingaQuantitativeRiskAnalysisisprovidedinWRC498(5).TheavailableUSSRU industry experience for an overpressure loss of containment of the processequipment due to a WHB tube failure is reported in Martens & Stern(2). In theauthors’ opinion, the reported industry experience would confirm that anyoverpressure loss of containment due to a WHB tube failure meets the “non‐credible”scenarioevaluationpertheWRC498(5)guidanceprovidedexamples.The‘‘consequence category” of this scenario meets the “low risk” evaluation per theWRC498(5)guidance(WRC498Figure4AisanexampleRiskmatrix,seeAppendixB).However, it may be prudent to add an additional layer of protection for loss ofcontainmentduetoadoubleendedtubefailurebyconfirmingthatthetubefailurescenarios developed pressure does not result in a rupture of the equipment byutilization of NFPA 69(6) deflagrationmaximum pressure for deformation but notrupturemethodology.




The ASME BPVC Section VIII Div 1(4) Nonmandatory Appendix H Guidance toAccommodateLoadingsProducedbyDeflagration,NFPA69(6)andindustrypracticehasbeenutilizedtosafelyandsuccessfullyaddressdeflagrationdesignpressuresinexcess of MAWP and corrected hydrotest pressures. Appendix H contains thisengineeringdutyclausestatement:“ThelimitedguidanceinNFPA69(6)requirestheapplicationof technical judgmentsmadebyknowledgeabledesigners experiencedin the selection and design of appropriate details.” This same engineering dutyclause can apply to aWHB double ended tube failure with the utilization of thedeflagrationmaximumdevelopedpressuremethodology.

o ItmaybereasonabletoconcludethatthelossofcontainmentoftheprocesssideofaWHBduetoadoubleendedtubefailureisanon‐crediblescenarioandisalowriskevent.However,itisalsoreasonableforgoodengineeringpracticetoprovideanadditionallayerofprotectionforamaximumdoubleendedtubefailurescenariodevelopedoverpressurebyaddressingthenon‐credible and low risk scenario evaluation as not exceeding theNFPA69(6)allowable maximum pressure for a deflagration methodology fordeformationbutnotrupture.

o If the non‐credible and low risk scenario for a double ended tube failurescenario developed pressure exceeds the NFPA 69(6) allowable maximumpressure for a deflagration for deformation but not rupture, the loss ofcontainment for the process equipment may present a risk that wouldwarrant the owner/operator to take measures to reduce the risk topersonnel and facilities by (for example) evaluating operating procedures,administrative procedures, limiting personnel access, and similar items.Increasingthedesignpressureofthesystemshouldalsobeconsidered.

4.0 Flowchart

Based on the rationale previously described and the authors’ experiences, a suggestedsequence of analysis Flowchart was developed to show the various steps and potentialrecycle loops that need to be consideredwhen evaluating an SRU reaction furnaceWHBtubeleak/rupture.




Figure1.SequenceofAnalysisFlowchart




Notes:1. Evaluate three double ended tube failure scenarios for process equipment

overpressureusingsteadystateanalysisprocedures[ReferenceMosher&Ogg(3)];

a. If developed overpressure does not exceed the Maximum AllowableWorkingPressure(MAWP)(ASMESectionVIIIDiv1(4)‐UG‐21,22)ofthelimitingpieceofequipment (including piping) within the overall system corrected for thescenario temperature to design temperature or the corrected hydrotestpressure (API 521 Pressure‐relieving andDepressuring Systems(1)‐ 4.2.2 ) thesystemdesignisacceptable.

b. If developed overpressure exceeds the MAWP (ASME(4)) corrected for thescenario temperature to design temperature or the corrected hydrotest testpressure(API521(1))thesystemdesignisnotacceptable.

i. Proceedtonextstep,theuseofdynamicanalysissimulationtoconfirmthethreescenariosmaximumdevelopedpressure.

2. Evaluatesamethreedoubleendedtubefailurescenariosforprocessoverpressureusingdynamicanalysissimulation[ReferenceMosher&Ogg(3)andCrockett,Moore.,&Jacobs(8)].

a. If developed overpressure does not exceed the correctedMAWP (ASME(4)) orthe corrected hydrotest pressure (API 521(1) – 4.2.2) the system design isacceptable.

b. If developed overpressure exceeds the MAWP (ASME(4)) corrected for thescenario temperature to design temperature or the corrected hydrotestpressurethesystemdesignisnotacceptable.

i. Modify the system per ASME(4) and API(1) methods to reduce theoverpressure.

c. Examples of possible minor/limited modifications to equipment and systemcontrols:

i. ASME(4);

