MATERIAL SELECTION FOR REPAIR OF DAMAGED PROCESS PIPING IN HIGH-TEMPERATURE SULFIDATION SERVICE IN THE NO.

4 CRUDE UNIT NO
December 12, 2012

Edwin H. (Ned) Niccolls Sr. Consulting Materials Engineer Chevron Fellow Prepared with the assistance of: Barbara Smith Senior Business Manager Richmond Refinery, CUSA

Edwin H. (Ned) Niccolls Sr. Consulting Materials Engineer Chevron Fellow

Materials and Equipment Engineering Chevron Energy Technology Company 100 Chevron Way, 10-3506 Richmond, CA 94801 Tel 510-242-4539 ehni@chevron.com

MATERIALSELECTIONFORREPAIROFDAMAGEDPROCESSPIPINGINHIGHTEMPERATURE SULFIDATIONSERVICEINTHENO.4CRUDEUNIT1 ExecutiveSummary OnAugust6,2012,afireoccurredintheNo.4CrudeUnit(CrudeUnit)attheRichmondRefinery, ownedandoperatedbyChevronU.S.A.Inc.(CUSA).Whiletheinvestigationintothecauseofthefire isongoing,preliminaryinformationindicatesthatthefireoccurredintheareaofaleakina5feetlong sectionofthe200feetlong4sidecutcarbonsteelpipeinhightemperatureservice.Itisbelievedthat theleakresultedfromacceleratedsulfidationcorrosioninthe5feetlongsectionofpipe,whichhad lowsiliconcontent(lessthan0.10wt%Si). CUSAhassubmittedapplicationstotheCityofRichmondforpermitspursuanttotheCaliforniaFire Code(CFC)toreplacefiredamagedpipinginhightemperaturesulfidation(HTS)serviceinthe CrudeUnitwith9Chromealloypipe(9Cr).Becausequestionshavebeenraisedbycertainmembers ofthepublicaboutCUSAsselectionof9Cr,theCityhastemporarilydelayedissuingpermitsforsuch workandBillLindsay,RichmondCityManager,hasrequestedadditionalinformationconcerningCUSAs selectionofmaterial. Asexplainedbelow,CUSAsselectionof9CrpipetoreplacefiredamagedpipeinHTSservicesatisfiesall engineeringandfiresafetystandards,andotherindustryrecommendedpractices,fortheuseinthe CrudeUnit.WhilecertainmembersofthepublicandtheChemicalSafetyBoard(CSB)havesuggested 300seriesstainlesssteel(300SS)asanalternativematerialbasedonitsincreasedresistanceto sulifdationcorrosion,theuseof300SSwouldintroduceanewdamagemechanismnotpresentwith9Cr thatismoredifficulttomonitorandinspectthansulfidationcorrosion. MaterialsSelectionProcess AstheownerandoperatoroftheRefinery,CUSAhastheresponsibilityandtechnicalexpertise necessaryforselectingmaterialsforaparticularservicebasedonsoundengineeringandindustry practices.Inselectingmaterials,CUSAreliesuponexperiencedmaterialsengineers,whousetheir expertjudgmenttochooserobustandpredictablematerialssuitablefortheplannedservice,takinginto accounttheriskspresented.Anymaterialselectionmustbesupportedbyacomprehensivemonitoring andinspectionprogramtoensurethattheselectedmaterialisperformingconsistentwithexpectations.
1

PreparedwiththeassistanceofBarbaraSmith,SeniorBusinessManager,RichmondRefinery,CUSA.

December12,2012 Page2

Thematerialselectionprocessiscomplexandbaseduponconsiderationofmultiplefactors,including operatingconditions(particularlytemperature),operatinghistory,processchemistry,velocitiesand otherflowconditions,localunitconditions,potentialunusualoperatingconditions,andturnaround considerations.Itisalsoimportanttoconsidertheriskfromallpossibledamagemechanisms,aswellas theabilitytomonitortheequipmentagainstthosedamagemechanisms.Otherwise,adecisioncould inadvertentlyshifttheriskfromonedamagemechanismtoanother.Wheneverpossible,CUSAselectsa materialthatbestaddressesthesemultipleconsiderations. TheCitysScopeofReview Oncematerialsareselectedfortherepairwork,CUSAappliesforpermitsfromtheCity,whichreviews thepermitsforcompliancewiththeCaliforniaBuildingStandardsCode,inclusiveoftheCalifornia BuildingCode(CBC)andtheCFC,asimplementedbytheCity.2Theinstallationofprocesspiping suchasthepipingbeingreplacedaspartoftheCrudeUnitrepairisregulatedintwoways: SupportstructuresforthepipingsystemsrequireabuildingpermitpursuanttotheCBC; Thematerialsforprocesspipingandthedesignofthepipingsystemarereviewedfor compliancewiththeCFC. PursuanttotheCFC,theCityspermittingroleistoconfirmthatthematerialsCUSAhasselectedto replacefiredamagedpipinginHTSserviceintheCrudeUnitcomplywiththeengineeringstandardsof theAmericanSocietyofMechanicalEngineers(ASME)CodeforProcessPiping(ASMEB.31.3)and thefiresafetystandardsoftheNationalFireProtectionAssociation(NFPA)Flammableand CombustibleLiquidsCode(NFPA30).3 TechnicalAnalysisforReplacementofFireDamagedPipeinHTSServiceintheCrudeUnit PursuanttoASMEB31.3andNFPA30,carbonsteel,5Chromealloy(5Cr),9Cr,and300SSaresuitable forserviceintheCrudeUnit.Inadditiontoadheringtotheseengineeringandfiresafetystandards,itis importanttoconsideramaterialssulfidationresistancewhenselectingmaterialsforHTSservice.API 939Cidentifiescarbonsteelwithadequatesilicon,5Cr,9Cr,and300SSasexamplesofmaterials suitableforHTSservice,dependingonvariousfactors.4Anadditionalconsiderationinselecting materialsforHTSserviceispriorexperienceandinformationconcerningamaterialspastperformance inthatservice.5
SeeRichmondMunicipalCode(RMC)6.02etseq. Afurtherdescriptionofthemannerinwhichengineeringandfiresafetystandardsarepromulgatedandincorporatedinto theCitysreviewisprovidedinAppendixI. 4 SeeAppendixII. 5 Id.
3 2

December12,2012 Page3

Takingintoaccounttheabovedescribedfactors,aswellastheriskspresented,theRefineryMaterials Engineerexercisedherexpertjudgmentandselected9Crastheappropriatematerialforreplacement offiredamagedpipingsystemsinHTSserviceintheCrudeUnit.Thisdecisionwaslaterconfirmedby otherexpertswhohavesincereviewedthedecision. Asnoted,9Crsatisfiesallregulatoryengineeringandfiresafetystandardsforcontainmentand processingofcrudeoil.Further,asshownbytheModifiedMcConomyCurvesintheAPI939C,the sulfidationcorrosionratesofcarbonsteelwithadequatesilicon,5Cr,9Cr,and300SSdemonstratethat eachmaybesuitableforHTSserviceintheCrudeUnit,inparticularwhenonetakesintoaccountthe CrudeUnitsoperationalhistory.6Thus,basedonallapplicabletechnicalstandardsandrecommended practices,9CrisasuitablematerialforreplacingthefiredamagedpipingsystemsinHTSserviceinthe CrudeUnit,andprovidessignificantlyincreasedresistancetosulfidationcorrosionwhencomparedto thelowsiliconcarbonsteelcomponentinvolvedintheAugust6incident. CUSAunderstandsthatcertainmembersofthepublicandtheCSBhavecommentedthat300SSmight beabettermaterialfortherepairworkbasedonitshighersulfidationresistanceascomparedto9Cr,in particularinlightoftheFebruary2012lossofcontainmentthatoccurredattheBPCherryPoint Refinery,wherethepipethatfailedwas9Crandthedamagemechanismwassulfidationcorrosion. WhiletheBPCherryPointincidentmayseemrelevanttotheselectionof9Crforpartsoftherepair work,basedonthepubliclydisclosedBPinvestigationofthisincident,wedonotbelievetheincident presentsananalogoussituationtotheRichmondCrudeUnit.Thepipinginthatinstanceappearsto havebeenasemistagnantdeadlegwhich,after29yearsinhightemperatureservice,allowed corrosivestobuildupinavaporspaceatthetopofthepiping,leadingtothefailure.Weunderstand thattheflowinglinesinHTSserviceintheBPCherryPointrefineryhadnoproblems,andthatBP replacedthepipeinquestionwith9Cr.Thus,thisincidentdoesnotsupportaconclusionthat9Crisnot suitableforHTSservice,butrathersupportsindustryeffortstoeliminatedeadlegsystemsasmuchas possibleandemphasizestheimportanceofexistingindustrystandardsrequiringaspecificdeadleg inspectionprogramfordeadlegsremaininginservice. Importantly,theselectionof300SSwouldalsointroduceanewdamagemechanismtotheCrudeUnit intheformofstresscorrosioncracking(SCC)fromchlorides,andpotentiallyfrompolythionicacids, thatwouldnotoccurwith9Cr. Whereas9CrisimmunetoSCC,chloridesinthepresenceofwatermaycauseSCCofany300SSpiping attemperaturesaboveabout140F.LossesofcontainmentduetochlorideSCCarewelldocumented intheliterature.7Further,CUSAhasidentified10instancesofSCCinstainlesssteelpipesinhigh
Id. ASMMetalsHandbook,Volume13CCorrosion:EnvironmentsandIndustries;CorrosioninPetroleumRefiningand PetrochemicalOperations,R.D.Kaneeditor
7 6

December12,2012 Page4

temperatureserviceincrudeunits,mostlyfromchlorides.8Thepipesthatcrackedintheseinstances wereinsimilaroranalogousservicetothepipingbeingreplacedaspartoftherepairoftheCrudeUnit, whichissubjecttopotentialrisksfromchlorideSCCbecausethecrudeoilsprocessedcontainchlorides, asdoesambientmoisturesuchasthatfromthedriftfromtheadjacentNo.3CATCoolingTower. Anotherpotentialdamagemechanismwith300SSispolythionicSCC,whichoccurswhensulfurscales combinewithoxygenandwatertoformsulfurousacidsthatcancracksensitizedstainlesssteel. Althoughthepotentialforthisdamagemechanismcanbemitigatedbyusingtheappropriategradeof stainlesssteel,itisstillarelevantconsideration.9 Afinalbutvitallyimportantconsiderationinselectingmaterialsforaparticularserviceistheabilityto monitortheequipmentagainstdamagemechanisms.Akeyreasonfortheselectionof9Crisits predictablecorrosionrate,whichmakesmonitoringofsulfidationcorrosionmoreeffective.Onthe otherhand,SCCfromchloridesorpolythionicacidsresultsinmicroscopiccracksthataredifficultto detectpriortofailure.Thus,theuseof9Crpresentslessoverallriskthat300SSwhenitcomesto detectingandpredictingcorrosion,anddoesnotintroduceanewdamagemechanismtotheCrude Unit. CUSAsselectionof9Cradherestotheapplicableengineeringandfiresafetycodesandisthebest choiceforpurposesoffireandoperationalsafetybecauseiteffectivelyreducestheriskfrom,and providestheabilitytoeffectivelymonitor,sulfidationcorrosion,whileavoidingtheriskofSCC altogether.

SeeAppendixIII. ThebenefitsofstainlesssteeldosometimesoutweighthepotentialforSCC,suchaswhenarefineryprocessesnaphthenic acidcrudes.TheRichmondRefinerydoesnotprocesssuchnaphtheniccrudes,however,sothisisnotaconsiderationfor theselectionofmaterialsforrepairoftheCrudeUnit.

9

December12,2012 Page5

Appendices AppendixI AppendixII AppendixIII ApplicationofTechnicalStandardstoNo.4CrudeUnitRepairWorkThroughCalifornia BuildingStandardsCode AmericanPetroleumInstituteRecommendedPractice939CGuidelinesforAvoiding Sulfidation(Sulfidic)CorrisionFailuresinOilRefineries IncidentsofStressCorrosionCrackinginStainlessSteelPipingCircuitsinHigh TemperatureServiceinCrudeUnits

APPNDIX I: APPLICATION OF TECHNICAL STANDARDS TO NO. 4 CRUDE UNIT REPAIR THROUGH CALIFORNIA BUILDING STANDARDS CODE The City of Richmonds Permitting Authority The City of Richmonds review of permit applications for the repair work being undertaken by Chevron to the No. 4 Crude Unit and related equipment (Crude Unit) is conducted pursuant to the California Building Standards Code (CBSC), Title 24 of the California Code of Regulations (CCR), which has been adopted and incorporated by the City in Chapter 6.02 of the Richmond Municipal Code (RMC). o State law (Cal. Health & Safety Code 18930 et seq.) provides that the CBSC are the minimum standards applicable to all "occupancies," i.e., all buildings and structures, in the state. o The California Building Standards Commission adopts these technical codes based on input from expert agencies, and reviews and amends the codes on a regular cycle. Cal. Health & Safety Code 18929.1; see also (http://www.bsc.ca.gov/codes/adoptcycle.aspx). o Local agencies such as Richmond may adopt local modifications to provide more restrictive standards provided the agency makes express findings that amendments are necessary because of local climatic, geological or topographical conditions. Amendments are not effective until copies of both the express findings and the amendments have been filed with the California Building Standards Commission. Cal. Health & Safety Code 18941.5. The CBSC includes multiple technical codes, divided into Parts, including as applicable to the repair work: o The California Building Code (CBC), 24 CCR Part 2, which regulates demolition and repair of buildings and structures. 24 CCR 1.1.3. o The California Fire Code (CFC), 24 CCR Part 9, which regulates structures, processes, premises and safeguards concerning, among other things, the hazard of fire and explosion arising from the storage, handling or use of structures, materials or devices. CFC 101.2. o The California Electrical Code (24 CCR Part 3) and the California Mechanical Code (24 CCR Part 4) are also applicable. The installation of process piping, i.e., the piping in the Crude Unit through which flammable and combustible liquids flow, is regulated in two ways:

o Repairs of the support structures for the piping systems require a building permit pursuant to the CBC. A building permit must be obtained from the Citys Building Official for all such work under the CBC. CBC 3405.1. o The selection of the materials for process piping and the design of the piping system require fire construction permits pursuant to CFC 105.1 (General), 105.1.2.2 (Construction permit), 105.7 (Required construction permits), and 105.7.7 (Flammable and combustible liquids). The fire code official, which in Richmond is the Fire Marshal, is responsible for reviewing and issuing fire construction permits. RMC 8.16.030(c); CFC 105.1.1. Technical Standards of the CFC In reviewing and issuing fire construction permits, the Fire Marshal determines whether the proposed process piping systems comply with the general provisions of the CFC, as well as technical standards that are incorporated into the code as set forth in CFC Chapter 47 Referenced Standards. The General Provisions applicable to storage, dispensing, use and handling of hazardous materials, including combustible liquids and flammable solids, liquids and gases, are set forth in CFC Chapter 27. Section 2703.2.2.2 of this chapter provides that supply piping for these materials shall be in accordance with ASME B31.3. o American Society of Mechanical Engineers (ASME) standards are promulgated in accordance with American National Standards Institute (ANSI) protocols. This is a rigorous, public process which brings together technical expertise from all sectors (e.g., industry, government, NGO, education) to formulate technical standards for specific subject areas. (http://www.asme.org/kb/standards/aboutcodes---standards). The Committee for ASME B31.3 is a continuing one, and keeps all sections current with new developments in materials, construction, and industrial practice. New editions are published at intervals of two years. (The history of ASME B31, as well as process used by the B.31.3 Committee to develop and publish the standards, is discussed in further detail in the Forward and Introduction to ASME B.31.3.) o ASME B31.3 - Process Piping sets forth the engineering requirements deemed necessary for the safe design and construction of process piping typically found in petroleum refineries; chemical, pharmaceutical, textile, paper, semiconductor, and cryogenic plants; and related processing plants and terminals. ASME B31.3 sets forth, among other things: References to acceptable material specifications and component standards, including dimensional requirements and pressuretemperature ratings Requirements for design of components and assemblies, including piping supports

2
704007280v2

Requirements and data for evaluation and limitation of stresses, reactions, and movements associated with pressure, temperature changes, and other forces Guidance and limitations on the selection and application of materials, components, and joining methods Requirements for the fabrication, assembly, and erection of piping Requirements for examination, inspection, and testing of piping

In conjunction with CFC Chapter 27, CFC Chapter 34 Flammable and Combustible Liquids, sets forth the requirements for prevention, control and mitigation of dangerous conditions related to storage, use, dispensing, mixing and handling of flammable and combustible liquids. Section 3403.6.2 - Design, fabrication and installation of piping systems and components, provides that piping system components shall be designed and fabricated in accordance with the applicable standard listed in Chapter 27 of NFPA 30. o The National Fire Protection Association (NFPA) is an international organization that develops scientifically-based consensus codes and standards, research, and education for fire and related safety issues. NFPA is an International Codes and Standards Organization (ISO), and all NFPA standards are developed in accordance with ISO protocols requiring a full, open, consensus-based process that brings together technical expertise from various fields to develop, revise and update standards on during a three to five years revision cycle. (http://www.nfpa.org/categoryList.asp?categoryID=124&URL=Codes%20&%20 Standards&cookie%5Ftest=1). o NFPA 30 Flammable and Combustible Liquids Code sets forth the fundamental safeguards for the storage, handling, and use of flammable and combustible liquids. Chapter 27 Piping Systems, applies to the design, installation, testing, operation, and maintenance of piping systems for flammable and combustible liquids or vapors, and 27.4.1 - Materials Specifications, provides that pipe and other materials used in piping systems shall meet the material specifications and pressure and temperature limitations of ASME B31. o NFPA 30 also provides standards for the installation of piping systems ( 27.6) and the testing of piping systems ( 27.7).