RigorousanalysisprocedurestoverifyactualequipmentMAWP(i.e.,ASMESectionVIIIDiv 1(4) ‐UG‐21, 22 andAPI 521(1) ‐ 4.2.2 correctedhydrotestlimitations), minor revisions to equipment to increase the MAWP andhydrotestlimitations.

ii. API(1);

Change the flow system conditions such as adding a check valve after thenon‐returnvalve(twodissimilarcheckvalvesinseries)toreducethesteamflow that could return to the boiler from the connected steam system[ReferenceMosher&Ogg(3)]




d. Re‐evaluatethethreedoubleendedfailurescenarios;

i. If developed overpressure does not exceed the corrected MAWP orcorrectedhydrotestpressurethesystemdesignisacceptable.

ii. If developed overpressure exceeds the corrected MAWP or correctedhydrotestthesystemdesignisnotacceptable.

iii. Proceedtonextstep.

3. The use of ASME Section VIII Div 1(4) UG‐140 and WRC 498 Guidance on theApplication of Code Case 2211‐Overpressure Protection by System Design(5)methodology to determine the credibility and risk associated with themaximumdeveloped pressure the system must withstand is a reasonable and acceptedengineeringapproach. Thecomparisonof the scenariodevelopedpressure to theNFPA69StandardonExplosionPreventionSystems(6)maximumallowablepressurefor deflagration, based on deformation but not rupture, will provide guidance todetermine if the use of the WRC 498(5) Quantitative Risk Analysis (QRA)methodologywouldbereasonableandappropriate[Martens&Stern(2)].

a. ThisNFPAmethodologydevelopsasystemmaximumpressureallowable forascenariowhichisexpectedtoresultinequipmentdeformationdamagebutnotrupture.Typicallythemaximumcalculatedpressureresultsinastressequalto2/3 of the material minimum specified tensile stress. This methodology isapplicableonlytoductilematerials.

b. EvaluatethethreedoubleendedtubefailurescenariosdevelopedpressurestotheNFPAmethodologymaximumallowablepressure:

i. If thescenariosdevelopedoverpressuredoesnotexceedtheNFPAmaximumallowablepressure,thesystemdesignmaybeacceptableperWRC498(5),iftheQRAdetermined the scenario tobenon‐credible and low risk. Note that theterms “low risk” as used here is not the R (Risk (R) = Frequency of eventoccurring (F) times Probability of Failure on Demand (PFD)) but is anevaluation of the event probability and consequence as described in WRC‐498(5). WRC498(5)guidanceindicatesthataneventriskevaluation(seeWRC498Figure4AforanexampleRiskmatrix,seeAppendixB)ofModerateRiskorHighriskshouldrequireamoreconservativeacceptableRvalue.

ii. If the scenarios developed overpressure does exceed the NFPAmaximumallowable pressure, the QRA risk assessment for the scenario would beexpectedtobecredibleandnotlowrisk.

iii. Proceedtothenextsteptoeither:

1. ConductaQuantitativeRiskAnalysistoevaluatescenariosandestablishcredibilityandrisklevels.

2. Redesignthesystem.




4. The use of Quantitative Risk Analysis (QRA), such as LOPA [Reference Martens &Stern(2)andLayersofProtectionAnalysisSimplifiedProcessRiskAssessment(7)],forthedoubleendedtubefailureistoconfirmiftheoverpressurescenariosarecredible,ornon‐credible,andisorisnotlowriskevaluation[ReferenceWRC498(5)andMartens&Stern (2)]. Itmustbenoted that theuser is responsible toconduct theQRAandestablishtheacceptancecriteriafortheanalysis.WRC498(5)providesmethodologyandguidanceexamplestoevaluatethecredibilityandrisklevelofascenario.WRC498(5)providestheguidancethatthecodecaseconsidersaRisk((R)=Frequencyofeventoccurring(F)timesProbabilityofFailureonDemand(PFD)),basedonactualindustryexperience,of lessthan0.0001(1in10,000yearsofoperation)asanon‐crediblescenarioanda lowriskevaluation isdetermined if the failureexperiencedoes not include multiple fatalities or major long term environmental impact.Martens&Stern(2)providedcurrentUSSRUindustry lossofcontainment(processside of WHB) and fatality and injury data. This data reports a tube failure of 1occurrence of loss of containment attributed to entering feed water after loss ofboiler water level and associated tube failure in 20,734 accumulated operationalyears. No associated fatalities or injuries occurred. The reported industryexperience confirms that aWHB tube failure resulting in loss of containment is a“non‐credible”scenarioanda“lowrisk”evaluation.TheuseofthecriteriathatanytubefailuredevelopedpressuredoesnotexceedthemaximumallowablepressurecalculatedbytheNFPA69(6)methodologyprovidesanadditionallayerofprotectionforlossofcontainmentduetoatubefailureandfurtherenhancesthesafetyofthesystemdesign.