There are numerous other standards and codes incorporated into the CFC as set forth in CFC Chapter 47, including certain American Petroleum Institute (API) standards. API Recommended Practice 939-C Guidelines for Avoiding Sulfidation (Sulfidic) Corrosion Failures in Oil Refineries, is not one of the standards incorporated into the CFC; nor is API Recommended Practice 580 Risk-Based Inspections or API 581 Recommended Practice Risk-Based Inspection and Damage Mechanisms. These recommended practices are applied voluntarily by the industry. 3

704007280v2

Scope of Fire Marshal Review and Approval of Permits When the Fire Marshal reviews and issues fire construction permits pursuant to the CFC, he is determining that the proposed process piping, including materials, complies with engineering (ASME B31.3) and fire-safety (NFPA 30) standards that have been promulgated and peer-reviewed by international standards organizations. CFC 105.2.4. The Fire Marshal does not determine in the first instance whether a material proposed for use by an applicant, such as 9 Chrome alloy pipe, is adequate for the proposed service from an engineering or fire-safety perspective. Rather, the Fire Marshal confirms that a proposed material such as 9 Chrome has been recognized as suitable for the proposed service in the applicable technical standards. CFC 105.3.

4
704007280v2

APPNDIX II: API RP 939-C (SUBMITTED IN HARD COPY ONLY)

APPNDIX III: INCIDENTS OF STRESS CORROSION CRACKING IN STAINLESS STEEL PIPING CIRCUITS IN HIGH-TEMPERATURE SERVICE IN CRUDE UNITS

SCCIncidentsin300SeriesStainlessSteelHotPipingPortionsofCrudeUnitsChevron
188SS Issue Number

Date

Author

Source

Summary

Chevron1

1970s

DLCooke

Chevron2 Chevron3 Chevron4

1971 1975 1986 2000(reportedin 2006metallurgy review)

WGHalsted DEFerrell JWCoombs

PrivatecommunicationbetweenDave SaltLake:Externalcrackingofstainlesssteellineattributedtowater CookeandformerSaltLakeinspector andchloridesusedduringfirefightingoperations. ElSegundo:Externalcrackingof316SSpipingatoutletofE356 duringstartup,followingshutdown,duetoeitherchlorideSCCor PTASCC. ChevronE241.01 ChevronRT9915844 Chevron75.16.56 Historybriefsummarycapturedin ChevronPasc.Crude1MetReview (relevantexcerpts) Hawaii:InternalPTASCCof304VacuumColumnbottomline. ElPaso:InternalchlorideSCCofNo.6CrudeUnitAtmospheric 304SStransferline. Pascagoula:InternalcrackingfromchlorideSCCto317LSSpipingat #6S/C.

Chevron5

EHNiccolls

*"Hotpiping"referstopipingthatoperatesabovewhereliquidwaterormoisturewouldbeexpectedduringnormaloperation,nominallyabove~400F Footnote: 1)ThesedonotincludecasesofSSinstrumenttubingcracking.

SCCIncidentsin300SeriesStainlessSteelHotPipingPortionsofCrudeUnitsIndustry
188SS Issue Number Industry1 Industry2 Industry3 Industry4

Date 1997 1997 1998 1998

Author TomFarraro(Citgo) RalphBlee(Exxon) TomFarraro(Citgo) CAShargay(Fluor Daniel) TomFarraro(Stress Eng)

Source 1997NACERefincor 1997NACERefincor 1998NACERefincor 1998NACERefincor

Summary InternalchlorideSCCof316SStransferlineatalowspotthatwas likelynotdrained. InternalSCCofSStoCrMotransferline(eitherchlorideorPTA SCC). InternalchlorideSCCaftersodaashwashingofSSincrudeunit. ExternalSCCin316SStransferline. InternalchlorideSCCof304SSvacuumtransferline(SCCattributed tocondensationatlowpointsandconcentrationofsaltsduringstart up).

Industry5

2004

2004NACERefincor

*"Hotpiping"referstopipingthatoperatesabovewhereliquidwaterormoisturewouldbeexpectedduringnormaloperation,nominallyabove~400F Footnote: 1)Thesearchforstresscorrosioncrackingincrudeunitsalsouncovered"dozens"ofcasesofchloridecrackinginotherunitssuchashydroprocessing.Aswediscussed duringoutmeetingonNovember26,2012,thehotpipingportionofcrudeunitscanexperienceduringstartupandshutdownconditionssimilartohydroprocessing units.

Chevron

(Private communication between Dave Cooke and former Salt Lake inspector)

Chevron2

I'fI

. L. l .

SHAN\-

{'; le

This lrork wab requested by D. D. Prlce. J G tlerr/p7;6J. G. KERR

9-24-7L
MATERIALS IABOITATORY
RICHT"OND

PIPE, EL SEGUNpO N0. 2 C8,U!E ul,EL_

FALLURE OF STAINTBSS STEET

.\ ,. f fIE z E-24L.0L t-r+ffi!nn eL igJ.lt

)'

I
I

1l'' .-.-t I
I

t.

il*( tt ,,'(
.y

nj t'|\

.'r

AL your request the MateriaLs rboratory conducued st,eeL piping epooi inLo the circumferential cracking of the stan1"es6 ""AD.JSJ"l?H'Z fron the outlet of exchanger E-356 n the EL, Segundo No. 2 Crude ltnfr/ The falure sras discorered July 30 when th9'pplng spooL started Leaklng {g5gg the start-up after a shutdown..' During the shut,doldn a verticaL T-joint was wetded into the tw year oLd stainless steeL pping spool. The field welds were stress reLieved at about 1600F with an Ex+o-met, sress relef kit while the piping spooL.was boLted up and highly rest,rained. The'spool was found to be Leaktng at two circrunferenEial icracks about 1/8 inch apart.and about, LlL inch long. The cracks were ilocated on the opposte side of the plpe fronr the ltTtt, and.about 8-lnchee rfrom one of the T-joint weLds. et nonnat operating conditione the 6 Llzinch 0D stanless steel plpe containe gas olL at about 450F and about 225 psL,

A chemical anaLysis of the pipe lndicates that it 1s made of Type 316L sEainless steeL. The result,s of the chenricaL analysis are glven in TabLe 1. The pipe has a hardness of about Rockwell B 78. This lndicates that the stainless steeL plpe is annealed, and hs a yield strength of about 30 ksi and a tensLe strength of about B0 ksl. A Lo'ngitudna1 cro6s sect,ion contaning both circumferentlal cracks was cut from the pipe. Figure L shows photomicrographs aken at the ID and OD of a crack excendng through the pipe wall. Apparently Lhe cracklng started at the OD snce the crack is more open there. Also, several other cracks startng at the OD and not extending through the alt were found within a L/4 fnch of the crack sho!i ln Figure 1. The tip of one of these cracks is shown ln Fgure 2. This crack appears to be intergranular and tightly cLosed.
Figure 3 shows an etched cross sectons of one of the cracks noc extendlng through the pipe lall. This photomfcrograph cLearLy lndlcaces tht the cracklng is intergranuLar The photomicrograph in Figure 4 rhows that carbides have precipitated at the graln boundares, which suggesLs hat he Type'316L scafnless seel pipe has been sensitized.

Sensitization is the harmful preclpftation of chromium carbdes in a nearly cont,inuous network around the metal grains of an ausEenit,ic stainless steel. The chr<rnirm cont,ent, of these carblde particles is theorized to be so dfeproportlonately high with respect to the metal iself , t'hat a thLn surroundng enveiope 1s depleted of fts chrmiun contenL. Thls reduces the thln envel-ope t,o nr stalnl-ess steel behavlor, and renders lt susceptible to subsequent intergranular eorroslon. The t,enperature range n which BensitizatLon occurs Ls from about 800F to 1650F. Sensitza1on ie a tLme-emperauie effect: wtrlch ls mos rapid

frm

about .250F

to

1350F;

:.2
FAITURI] OF STAINI,BSS STEEL PIPE, E,L sBcUNpO NO. 2 gRupp lrNrr _

_"

FILE:

gEPEMB,.ER

E-24L.OL 2f L97.1

It does not appear lkeLy that thls failure vlas caused by stress relief cracking sLnce Type 316L stainLess steeL is not generally belleved to be susceptible to this kind of cracklng. Ln generaL only the grades of stafnLess st,eeL containing coLumblum are Busceptib].e to stress rellel cracking Type 316L stanless steel would not be expected o lose much ductility durlng a stress relief'heat treatment eince no coLubium ls present to form columbium carbidee and cause precipitation hardening, Therefore, Type 316L sLainLess steel should be abLe to pLastically deform enough during a st,res6 reLef heat treatment t,o reJ.eve the stresses needed to cause crackfng. For these reasons it is not believed that this feLure l'ras caused by stress reLief cracking even though the pping spool was boLted up and highly restrained during the stress rellef heat treatmen. It is beleved that the cracks are from ntergranular etress corroslon cracking of the senstized Type 3L6L stainLess steeL. Chlordes and poLythioaic acids are the most common corroslve agents found around the Refinery tha cause'stress corrosio'n cracking n Type 3L6L stainless steel. ChLoride crackng generalLy starts as int,ergranular cracking and then develops as transgranuLar crackng. Polythionic acids onLy cause intergranuLar stress corroson cracking. Therefore, it appear.s thag the fallure of the sLalnless steel ppe vras caused by polythionic aclds rather'than chLorides, slnce only

intergranuLar cracklng was found. rrPolythionc acidsrr is used as a general term referring to sulfur acids, such as sulfurous acid., The lntergranuLar faiLure of sensitized Type 3L6L stainless steel nay occur 1n any plane where ron suLfide corrosion products are formed. Durlng shut,dons the lron sulfide oxldizes on contact wth air to form sulfur dloxide. Ir the presence of moisturereulfurous acid 1s created. Thls process ls descrlbed by the follovrlng two reactlons:

(1)

'4Fes + 702

----+

4sO2

+ ZFe2Og

(2) soz + H2o 4

H2so3

spooL fron poLythionic acids. the outlet of exchanger E-356 came n'contact wth However, it appears that ths dfd occur after the stress relief heat treaLment, during whlch the Type 316L stafnless steeL was sensitized. Even though the exposure of the outside of the plpe spooL to polythionic acids is probabLy.not coirunon, precautons shouLd .be'taken to prevent, recurrence by keeping suLfur containlng materLal"s off he pipe. At the operating tmperacure of 450F it is possible but improbable that sulfur contaning materials on the pipe could nake'lron sulfide. However at stress relief temperatures any sutfur contalning material would corrode the staLnless steeL pipe leaving an iron sulfide residue that wouLd produce polythlonic aclds af,t,er contact wlth alr and vrater. 'l.d. .,b.''tlotr( I^IGH/ hh W. G. HALSTED Attah.

It is not cLear how the outsde of the Btainless steeL piping

cc: B. P.

Faae

Table 1.

Elemgnc

Pgr.ceng

c
Cr
Gu
Mo

0.03

17.00 o.L4
2,2(+

ln

..63
13. L5

Ni
P

0.030
0.013

si

'

0.65

Pipe

Pipe ID

FIGURE

1. Cross sectf.on of circumferenial crack extending through pipe wall.

unetched 50X.

FIGIIRE

2.

Cross

eection of crack

tiP not exending

through. pipe.

unetchd 250X.

)/

-oD

.;'f.

\;
1.1

Crack

.';
'::i4
!!t'.

FIGURE 3.

Cross sectlon of crack not exending through pipe. etched 42.5X.

qIcURE 4.

Cross secion of crack tip showing carbide preciptates at grain boundaries. etched 500X.

MEMORANDUM Go.I44

ro,

i - ji-,":irlr\-'------ [ 1_j--e2-..cr-i-\cl-l\,g\\+sej\rlq--L)c;l"oir?.t---$-rc 1! Fr un. -(il- h-l,lcr[srecl- -]Ylcr!qlt'-L(Js--\clr-r-lc''-\,' '--kicLriroixl---: [l." L{.\-CIj-uqr i lj=Ler'rl**':--l:ies{- --tt E-\ -$cundan,--r
MR.

No.l-,-Clrcle--11nl

fl,'-

re

Y()UR F I tT:

\ci-r\-onr*L 'r-t-{or:meioq---Cbnc.gia-rrr,-]lr-mtecrca,nu\c\o s}.c-.a'. -)!i1LftQ 'rxtc\ii.rr -s*:qrUzKl-tps- ---ai qe:il, llac..l /\ 'l\ ,r \. \ . ' , 1 1l, iibc--ctociclEa--.sr-rocf r-,n\ t* iibc--t*o#,LE*-ft'-)rocJ turce--cn -\>ptii:en -\>prii:en.yq
-crc
r

-il--l-i-$rsrhle-il.-iir,:,rhle-rn-',;sifi:.\ceniJcitt clc'cki,%l-I!)dIL-rt lrxrn+\C!.u$1-CCrrc$?'l

-:f-errv-r \*,sstfii:L,\::.c< /\ 1\

i.,

'--'.,-t I t r '\ ,n* r',.t*r,..r\c'.o ..- 5:i\.-'-\ ^-': S,L- :;\((. 'ho,
r r

-p(,ItJir-r_aors---aeiCs--cn--cb-\olle[e$-ecr icl hr\\/., ..

\ .. . \

li) r rr i r I fi.fnt- rPe'c'l mg:an-&-n--qnnckrnf :, -^-1he--:Uornl*'x,-ds lmq itirp,.,,:tt"t o-e*hanq9c-:g5(' r-. n'ior- l\ke\q,---inHlhp ori\siclp cl -thp rlp cnrrp, l*\, \. (r ( r . n ------i--r -rn-cor-t<sl-t.rdlf -cblc,rrLcx--lharr-i: _tl _ls_ N:,r_9__no,\robLe=_thci_l= he_ cI cc-Lc-ie -- sLrgss-s \ -tt\cl p:s:c.(l-ruors- r$rs-el- ne--rrerr' io p--vc )\'t-, t-c. c( \olenc \)
r

e(

lst<l-I-ne.

- l---i:q-Leepur<*--h\onicle a onc,l .st,tific-cons

rHEr

REPLY

tS RtgUtRED, F0RR0 0Rt0illAL Ail0 0r(

C0PY

Chevron 3

rVTEfu?TRA\D{-JM
RT01 12843 Engineering Department taLerial-s Laboratory Rchrnoncl, Calif ornia
(,.t,
Noveruber

@s

r:11 tts

26,

L975

FrLtiD TYPE 304 VACUTM BoTTOMS LNE,

HATNAIIA,{ RIFTNERY

*fateri.als Lab Fii-e

TviR.

?-100, 43

iavrai

R. L.

BREEHOFT:

The attached Maeral-s Laboratory repor:L concLudes that a sensitzeC Type 30lr line n vacuuin bottoms sersce at 6B0F fail-ed by sulfur acid crackng fronr the ID. Sensitz-aton may have occurred during fabrication bu more probabl-y was the result of servce. Cracking by thi-s mechanisnr wouLd have occurred durng shutdor,ms cvhen raosture eontaced inernaL sulfide scal-e. Because of the fLre ltazerd presented by the possble fornation of addtj.onal cracks, we recomrnend tha al-L Type 304 parts of the pipe spooL be replaced. The use of a stabil-ized stainlesg steeL, J-ike Type 32L or Type 347, in stress reLieved repLaeement piping, should preven both cracking and corrosion, Carbon steeJ- or chrore-moJ.y steels rriL1- preven cracking, but, ca corrode rapdi-y at 6B0F in the presnce of. H2S, the corroson rat,e depending on the quantty of I{2-S presen, Careful study of projecEed corrosion rates based oa servce experience v.th related equipment should be made before any of the chrome-rno1-y steeLs are used.
\

i; Ii;;{,:,+
J. G. KERR

,...'