a. Ifriskanalysisconfirmsthedoubleendedtube failure isnon‐credibleand lowriskandmeetstheusesriskacceptancecriteria[ReferenceWRC498(5)]andthedeveloped overpressure does not exceed the NFPA 69(6)maximum developeddeflagration pressure (deformation but not rupture criteria) the design isacceptable.

b. If risk analysis confirms the developed overpressure exceeds the NFPA 69(6)maximum developed deflagration pressure (deformation but not rupturecriteria) the authors recommend the double ended tube failure resulting inrupture scenario should not be considered as non‐credible or low risk[Reference WRC 498(5)] and the system design is not acceptable unless theowner/operator further evaluates the operating procedures, administrativecontrols/procedures, limitingpersonnelaccessand similar items to significantlyreduce the risk topersonneland facilities such that the systemnowmeets theirQRAacceptableriskcriteria.

c. If the final risk analysis confirms the double ended tube failure is a crediblescenario and remains a significant risk [Reference WRC 498(5)] the systemdesignisnotacceptable,andthedesignmustbemodified.




5.0 Examples of How to Apply the Flowchart

The followingexamplesare theauthorssuggestedmethodology forusing thesequenceofanalysis Flowchart for an SRU evaluation for overpressure, due to WHB tube rupturecriteriaforadoubleendedtubefailureresultinginlossofprocesssidecontainment.TheseexamplesareprovidedforguidanceforuseoftheFlowchartandareintendedtoillustratetheprocessbutarenotintendedtobeapplicableforallsituations.TheSRUunitutilizedforalltheseexamplesistakenfromtheMosherandOgg(3)referencePlantB analysis having basic design information of (additional plant information isprovidedinthepaper):

Capacity of 155LTPDof sulfur. When ammonia combustion is accounted for thetrue plant capacity is 174 ELTPD with air only operation and 265 ELTPD withenrichmentupto39%oxygen

WHB with Two pass design with separate steam drum generating 600 psigsaturatedsteam

Tubes–3inchschedule80pipe

Reaction furnace andWHB process sideMAWP 75 psig (hydrotest pressure 97.5psig)

FirstCondenserMAWP55psig(71.5psighydrotestpressure)

Example1Thisisanexampleofthesimplestevaluation;Referencing theFlowchart, the firststep for theevaluationof theplant isaclassicsteadystate evaluation using the three scenarios listed and referring to Flowchart Note 1. Theresultingpressure foreachof thescenarios iscompared to therespectiveequipmentandpiping MAWP and Corrected Hydrotest Pressure (see Flowchart Note 1) and the firstdecisiondiamondisevaluatedperthecriteriaof;

IfthecalculateddevelopedbackpressuresdonotexceedeitherthecodeMAWP,orthecorrectedhydrotestpressure,theevaluationanswerisNOandtheanalysishas“EvaluatedAll SRUWHBTubeRuptureOverpressure Scenarios to beAcceptable”andtheevaluationiscomplete.

If the calculated developed back pressures do exceed the code MAWP, or thecorrectedhydrotestpressure, theevaluationanswer isYESandtheanalysis isnotcompleteandadditionalevaluationisnecessary.




InReference3,MosherandOggcalculatedasteadystatebackpressureof117psigat the1stCondenser. Suppose instead that the calculated backpressure had been say; 68 psig.Thispressure is less thanthecorrectedhydrotestpressureof71.5psig,so theevaluationdecisiondiamondanswerisNOastheanalysishas“EvaluatedAllSRUWHBTubeRuptureOverpressureScenariostobeAcceptable”andtheevaluationwouldbecompleted.Figure2belowshowsthepathtakenthroughtheFlowchart.

API-521 SRU WHE Tube Rupture with area equivalent to twice the cross sectional area of one tube.(Note 1)

Complete Steady State Analysis with 100% steam passing through the tube break. Include all possible open relief paths (vent through tail

gas line, vent through sulfur seals, etc) {Leak approximates the conditions when a top tube in the bundle is uncovered and damaged}

Is the calculatedpressure above the

corrected hydrotest pressure of any component?

(Note 1)

No

YesComplete Dynamic Analysis of tube

break Scenarios. Include all available equipment and piping volumes in

appropriate locations. (Note 2)

Analysis CompleteEvaluated All SRU WHE Tube Rupture Overpressure Scenarios to be Acceptable

Is thecalculated

pressure above thecorrected hydrotest pressure

of any component?(Note 2)

No Yes

Complete Risk Analysis, such as Layers of Protection Analysis (LOPA) to determine if the overpressure

scenarios are credible (quantified risk evaluation), or low or non-credible and low risk for tube rupture

resulting in loss of containment (WRC Bulletin 498 Guidance of Application of Code Case 2211 & 2211-1)

(Note 4)

Is thecalculated

scenario pressureabove the allowable pressure

using methodology for NFPA 69 maximum developed deflagration

pressure (deformation butnot rupture criteria)?

(Note 3)

No Yes

Complete Steady State Analysis with 100% BFW passing through the tube break. Account

for steam production from flashing due to pressure drop through tube break and contact with hot refractory. Include all possible open relief paths (vent through tail gas line, vent

through sulfur seals, etc) {Leak approximates the conditions when a

bottom tube in the bundle is damaged}

Complete Steady State Analysis with normal BFW/Steam mixture passing through the tube

break. Account for steam production from flashing due to pressure drop through tube

break and contact with hot refractory. Include all possible open relief paths (vent through tail

gas line, vent through sulfur seals, etc) {Leak approximates the conditions when any

tube in the bundle is damaged}

Can thesystem be modified

by ASME & API methods toeliminate overpressure

scenarios or reduce the severity of theoverpressure?

(Note 2)

Yes

No

Is thecalculated pressure

above 2/3 of tensile strengthwhich is the criteria for NFPA 69

maximum developed deflagration pressure (deformation but

not rupture criteria)?(Note 4)

Does risk analysis confirm the

scenario is non-credible and low risk (WRC 498)?

(Note 4)

Risk analysis confirms the scenario is credible and significant risk

(WRC 498)

Yes

The risk warrants that the owner/operator further evaluate operating procedures, administrative

procedures, limiting personnel access, and similar items to further reduce the risk to

personnel and facilities to acceptable levels.(Note 4)

Yes

No

Should consider redesigning or reconfiguring the system

Redesign or reconfigure the system to lower the back pressure or increase the

corrected hydrotest pressure.

No

Figure2.PaththroughtheFlowchartforExample1.

Example2The Mosher and Ogg(3) reference determined that the steady state maximum calculated100%steamscenariopressureof123psigisthemaximumpressurescenarioresultbasedoninfinitesteamavailability.The Figure 8 from Mosher and Ogg(3) reference indicates the Steady State 100% steamcalculated developed back pressure of 123 psig shown as the top blue line (assuminginfinite steam availability), did exceed the code MAWP and the corrected hydrotestpressure and the first Flowchart evaluation decision diamond answer is YES, thereforeadditionalevaluationisnecessary.Notethelowyellowlineisadynamicanalysisresultofa~30psigpressurebasedononlythesteamavailableinthesteamdrumwhichwillbediscussedlaterinthisexample.




Figure8 PlantB–174ELTPDScenario1AllSteam–SteadyStatevsDynamicResults

Theauthorsconsider theuseofdynamicanalysismethodology (seeFlowchartNote2)ofthe unit using the same three tube failure scenarios to be an appropriate next step asindicted in the evaluation Flowchart. The dynamic analysismethodology is presented inMosher and Ogg(3) reference. The yellow line in Figure 8 above indicates the dynamicanalysiscalculatedmaximumdevelopedbackpressures,the100%steamscenario,didnotexceed~ 30 psig, however this analysis considered only the steam volume in the steamdrum as theWHB steam non‐return check valvewas assumed to be fully closed and nosteam entering the steam drum from the plant steam system. With this assumption thescenariomaximumpressureis~30psig.However,APIStandard521(1)indicatesthatthesinglenon‐returncheckvalve(whetheritisinspectedornotinspected)istobeconsideredtoremainfullyopenandhasthesameflowresistanceinthereverseflowdirectionasintheforwardflowdirection.MosherandOgg(3)referencecompletedanadditionaldynamicsimulationforsteamenteringthesteamdrumfromtheplantsteamsystemforasinglenon‐returncheckvalveandtheresultsreportedinFigure13,below.Thedynamicanalysisforthe100%Steamscenario,assuminganinfinitesteamsupply,withthefailedsteamnon‐returncheckvalvewithnoresistance,developedamaximumbackpressureof~120psigasshownbytheyellowlineinFigure13fromMosherandOgg(3)reference,whichisessentiallythesameasthesteadystate100%Steamscenarioshownbythepurpleline.Thegreenlineindicatesthedynamicsimulationdevelopedbackpressureof~110psigusingtheAPI521(1)checkvalvereverseflowresistantcriteria.Thelowerbluelineisaddressedlater.