DEF/Jhh

cc: E. H. Edwarcls - S.F, R. D. Ssvitters - S, F. B, P. Faas - El Segundo C. 1't. Angi-in - Harva F. A. Mcr\IlLin - CRC D. F. Boaz - Inspectfon If. B. Floyd - El Segundo R. C. Dover - Pascagoula

;S C A .1 N E D iF:tri?5tuS|,l

l.iATJlliI\LS LAUOP.ATOIY

P.IC( rNl)
:

IAILED TYPE 304 VACUM{ BOT'IO, LTI'E, IIAIATIAN REFINERI

RACKGROU}ID

FILE:

P-100. 43
L97 5

NOVEI,IgER 25

Irr nid=Septernber t-re 6-inch clj.aneter portiorr of the vacrtlim bottoms lir,e in lraruaiit" ruo Unit clevelope<l cracl:s acljacent Lo r rv?-I.-d. joining the line E,o .,ti)'l iaricatecl of Type 3QLr',' ada"' a redtrcing elboi.r. The bottoms line "-apotd*tl otcl-oc.k position cin the sLainless steel-. The craclcs, r,'hch rere in the l-2 line ru'ere vreld repaired rth E-308 manual r,reldmetal and the Line r''as pi-aced back in service. ter only one day of operaLion, a. secolrd crack appeared at about the 4 orclock postion n he sare general area. The line had been in vacuum bottons serviee at 680f ancl 350 ps:'-g for 15 years rhouL ilrcidenE. Becruse the line s nsulated, metal- lqal-J- t.emperature l,ras probably very c1.ose to the 6B0F sock ternperature, No records vre aval-abJ-e to indicate thether the pipe had been stress relieved.

reituce the loads appliecl by a nearby valve rnanfolcl, The ci:acked erea vlas f lane-cut frorn the 30t' p.pe section removecl; both pieces t/ere senL to the l'later.als Laboratory for an nvestigation of the failure mechanisrn.
INVESTICATION

Thrty inches of the line, irrcluding the cracked segment, &'ere replaced rvith Type 321. Welds r.lere rnade ,ith E-347 manual rel-drnetal ancl rvere sress relieved at l-600F for one hour. A brace ras installed near the fal-ed area to

After c1-eaning wth a solvent o remove tar, the ID of the cracked piece from the l.ne appeareC es shor,m in Figure 1. Several cracks are vsble., Simflar examinaton of the OD found onLy one short crack Lhat could be observed vsual-l.' but thaE vras nE pked up by dye penetrant nspection, even after he piece \.ras l.rarmed to 125F. The cracks are either very tigh or are pl-ugged with scaLe and vacuum resicl. All--Sa,ks -are avray fiorn rveld zones,
The resuLts of a chercal- analysis of a sample of the pipe materal, l-isted in Tabl-e 1, shoroed the pipe to be Type 304 and not Type 304L.

Metal-lographc examination of a cracked portion of the ppe r'rall found that cracks were entirely ntergranular and originated at the line ID, as shotrn in Fgure 2. No cracks r,ere ob.seryed in .eldmetal or r: r,eld heat affected zones. Huey tests of wo p:ipe sampl-es, one neer the cracks and the other 30t'arvay from t,tle r.eld, foun<l that the entre secton of pipe rras severely sensitized. Corrosion raEes measured in the Huey tests, listed in Tabl-e 2 were virtual-ly dentical for both samples. Figures 3 and 4 suggest that rvel-d heat affected zones are not as severeLy sensitzed as ppe materT,.
DISCU SSTON

Failure most probably oecurred by a sr.rl-fur acid or polythionic acid cracking mechanism in he sensitized Type 30r pipe, These acids can form when iron sulfde scales come in contact .riEh rvater. Since the inside of the line is r.vell above the der.r point of valer, even at he 350 psg operating temperature, crackng most probably occurred during shutdor.rns, Simlar craclcs may have formed elser'rhere n the piping. Even if t vere possible'Eo find and reoair all craclcs in the approximately 1-00 feet of Type 304 pipe remaning, additional
cracking is a distinc possib1ty.

*2Records do not indj.cate vhethe.r the pipe spool \,as stre-ss relevecl . Tf j.t had been, loc.al slress re.lief of r.relds r,ou1d have been the nost probabl.e methocl because of tiie size a.nc1 configurati-on of the spool. Local sress Telief of welds rvould ror-, have caused sens;itizaicn of the portion of ppe f ron r':hich Sanpte B (see Table 2) r.ias dra'.,m. Therefore, service expobrrre seems Lo be the mosU pr:obzrbJ-e cau.se of sensitidat.ioir, al-tholrgh a furnace stress reliaf cor-rld haive p::oduced the sa.li:e effect. Tha bottoins tenperature , !,,,-lof 680ir is some..hat below th-* 700F tempeatrlre that Socal considers to be the itJ ,{,'," /lorer lrnt for sensiti.zation of Typa 304 stainless steel. }lorever, lve have settled on this 1o'er limit from tests using exposuT:e times shorter than the n,.t'.t,' '1" ' ,: l,\ I 100,00C hours of service accumulated by the \racuum bottons line. n'.' /- \.) ,t. .,.
.1,.,

adclitional c::ackng rnay occur n ' ,^',, ,,r'.Becauseline, not just r.;el-d areas, may be sensitized areas, and because the entre sensitized, Ehe Type 304 pipe spool.,,q it'" -,,r:|'7' bt.' l,", I'shoul-<l be replaced. Leaks that may forrn r,ill probably be snral1-, but he r_ i,'l rislc of spontaneous ombustion of ol- saturated. insul-ation presenLs a .r-,1 substantial fire l;azar. Type 304L sections may not need replacement, but no ,,Lt !' , .,^.,trtt..| nonclestructve means presently exists to di.stinguish either Type 304 fron Type . !!lr'i,, ,-r"' 304L or sensitized areas fronr unsensitized areas. llul'' .t,, " .'l .o,y''" .,,,;'' FasLors to b considerec{ in he sel-ector oi mat.erial-s for repla,ceinent piping are presented n Table 3, Sensitized ar.rstenitic stainless can fail as '' l[, ^,' described above, ancl so Type 304 should noi be use",,. Although Type 304L .,zil ,.. ')-ri"':; .,.!, \ rvorrl<l probably not have failed as above, Type 321''fype 347 s preferecl for ''', J irlproved resistance to sensitization cluring seryice. All. austenitj.c stainless "'' ^ij , ^.r91'.",,,)<' st,eels may be suscepcj-trle to chl-oride cracking fr:oin the 0D under insul-ation, eilte:: froin salts leachecl from the nsul-ation ol: from salts carried n the |'e'' t,\L"' , ,,t at.nosphere in Hawaiits ocea.nsde envronmeni. I,le, therefore, reconrneind stre-ss t\ /,, i.r., -..i.'.gf .
'^?, ''
'-,

..ar1oi

l'{,'ri';,'"'', carbon steel or chrorne-moly steel piping rvoulcl eliminate all risk of failr-rre by ,,'.'. cracking, if properly fabricated. Iiorever, at 680F, hese materi-al-s can "-.-!.,!o' corrocle irr the presence o 125. Corrosion rates will be dependent upon lI2S I.i"".l 0 coneentrations and on erlJ-oy coingosition. The corroson rates of carbon steel \..*u|.y't,.1.,1,.r, or Cr-Mo equipment operating at smiJ-ar rnetal- t,emperatures (not stock temperatures) in resid service 4'"^lu'' A used. Increased should be examned careful-l-y before any of these materials are '4t' chromium concentraon improves resistance to H2S,
.,n)u'' '.)
CONCLUSIONS AND RECO}fu',IENDATIONS

i-. 2.

The Type 304 v'acuum botlorns Line r,as severely

failed by sul-fur acicl cracking. Senszaton

service.

sensitized and probably

may have occurred during

3.

Type 304L secLons nay not be sensitized and r.roul-d, therefore, noL requre repl-acement. Simple Lests on sma11 sampl-es (as small as a sl-iver l-/4r'wide L/2" Long) sent to the MaEerials Laboratory could determine rEhether he material is sensitLzed. The 4-inch pip:i-ng should be tested in this rvay before replaeement,. For replacement piping, we recormend the use of stress relieved rype z or Type 347l- Carbon steel.o a Cr-l'o steel (up o 9 Cr-r1/2 to) or Type 410 stan1ess steel rnay be used if operaLors carr estabiish that il2fi concentations and, therefore, corrosion rates r.ri,11 stay at satsfactory 1or, level-s.

Because the rest of the J-ine presents a fre hazard from possble formaEon of additional craci<s, we recommend repl-acement of the entre pipe spoo1.

DEF/jhh

D. E.

$,

i*.**,-tL
}'ER.I{.EI-L

(
Tabl.e
Cn-ernicaL

(, I
Analyss

Btms

Lne

304

304L max.