Figure13 PlantB‐174ELTPDScenario1AllSteam‐LeakingSingleCheckValveperAPIStandard521Entering the decision diamond (see Flowchart Note 2) the calculated developed backpressure for the non‐return check valve failed open does exceed the codeMAWP or thecorrectedhydrotestpressure,theevaluationdecisiondiamondanswerisYESandthenextdecision step is either modify the design by use of ASME and API industry consensusdocumentsorproceeddowntheFlowchart.For this example the choicemade is tomodify the design by the addition of a dissimilarsecondcheckinthesteamlinefromtheWHBsteamdrum.APIStandard521(1)providesguidanceforleakratesoftwocheckvalvesinseriesresultinginsignificantreductioninreverseflowcomparedtoonenon‐returnvalvefailingfullopen.The flowrate,using theAPIStandard521(1) guidanceback flow leakage rate, isbasedonassuming the smallest check valve has completely failed and the larger check valve hassevereleakage.Severeleakagecanbemodeledbytreatingthecheckvalveasanorificethatis sized topass10%of thenormal forward flow. For thisexample it isassumedthat theleaking checkvalveback flowrate is10%of thenormal forward rate,which reduces the100%steamscenario calculatedbackpressure from~110psig (green lineonFigure13above)to~30psig,whichisessentiallythesameasthelowbluelineinFigure13.(IntheMosherandOgg(3)referenceFigure13thebluelineinFigure13wasforadifferentsetofcircumstances but the two resultswould basically overlap). Note the low yellow line inFigure 8 is the dynamic analysis result of a~ 30 psig pressure based on only the steamavailable in the steam drum. For this example, by using two dissimilar check valves inseries the dynamic analysis results are roughly the same as assuming the only steamavailableisfromthesteamdrum.




Also it is assumed, for the purpose of this example, that all three scenarios calculatedmaximum pressure are less than the individual equipment MAWP of 55 psig or thecorrectedhydrotestpressures.AsthecalculateddevelopedbackpressuresdonotexceedthecodeMAWPorthecorrectedhydrotest pressure, the evaluation decision diamond answer is NO as the analysis has“EvaluatedAllSRUWHBTubeRuptureOverpressureScenarios tobeAcceptable”andtheevaluation would be completed. Figure 3 below shows the path taken through theFlowchart.






(Note 1)

No





Is thecalculated



No Yes




(Note 4)

Is thecalculated




(Note 3)

No Yes













(Note 2)

Yes

No







(Note 4)


(WRC 498)

Yes




Yes

No




No


Example3This example uses Example 2 input and is an example of the use of Quantitative Risk AnalysissuchasLOPA,andincludestheNFPA69(6)Deflagrationbasedmaximumallowablepressure methodology. Examples of possible “simple” LOPA analysis for over pressurescenariosforlossofcontainmentriskduetoWHBtubefailuresfromMartens&Stern(2)areprovidedforreferenceinAppendixA.




Forthisexampleitisassumedthattheownerhaselected,basedontheirexperience,nottoaccept theAPI521(1) suggestedapproach that theadditionof a secondnon‐similar checkvalvecouldreduceback flow(thereforenoreduction inreversesteamflowrate fromtheheadertothesteamdrumforthe100%steamscenario).The100%steamscenariowithfailedcheckvalve,evaluatedbydynamicanalysisandstaticanalysis,maximumbuiltupbackpressureof~110psigexceedstheMAWPandcorrectedhydrostatictestpressureasindicatedinFigure13fromMosherandOgg(3)reference.Forthisexamplethe100%steamisassumedtoresultinthemaximumpressurescenario.Notethatanyofthethreescenariosmaydevelopthemaximumpressureforaspecificplantevaluation.TheNFPA69(6)evaluationdecisiondiamondstepresultsinaNOwhichleadstoaQuantitativeRiskAnalysissuchasLOPA(seeFlowchartNote4).NotethatforaYESresultit would lead to either further modification to the design or to also conducting aQuantitative Risk Analysis such as LOPA (see Flowchart Note 3). For this example theowner/operator does not select redesign ormodification andproceedswithQuantitativeRiskanalysisutilizingLOPA.Flowchart Note 3 includes comparing the maximum built up back pressure to theNFPA69(6) maximum allowable pressure based on deflagration methodology for adeformationbutnotrupturewhichisa2/3tensiledesignbasisforavoidingtensilerupture.ThismethodologyisacceptedbytheASMEBPVC(4)asprovidinganacceptableriskforalossof containment failure for a deflagration maximum pressure scenario based on specificdesignrequirements in thecode,althoughthismethodologyutilizesa lowersafety factor.Note the top line in Figure 13 indicates the Thermal Reactor Pressure (Avoiding TensileRupture)of180psigandalowerlineindicating132psigforthe1stCondenser.InMartens&Stern(2)theresultswerepresentedfora2014surveyofClausThermalReactorWasteHeatBoilersafetyhistoryforoverpressureduetotubefailures.Inthatsurveytherewasasinglereported lossofcontainment(nopersonnel injuries)ofa lowpressure‐ratedSRUduetoatuberuptureoverpressureevent.Thesurveycoveredagrandtotalof20,734SRUoperatingyears.Theonelossofcontainmentevent,resultsintherisk(R)ofalossofcontainmentduetoatuberuptureeventfrequencyof1/20,734=4.8x10‐5annually.This reported SRU industry experiencewould indicate that perWRC 498(5) guidance thetypicalindustryutilizedacceptableRiskCriteria(R)of<1x10‐4wouldqualifyaWHBtuberuptureeventresulting in lossofcontainmentasanacceptableRbasedonanon‐crediblescenarioandlowriskevaluationasnoassociatedinjurieshavebeenreportedandthatanadditionalQuantitativeRiskEvaluationmaynotbewarranted.HoweverMartens&Stern(2)proposeitisreasonableandgoodengineeringpracticethatanadditionallayerofprotectionbeprovidedeven though the tube failureresulting ina lossofcontainmenteventmaybeconsidered to be an acceptable risk. It is proposed inMartens& Stern(2) that the use ofNFPA 69(6) deflagration maximum pressure methodology, based on deformation but notrupturedesign,providesareducedriskforalossofcontainmentfailureandthereforethisprovidesanadditionallayerofprotectionforacalculatedmaximumpressurethatexceedsthe MAWP or the corrected hydrotest pressure but does not exceed the NFPA 69(6)allowablepressure(seeFlowchartNotes3and4).




TheuseofQuantitativeRiskevaluationandLOPAarediscussedinMartens&Stern(2)andabove. The Owner/Operator is responsible for establishing the appropriate Risk criteriaandotherLOPA inputvalues for theevaluation. TheASMEBPVC(4) recognizes theuseofWRC 498(5) which provides guidance for Overpressure Protection By System DesignincludingtheuseofQuantitativeRiskanalysisandLayersofProtectionAnalysisSimplifiedProcess Risk Assessment(7) reference provides guidance for LOPA analysis. Applying theNFPAmethodology,Figure13abovewouldindicatethatthedevelopedbackpressuresforallthreescenarioswouldnotexceedthemaximumpressureatthe1stCondenserbasedontheNFPA69(6)deformationbutnotrupturemethodologycriteria.ForthisexampletheOwner/OperatorQuantitativeRiskandLOPAanalysis,foraSRUWHBlossofcontainmentduetotuberupture,determinesthescenariostobenon‐credibleandalowRiskutilizingWRC498guidanceas themaximumdevelopedbackpressuredoesnotexceed the NFPA 69(6) methodology calculated maximum pressure. The first decisiondiamondstepaddressingnon‐credibleandlowriskisYES(seeFlowchartNote4)andthefollowingdecisiondiamondstepanswerisNOastheanalysishas“EvaluatedAllSRUWHBTube Rupture Overpressure Scenarios to be Acceptable” and the evaluation would becompleted.Figure4belowshowsthepathtakenthroughtheFlowchart.






(Note 1)

No





Is thecalculated



No Yes




(Note 4)

Is thecalculated




(Note 3)

No Yes













(Note 2)

Yes

No







(Note 4)


(WRC 498)

Yes




Yes

No




No





Example4This example is based on Example 2 however the Owner/Operator selects to utilize theSteadyStateanalysisdevelopedmaximumpressuresandnotconductaDynamicAnalysisbutproceeddirectlytoQuantitativeRiskandLOPAanalysis.ThisExamplebecomesthesameasExample3,includinganalysisresults,exceptfortheuseoftheSteadyStateanalysiscalculatedmaximumpressuresonly.Figure5belowshowsthepathtakenthroughtheFlowchart.






(Note 1)

No





Is thecalculated



No Yes




(Note 4)

Is thecalculated




(Note 3)

No Yes













(Note 2)

Yes

No







(Note 4)


(WRC 498)

Yes




Yes

No




No


Example5This example is the same as Example 3 or Example 4 except the last decision diamondresultsinaYESanswerasanyoneormoreofthefollowingassumptionsapply;

TheOwner/Operatordetermines that theiracceptableRisk(R)criteria is<1x10‐5(insteadof<1x10‐4)whichwouldnotallowtheSRUWHBtuberuptureresultinginlossofcontainmenttobeconsideredasnon‐credible.




TheSteadyStateandDynamicSimulationcalculatedmaximumpressuresexceedtheMAWP, corrected hydrotest pressure and NFPAmethodologymaximum pressurecriteriaresultinginanunacceptableRiskcondition.