C 0.056 Cr L9.00 Ni L0.60 Mo 0.18 Mn 1,74 st 0,50 P 0.025 s 0.007 Cu 0.17 '"
(per
Tabl-e
ASTII
2

0.08 max. 0.0.035

1-8-20 8*LL 2.0

0,7

L8-20 8-13

2.O
0 .7 5

5 0,04 0,03

0.04 0.03

)
secton
rrCrr)

lluer Tests on Two Samples from Pipe I'Ia1L

Lst /r8 hours

Samp1e

^262,

2nd 48 hours'
2150 mpy
2L2O mW

A Sample B
Sanple A

LL50 mpy 1065 mpy

- CLose to qreJ.d, but away from IIAZ SampJ.e B - At opposfte.end of 30" pf-pe sanrple frorn SampJ.e A

TabLe

Choce of RepJ-acemen l4aterials PossibLe FaiLure Mechansms

Stress ReLeved

Not St_resg Rel-ieved

304 30{lt 32L 347

arb b

brc
c c
c

c,s.
5Cr

d
d

d
e

a - Ilill senslt,ze during stress reLief . Can faL by su3-fur acd cracking from ID during shutdowns when 1-iquid 'rater

-_ \ " I ..

is

present'

'-

lf.y ensltize during servce.

crackng.

Can

faiL by sulfur

acf.d

c-

May faL by chLoride crackng under fnsuLaton durng shutdowns, when lquid water is present and met,al temperature

is )rsor.

d - May fal by H2S corrosion durng servce.

- May fail by sulfide craetcing from ID tf
shutdorvns when

liquid

BHN

aer

s present.

) ZtS during

FIGURE 1.

of vacLrum bottoms lne shcx,'ing cracks on cleaned metal surface (left) an<l penetrant inspecte<i surface (right). l\eld 1n 1ine is at bottom of pictttre'Dye check clid not reveal all cracks' inclicating tfrat cracks re tilht and/or f illed wj'tli vacuum resicl'
photogr:aphs <lf IL)

FIGURE

2. Lght etch of cracked pipe cross secEion shor's that crack is entirelY inter:granular and that material is ProbablY severely sensitized.
I 00x 10% oxatic

acicl etch

15 sec.

L./
,-v ,/,
f

]TIGURE 3.

llAZ of pipe welcl is PartiallY ditcheci af ter 1Y262 etcb for 30 sec.

l-:t.\
\: i ({.-:. 1..t:
\ /,.1

' .

\1,.
\.,i,

za

.\
)--.r
. ,,
.

)
tr'
i-- '-iF

{',

500x ASTII 262

oxalic acid etch, 30 sec.

4. PiPe material awaY from heat effects of welding is severely ditched aft.er identical etching treatment as Fgure 3.
FIGURE

Huey

tests show this Part of is severely senstLzed.

PiPe

500x
ASTM

4262 oxallc acid

etch, 30 sec.

Chevron 4

MEMORANDUM GO.I44

Lt/ - 9.4
t-t
a

1,

?os.

/-

| 4'L
) 5 ko,.;
s
u./1

fit ls

/n",

- r;/,L

,-u,tll

MEMORANDUM

zl.
our FlLl:

,J

3UIJECT:

'*rE
,

-0/-7o
l.rl4a a

/
4

Sc!

/**

e-,

frtr l?LY 3 rEQul80. t0llD 0l8tl^L 0 0lt

g0Y

MEMORANDUM Go.4

19_

^ta,^

.L

.9

(u
lor-,h;t

tcunt9, trrlD

0t

r\,1

MEMORANDUM GO.44

19-

tt?LY l3 rEgulto' Fo^D OlltlL 10 0lt

c0?Y

...r. l.t01

MEMORANDUM Go.I4,I

t9_

ol

riltr lEtly ts rEqulSto, FOrD ort6ta! D

OtE

COpy

MEMORANDUM Go.I4.I

C1,,G,.,

T-*

Nrt s =,4

T
tr [pLy ts rfQutt0, toA0 Ot6ttlL ^t0
orE c0py

MEMORANDUM Go.I44

rtq b
oul
F

tlt:
FI

rour

t:

|r*
\,q

1,ry

,l

rilEr t?Ly rs tgutrtD, Fo10 0rt6tAL Ato ett

copy

Chevron 5

Chevron

Corrosion Mtgation Plan

Chevron Products Company

March 2006 (Updated May 2007)

Ned Niccolls

- Author,

Energy Technology Company

Waldo Jurisson and Morris Bryant, Inspection- Contributor

Alice Burch and Angie Clark, Process - Contributor Andrew O'Connor, Metallurgy - Contributor

(:()Rr{osro\N'il'r(;,\'ilo\l't r\-l'\s(,rr;r)r tr('ttr

Code

t)t.. 1-- j\1

\t{(il1{}l) rr(:l'trr.t t.tr1

rl

llX}71

3 l. I piping advises against using carbon steel above 775F to avoid graphitization of carbon steel, and only gives allowable stresses to 800F for carbon steel. The ASME/ANSI Process Piping Code 31.3 advises against using carbon steel above 800F to avoid graphitization of carbon steel, although it provides allowable stresses for much higher temperatres. Currently Crude Unit I limits the carbon steel piping to 825F which is safe for a number of years because at 825F the rate and amount of graphitization will be slight. However, even allowing this higher 825F limits plant operation.

The alloy piping should be used until the temperature drops to below 800F. Note the corporate piping manual currently advises limiting carbon steel to 750F to assure no graphitization, but we will revise the manual to match the Code wording. [Update 5/07: Rechecking the temperature found the carbon steel to be

actually operating at only 680'F

no action required.l

8.

C-1102 Vacuum Column Type 317 Stainless Steel Packing and #6 s/c Stainless Piping Recommendation: Upgrade the C-I102 top bed packing to Inconel or ceramic, whichever is most economic. See also related recommendations below for online monitoring and shutdown inspection work.

Background
challenged by having very little knowledge as to the actual process or corrosion conditions of this corrosion loop. The type 3 I 7 stainless steel packing shows signs of conoding, and the #6 s/c piping system contains chlorides which could lead to pitting and stress corrosion cracking of the 317L stainless steel.
Vy'e are

In 1998 the top packing in C-l102 was heavily fouled and the 317 stainless packing had holed through in some center portions of the bed. We believe this was due to corrosion at chloride deposits. Naphthenic acid conosion is possible, but less likely with the 317 stainless metallurgy. The crude unit is now again seeing signs of packing in various components of the 6th sidecut. An informal examination of packing samples in 2004 indicated chloride corrosion. In 2000 an 8" discharge line from P-I116 cracked underneath scale deposits from what was believed to be chloride cracking.
Reviewing the UT and RT data of the #6 slc indicates we could be seeing moderately high conosion rates of the stainless piping. We believe that is an artifact of the relatively new piping and the diffrculty of monitoring thin stainless piping. While we camot yet prove that the general corrosion rate of the stainless is low, we expect that it is low and the real threat is from pitting or cracking at locations where deposits and water can collect.
Fortunately the 2003 inspection of the bare C-t 102 column shell at the top packed bed location showed only mild pitting of the carbon steel. The exchangers in the 6th sidecut circuit also have shown little damage, due to2205 Duplex stainless installed in 1998 in E-l103 and E-l146A'. The carbon steel E-11468 bundle showed moderate ID and OD pitting and was retubed in kind in 2003. The 4O-year-old E-l128 carbon steel frn fan exchangers showed fouling but very little

Chevron Corp

10/98

(.'()ril<rst();\iUlil(;.\'r

lo,\l't.\\-l',rs(.\r;0t].r('trr t)t:l-Nl,\t(il2006(tJt'r).\il..r)N'1,\l

2()ll7)

Recommendation 2: Conduct annual AUT (or scanning UT if AUT is impractical) inspection of the bottom head and nozzles of C-1105 (discuss), 2, Loop # 12 (System 12,6 slc piping)

Recommendation

Conduct AUT, and possibly some RT, upstream and downstream of P-I1164 (also simply do an external visual walk-through). Do the first AUT in the first quarter of 2006, conduct follow-up AUTs on about one year intervals as long as the chloride issue remains. Recommendation 2: Take a sample from as close to the column as practical, to measure chloride and water content. Take the first sample in the first quarter of 2006, plan follow-up samples once per quarter as long as the chloride issue

l:

remains.

Recommendation 3: Recheck the carbon steel piping portion that appears to be showing activity. Specifically note the carbon steel piping leading to the E1128's, and the E-l128 fouling.