The Owner/Operator does not accept the NFPA 69(6) methodology maximumpressure as suitable to provide a low risk situation and will use the MAWP andcorrected hydrotest pressure criteria which are exceeded resulting in anunacceptableriskcondition.

This would result in the decision diamond leading to the lowest Process/Informationsymbol stating “The risk warrants that the owner/operator further evaluate operatingprocedures, administrative procedures, limiting personnel access, and similar items tofurtherreducetherisktopersonnelandfacilitiestoacceptablelevels”.The output from this symbol states “Should consider redesigning or reconfiguring thesystem”.HowevertheOwner/OperatorhasthefinalresponsibilityfortheevaluationsanduseoftheFlowchartandmayevaluatealltheconditionsandattributes,suchastheuseofacomplex LOPA with additional layers of protection, taking into account their specificexperience and conditions to establish an acceptable risk meeting the Owner/Operatorestablished Risk Criteria (see Flowchart Note 4). Figure 6 below shows the path takenthroughtheFlowchart.






(Note 1)

No





Is thecalculated



No Yes




(Note 4)

Is thecalculated




(Note 3)

No Yes













(Note 2)

Yes

No







(Note 4)


(WRC 498)

Yes




Yes

No




No





6.0 Conclusions

Thefollowingconclusionscanbedrawnfromtheresultsofthispaperandfromwhattheauthorsexperiencedwhilepreparingthispaper;

API Standard 521(1), is being revised to include information regarding SulfurRecoveryUnit(SRU)ReactionFurnaceWasteHeatBoiler(WHB)tuberupturesandotherupdates. Themodifications toAPI521(1) thatwereballoted in thespringof2015statethattheusershouldevaluatethesystemtodetermineifthesystemcanrelieve a rate equivalent todouble the cross sectional areaof a single tube in theWHBwithout exceeding the corrected hydrotest pressure of the reaction furnaceandotherlow‐pressuresideequipment.Theproposedmodificationsdonotprovideanyrealguidanceonhowtheanalysisshouldbedoneorasuggestedsequenceforthevariousanalysesthatshouldbecompleted.

NoconciseguidelineonhowtoapproachtheoverpressureevaluationprocessforanSRUWHBtuberupturescenariocurrentlyexists inthepublicdomain. Thispaperwaswrittentohelpfillthatvoid.

EachOwner/Operatormustestablishtheirownprocessforevaluatingoverpressurefrom an SRUWHB tube rupture scenario. This paperwaswritten to express theauthor’sopinionregardingtheirsuggestedsequenceoftheevaluationandprocesssteps.

EachOwner/Operatormustdeterminewhattheyconsideranacceptablerisktakingintoaccounttheirspecificexperienceandconditions.

EveniftheresultsoftheQuantitativeRiskAnalysisevaluation,foranSRUWHBlossof containment due to tube rupture, determines the scenarios to be non‐credible(the event has an annual probability of occurrence less than theOwner/Operatoracceptable risk benchmark probability) and a low Risk, it is reasonable for goodengineering practice to provide an additional layer of protection. That additionallayer of protection is to evaluate the overpressure scenario and confirm that themaximum pressure does not exceed the NFPA 69(6) allowable maximumdeflagrationpressuremethodologyfordeformationbutnotrupture.




7.0 Recommendations

ItisexpectedthatthefinalAPI521revision(probablyin2019)willincludeallorsomeoftheproposedWHBtuberuptureeventlanguagewhichwilltasktheSRUowner/operatortoevaluatetheoverpressureofthelowpressureprocessside. TheauthorsrecommendthattheSRUowner(s)/operator(s)fullyengageandparticipateintheballotingoftheproposedWHB tube failure provisions to provide suitable industry experience and guidance to thedocument.It is the authors’ experience, as given in this paper, that the proposedWHB tube failureevaluationisanecessary,lengthy,andsomewhatdifficultengineeringprocess.TheauthorsrecommendthatSRUowner(s)/operator(s)initiatetheevaluationofpossibleoverpressurebefore the API expected 2019 revision is issued both to support input to API 521 ballotcommentsandtodeterminewhatpossibleimpactontheirrespectiveSRUmayresultfromcompliancewiththefinalAPI521update.8.0 Acknowledgements

TheauthorswouldliketoacknowledgeandthankBrimstoneSTSLimitedforprovidingtheforumforthe2014SRUWHBTubeRupturePanelDiscussionandforallowingthispapertobepresented.ThegoalofthepaneldiscussionandthispaperwastoinformandassisttheSRU industry toanticipate the inclusionof theproposedSRUWHBprovisions in thenextrevisionofAPI521.9.0 Author Contact Information

DennisH.MartensConsultantandTechnicalAdvisorPorterMcGuffie,Inc.Lawrence,KS,USAmartensdh@pm‐engr.com

AlanD.MosherPrincipalProcessEngineerBlack&VeatchCorporationOverlandPark,KS,[email protected]




10.0 References

(1) APIStandard521,Pressure‐relievingandDepressuringSystems,AmericanPetroleumInstitute,6thEdition,January2014.