Background Until proven otherwise, we need to assume the stainless sidecut piping is conoding like the stainless column packing. Unfortunately, this corrosion is likely to be highly localized and could even cause chloride sess corrosion cracking (there was one case of P-l l16A discharge piping that was suspected of chloride cracking in 2000). This system has had significant and possibly sporadic corrosion since at least the 1990s, but it appears to have worsened in recent years. It is very diffrcult to confidently determine conosion rates from the existing UT data, due to relatively recent upgrade to stainless for much of the system, and
substantial scatter in the data.
3. Possible Mismatch of Some Piping Metallurgy Entering the C-1101' C-1102'

C-I103 columns (See 2.1, items 1-5)

Recommer,rdation: If we determine there is a mismatch of piping metallurgy entering the C-l101, C-l 102, and C-l103 columns, then a TML at each mismatch should be established within the next few months to monitor that location until the piping or nozzle can be upgraded at the 2008 shutdown. See the detailed notes in Section 3.0 for specific locations which may have mismatches. 4. Loop #3 (System 3, Top of Atmospheric Column and Overhead Piping)

Recommendation 1: Conduct bi-annual AUT of the carbon steel sections of overhead piping, with order of priority being the piping into the E-l101s, and the piping in and out of the E-1120 fin-fans. Recommendation 2: RT additional low points/drains in the overhead piping systems, especially in area of dwg 003-l l (note point 30).

Background
Note that the atmospheric overhead compressor section appears to have areas localized corrosion.

of

Chevron Corp

1s/98

coRRosroN

MTTTGATTf)N Pr-.rn

PL,rN

- Psc.{GouLA Cnt oe I -

VlARcu 2006 (UPDATED MA 1 2007)

Equipment

Material

Description and Historical Data

Max
Temp

TAN

Sulfu
(wto/o)

Corrosion Rate
Hst. Calc.

Remaining Life
Range

(Channel: cw) E-1124 Tubes

NB

Tubes:

Admiralty SBI I l-445

TIS: NRB SB-

- installed; 04 uniform corr.

02

- A&B:

very light

200

NA

4 H2S

<3

>10

t7t464

3.12 Corrosion Loop #12 (System #12): Vacuum column E-l1464/8. From E-l146A18

6th sde cut, and 7/E sidecuts

This corrosion loop consists of the #6 sidecut piping and associated equipment from the C-l102 vacuum column, through E-I103 and the # 6 sidecut goes both back to the C-l102 cohmn via the E-I128 air cooler, and to the Isomax 2 plant via the E-l140 air cooler. This system also includes the C-I102 vacuum column #8 sidecut from C-l102 through E-l sidecut goes both to the Isomr I plant (and to Isoma:< feed storage), and back to C-I102.

l4lA

to E-l I12. From E-l I 12 the #8

Carbon steel has shown surprisingly high corrosion rates since at least the 1990s, and some historical notes link the higher rates with processing naphthenic acid crudes. We now believe water and chlorides are important factors. Upgrading much of this system to 3l7L SS in recent years will greatly reduce the general conosion, but the stainless is susceptible to pitting under deposits, or possibly even chloride stess corrosion cracking. The 31 7SS packing in the C-l 102 column has seen significant chloride corrosion. There is still some CS in the lower temperature piping that is showing some activity and needs rechecking.

This loop also contains#71#8 sidecut piping and associated equipment. Note that some of the lower temperature piping is still carbon steel and seeing moderate activity. The hotter portion of this piping is in the range where relatively smll increases in temperature (20-30F) and TAN levels could substantially increase carbon steel corrosion in this part of the circuit. Discuss if we should put in a probe here, or other monitoring. Make sure the emergency wash oil line near the outlet of P-1119 going from C-1102 to C-1103 is the right metallurgr and is covered somewhere.

Chevron Corp

70t98

Industry

I&2

97c5.1-01: GRUDE AND vAcuuM uNlr (NAPHTHEN|C ActD coRRostoN): Mike Beaton (Petro-Canada) - asked for experience with PASCC of stainless steels used in crude and vacuum units used for protection against naphthenic acid corrosion. After discussion this was further clarified as "L" grade (low-carbon) and regular-grade materials. A
show of hands indicated no problems with "L" grade material.

I I I

97c5.1-02: CRUDE AND vAcuuM uNtr (NAPHTHENTG AcrD GoRRostoN): Tom Farraro (Citgo) - reported a cracking problem in a solid type 316 SS transfer line that was probably caused by chloride SCC after it was soda ash washed one time. The cracking occurred at a low spot between two pipe supports that didn't get drained. This incident was about 10 years ago and since thattime soda ash washing has not been used, and nofurther
incidents of cracking have occurred.

determined whether the cracking was due to chlorides or pASCC.

97c5.1-03: CRUDE AND vAcuuM uNtr (NAPHTHEN|G ActD coRRostoN): Ralph Blee (Exxon) - commented that during a PMI inspection two short lengths of stainless steel were found at the inlet to a crude column and a vacuum column, welded to Cr-Mo transfer lines. These lines had been in service approximately 30 years and both were cracked in the weld HAZ area. The piping had never been soda ash washed; however, it was not

97c5.1-04: GRUDE AND vAcuuM uNtr (NAPHTHEN|G ActD coRRostoN): Kevin Ganschow (Chevron) - described some experiences in crude units many years ago with naphthenic-acid{ype corrosion on the undersides of trays and in the vapor space of columns. He asked whether anybody tried or had any experience using corrosion probes in the vapor
space of a column to take corrosion rate measurements in that area.

97c5.1-05: CRUDE AND vAcuuM uNlr (NAPHTHEN|C AcrD coRRostoN): Dannie Clarida (Conoco) - commented he has no experience with corrosion monitoring
devices installed in a vacuum tower, but there has been experience with a vacuum tower clad with 410S in an area where naphthenic acids caused corrosion in both liquid space and vapor space, in the packed section as well as some trayed areas. This is not exactly an answer to the question, but certainly corrosion in the vapor space areas was observed.

condensation on the corrosion probe because naphthenic acid corrosion does not occur in the vapor phase itself,

97c5.1-06: CRUDE AND vAcuuM uNlr (NAPHTHEN|G AcrD coRRostoN): Keith Lewis (Shell lnternational) - asked Kevin whether he was going to try to force

condensation to occur on the probe.

97c5.1-07: GRUDE AND vAcuuM uNlr (NAPHTHEN|G ActD GoRRostoN): Kevin Ganschow (Chevron) - responded that he was thinking about the possibility of an aircooled type of probe so that the end of the probe is cooled such that it causes some type of

97c5.1-08: GRUDE AND vAcuuM uNlr (NAPHTHEN|C ActD coRRostoN): Bill Fahey (Petrolite) - commented that some years ago he saw a probe consisting of a hollowed-out cylindrical coupon through which wash oil, or if it was an overhead, water, was injected to cool the probe down. He tried it but had problems with packing, so it didn't really work well, but the idea was to put a coolant through and then let the coolant enter into the
process.

Industry

corrosion cracking. More I types of units. quick specifically, he asked whether soda ash wash is being used in these A I survey was made of the audience: approximately 20 indicated taking precautions at all; approximately 4 indicated that some precautions are taken if the material no I is believed to be sensitized; no one indicated that they always take precautions. ii.--

98C5.1-07: GRUDE AND VACUUM UNIT (NAPHTHENTC ACD CORROSTON): Andy Gysbers (lmperial Oil) - commented that equipment in crude distillation units is likely to be coated with a thick resid material, which in turn poses less risk of a polythionic acid stress corrosion cracking problem. lf this material is removed, there could be a problem. 98G5.1-08: GRUDE AND VACUUM UNIT (NAPHTHENTC AC|D CORROSTON): Keith Lewis (Shell lnternational) - commented on one case in which soda ash washing was used in a crude unit stainless steel furnace in conjunction with mechanical pigging to remove coke deposits. ln this case the soda ash solution was used as the pigging fluid to protect the base material after the coke was removed. 98C5.1-09: GRUDE AND VACUUM UNIT (NAPHTHENTC AC|D CORROSTON): Kirk Baker (Chevron) - described a similar situation to the one mentioned by Keith. His case was a resid stripper furnace with type 317L SS roof tubes and the skin temperature was 538'C (1,000"F). These tubes were mechanically cleaned by pigging and left open for inspection for an extended period of time. For this reason a soda ash wash was applied. 98C5.1-10: CRUDE AND VACUUM UNIT (NAPHTHENTC AC|D CORROSTON): Jorge Hau (PDVSA) - asked whether anybody had experienced polythionic stress corrc-' crakinsinvacuumtower.

rherewerenoresponses

I I I I f

98c51-1 1: CRUDE AND vAcuuM uNtr (NAPHTHEN|G ActD coRRostoN): Tom Farraro (Citgo Petroleum) - commented that they do apply the soda ash wash k stainless steel piping in their vacuum units because of some past problems due to polythionic acid stress corrosion cracking. However, he added that in one case they did experience chloride stress corrosion cracking after using the soda ash, due to leaving soda ash solution in

:::::3

to* spots.

98c5.1-12: GRUDE AND vAcuuM uNtr (NAPHTHEN|C ActD coRRostoN): Jorge Hau (PDVSA) - asked why people are using "L" grade materials for nonwelded
components such as structured packing. There were no responses.