(2) Martens,D.H.,&SternL.H.,“DesigningtoproposedAPIWHBtubefailuredocument”,

BrimstoneSulfurSymposium‐Vail,September2014.(3) MosherA.D.,&OggD.A.,“SRUOverpressureScenariosWasteHeatExchangerFailures

with Different Sulfur Sealing Technologies”, Brimstone Sulfur Symposium ‐ Vail,September2014.

(4) 2013 ASME BPVC, Section VIII Rules for Construction of Pressure Vessels Div 1,

AmericanSocietyofMechanicalEngineers,2013Edition.(5) WRC498,Guidanceon theApplicationofCodeCase2211—OverpressureProtection

bySystemsDesign,TheWeldingResearchCouncil,January2005.(6) NFPA 69, Standard on Explosion Prevention Systems, National Fire Protection

Association,2014Edition.(7) LayersofProtectionAnalysisSimplifiedProcessRiskAssessment,CenterforChemical

ProcessSafety,AmericanInstituteofChemicalEngineers,Copyright2001.(8) CrockettS.,MooreA.,&JacobsG.,“SRUWHBTubeRuptureModelingInsightsLearned

@BPCherryPoint”,BrimstoneSulfurSymposium‐Vail,September2014.(9) TheBusinessofRisk,PeterGeraldMoore,CambridgeUniversityPress,1983.(10) Guidelines for Process Equipment Reliability Data with Data Tables, Center for

ChemicalProcessSafety,AmericanInstituteofChemicalEngineers,1989.(11) API Recommended Practice 581, Risk‐Based Inspection Technology, American

PetroleumInstitute,SecondEdition,September2008.(12) AdvisoryCircular,SystemSafetyAnalysisandAssessment forPart23Airplanes,U.S

DepartmentofTransportationFederalAviationAdministration,November2011.(13) “What are the Odds of Dying From…” National Safety Council, accessed July 16,

2015,http://www.nsc.org/learn/safety‐knowledge/Pages/injury‐facts‐chart.aspx




Appendix A ThefollowingtableisfromMartens&Stern(2).








© Black & Veatch Holding Company 2015. All Rights Reserved. The Black & Veatch name and logo are registered trademarks of Black & Veatch Holding Company. Porter McGuffie, Inc. and Black & Veatch 25 September 2015

AppendixBThefollowingtablesarearecreationoftablesfromWRC498(5).

ProbabilityCategory

A L M M H H H

B L L M M H H

C L L L M M H

D L L L L M M

E L L L L L M

F L L L L L L

VI V IV III II I

ConsequenceCategory

Fig.4A–ExampleofaRiskmatix,H=HighRisk,M=ModerateRisk,L=LowRisk

Category Description AnnualProbabilityRange

A VeryLikely ≥0.1(1in10)

B Possible ≥0.01(1in100)but<0.1

C Unlikely≥0.001(1in1,000)

but<0.01

D HighlyUnlikely ≥0.0001(1in10,000)but<0.001

E NotCredible ≥0.00001(1in100,000)but<0.0001

F PracticallyImpossible <0.00001(1in100,000)Fig.4B–ExampleofProbabilityCategories


© Black & Veatch Holding Company 2015. All Rights Reserved. The Black & Veatch name and logo are registered trademarks of Black & Veatch Holding Company. Porter McGuffie, Inc. and Black & Veatch 25 September 2015

Category Description Examples

I CatastrophicMultiplefatalities;majorlongtermenvironmentalimpact

II Major Fatality;majorshorttermenvironmentalimpact

III SeriousMajorInjuries;significantenvironmentalimpact

IV Significant SeriousInjuries;shorttermenvironmentalimpact

V MinorFirstAidInjuriesonly;minimalenvironmentalimpact

VI None NosignificantconsequenceFig.4C–ExampleofS/H/EConsequenceCategoriesCategory Description Examples

I Catastrophic ≥$10,000,000

II Major≥$1,000,000

but<$10,000,000

III Serious ≥$100,000but<$1,000,000

IV Significant≥$10,000

but<$100,000

V Minor ≥$1,000but<$10,000

VI None <$1,000Fig.4D–ExampleofEconomicConsequenceCategories

Documents

API 521 SRU RF WHB Evaluation for Overpressure due to …pm-engr.com/wp-content/uploads/2016/03/2015-Brin... · Brimstone Sulfur Symposium API 521 SRU RF WHB Evaluation for Overpressure