98c5.1-13: GRUDE AND vAcuuM uNtr (NAPHTHEN|G ActD GoRRostoN): Keith Lewis (Shell lnternational) - referred to some early laboratory work pubiished by Joerg Gutzeit that showed that austenitic stainless steels in the sensitized condition ar more susceptible to naphthenic acid corrosion at high TAN numbers. His question was whether anyone had seen examples from plant conditions in which accelerated naphthenic acid attack has occurred in materials known to be sensitized. His particular interest is a hydrocracker feed furnace that could have sensitized type 321H SS tubes and be exposed toeed with a TAN 1 to 2.

Industry

I I I

i-rr"u.r-14:

.RUDE AND vAcuu' uNrr (NA'HTHENT. AcrD Cathy Shargay (Fluor Daniel) - responded by saying that she knew of one case in a vacuum unit in which a type 316 SS transfer line was sensitized and did suffer from stress corrosion cracking externally, but there was no internal accelerated attack due to naphthenic acid.

GoRRosroN): I

98c5.1-15: GRUDE AND vAcuuM uNtr (NAPHTHEN|G ActD GoRRostoN): Brian Hopkinson (PDVSA)- added that he had never used "L" grade SS materials in his plants and he had never experienced accelerated localized naphthenic acid attack due to sensitized
material.

to naphthenic acid attack of sensitized material.

98c5.1-16: CRUDE AND vAcuuM uNrr (NAPHTHEN|C ActD GoRRostoN): Tom Farraro (Citgo) - also commented that in several of their units they have a mixture of "L" and regular grade SS materials and they have never seen accelerated localized corrosion due

services.

98c5.1-17: GRUDE AND vAcuuM uNtr (NAPHTHEN|G ActD GoRRostoN): Elizabeth Babaian-Kibala (Nalco Exxon Chemicals) - asked what experiences people are having with high{emperature corrosion probes for naphthenic acid and corrosive crude

sulfidation problems with high-temperature E/R probes in these services.

98c5.1-18: GRUDE AND vAcuuM uNrr (NAPHTHEN|G ActD coRRostoN): Deyauan Fan (Shell Oil) - responded by saying that he has experienced both coking and

Felix Perdieus (Exxon) - responded that they have had good experience with eight corrosion probes installed in hot oil services in their distillation unit. The probes give reasonable estimate of corrosion rates. A couple of the probes have had to be replace because they've reached their end of life. The safety people in the refinery are becoming more convinced ihat this type of probe can be handled safely.

98c5.1-19: GRUDE AND vAcuuM uNtr (NAPHTHEN|G ActD coRRostoN):

98c5.1-20: GRUDE AND vAcuuM uNtr (NAPHTHEN|G ActD coRRostoN): Walt Jenkins (Techsoft) - commented that his previous employer had 10 to 1S probes installed in high-temperature service. These probes did give reliabie results. He added that it is necessary to have special tools to pull these probes out of service. Probes were preferred to coupons because they had to be removed much less frequently. When asked about using different types of access fittings for removing and reinstalling probes, Walt commented that in most services they used a standard valve packing arrangemnt; however, in high-temperature high-pressure services they might consider using the more expensive cultom-designed special access fittings.
Mike Nugent (Tosco) - added that that in the CORROSION/98 paper no. ST7, "xperience with Naphthenic Acid Corrosion in Low Tan Crudes," he describes very good expeiience with g probes operating at about 260"C (500'F) for a period of six years. They now have 11 probes installed in this service. The readings from the probes correlate very weilwith the resulis from corrosion coupons and with their inspection experience.

98c5.1-21: GRUDE AND vAcuuM uNtr (NAPHTHENIC ActD coRRostoN):

Industry

t"-' I eoocs.2.04: cRUDEDISTILLATIoNTAcUUMUNITS:

Rob Smith (Marathon Asland) commented that they alloyed up areas to resist acid corrosion, but were concerned about cl-scc. He commented that previous REFIN'COR minutes did not contain a great deal of information relating to this particular concern. He wondered if that was because people had not experienced these failures, or everybody understood that they were chloride stress corrosion cracking and did not report them. He suspected that this may be the situation. He was looking for any experiences people actually had, whether they were related to shutdowns and/or to desalter upsets. lf so, were there any special measures to prevent such failures by alloying further up or special washing or neutralization?

naphthenic

if

2004C5.2-05: CRUDE DtSTtLLATIONruACUUM

UNTTS:

Tom Farraro (Stress Engineering) commented that they did have one vacuum unit that had a solid 304-SS transfer line. The furnace outlet operated at about 399oc (750'F). lt operated for about 20 years when they were not running any naphthenic acid crude. when they started running some naphthenic acid crudes the line failed. However, it did not fail from naphthenic acid or sulfidation corrosion, but rather cl-scc. some condensation occurred during an outage which accumulated in the low points in the line. The acidic salts then concentrated during startup and caused stress corrosion cracking. lt was subsequently replaced with a SS clad line. For a number of years, they also experienced numerous failures on solid SS piping in hydrotreaters and hydrocrackers, as well as, crude units. ln almost every case, the cause of the failure was not during normal operations. lt was always due to some type of upset or something that was done improperly during a shutdown or a start up. ln one case, a desalter upset resulted in extensive chloride stress cracking in the crude preheat train. The train was operating in the 288 to 31soc (550 to 600"F) temperature range and normally there should not have been any water present, therefore, chloride stress cracking should not have occurred. They managed to upset the desalter with sufficient severity that a slug of water went through the system. As a result, they had dozens of cracks. He recommended, whenever ossible, not to use solid austenitic stainless steel in a refining environment, but to use clad instead. lf solid alloys have to be used, use alloys other than 300-ss. one problem on clad lines is that it may not be possible to use clad on small-bore connections. ln such cases, he recommended to upgrade the small connections to Alloy-625 or Alloy-C, which are resistant to chloride cracking. He could not recall a single solid stainless steel piping system in their refineries that did not have a failure due to chloride stress ciacking at one time or another. Every single solid ss piping that he had in evry unit, regardless of the service, ended up with at least one failure due to chloride stress cracking. For some reason it was unforeseen or unexpected.

2004C5.2-06: CRUDEDISTILLATION/VACUUMUNtTS: cathy shargay (Fluor Daniel) commented that they always prefer clad
construction. However, solid construction may be more economical at times.

2004C5.2-07: CRUDE DISTILLATIONruACUUM UNTTS: Joerq Gutzeit (Consultant) also pointed out the there also exists an external CL-SCC risk for solid SS construction. 2004G5.2-08: CRUDEDISTILLATIONruACUUM UNTTS: Bill Neill (Corrosion & Materials Technology lnc.) commented that he was
working with a refinery on their first titanium bundle. This refinery's experience was very favorable with rolling the tube ends into the tube sheet. However, this particular client had already gotten a bid to weld the tube ends. He asked if people have an approach as far as number of grooves and the number of weld passes when they weld titanium tubes into the tube sheet.

2004C5.2-09: CRUDE DISTILLATIONruACUUM UNTTS: Andy Gysbers (lmperial Oil) commented that it was their practice to sea weld all titanium tubes. Their experience had been not to rely on mechanical rolls. He had seen bundles which passed the shop hydrotest, shipped from shops, and arrive in the field with enough relaxation to leak on the in-service hydrotest. Their practice was to then seal-weld all titanium tube bundles (in addition to mechanical roll). They will accept autogenous seal welding. 2004C5.2-10: CRUDE DISTILLATIONruACUUM UNTTS: Dick Horvath (Shell Global Solutions) asked if the seal weld is done before or after the mechanical rolling. 2004C5.2-11: CRUDE DISTILLATIONruACUUM
UNTTS:

Andy Gysbers (lmperial Oil) commented that they seated the tubes, did the seal welding, and then fully rolled the tube. Otherwise, there is a concern of creating a sealed gap causing pinhole leaks through the weld.

2004C5.2-12: CRUDE DISTILLATIONruACUUM UNTTS: Deyuan Fan (shell Global solutions) commented that Jim McMaster published a paper about the various tube-to-tube sheet joint designs for titanium heat exchangers and the pros and cons of seal welding and
mechanical rolling. The reference is below.

Jim McMaster, "Titanium for Mechanical Equipment in lndustrial Corrosion Service," page 34, "lndustrial Appll "
Symposium of ASTM Committee B-10 on Reactive and Refractory Metals and Alloys, New Orleans, LA, Oct 1979.

2004C5.2-132 CRUDE DISTILLATIONruACUUM UNTTS: Dana williams (Marathon Ashland) asked a question related

to

crude