ComputationalPipelineMonitoringforLiquids

APIRECOMMENDEDPRACTICE1130

SECONDEDITION,APRIL2022

AmericanPetroleumInstitute

CopyiohtAmericanPatroeumIhabte

SpecialNotes

APlpublicationsnecessarilyaddressproblemsofageneralnature.Withrespecttoparticularcircumstances,local,state,andfederallawsandregulationsshouldbereviewed.TheuseofAPlpublicationsisvoluntary.Insomecases,thirdpartiesorauthoritieshavingjurisdictionmaychoosetoincorporateAPIstandardsbyreferenceandmaymandatecompliance.

NeitherAPInoranyofAPI'semployees,subcontractors,consultants,committees,orotherassigneesmakeanywarrantyorrepresentation,eitherexpressorimplied,withrespecttotheaccuracy,completeness,orusefulnessoftheinformationcontainedherein,orassumeanyliabilityorresponsibityforanyuse,ortheresultsofsuchuse,ofanyinformationorprocessdisclosedinthispublication.NeitherAPInoranyofAPI'semployees,subcontractors,consultants,orotherassigneesrepresentthatuseofthispublicationwouldnotinfringeuponprivatelyownedrights.

APlpublicationsmaybeusedbyanyonedesiringtodoso.EveryefforthasbeenmadebytheInstitutetoensuretheaccuracyandreliabilityofthedatacontainedinthem;however,theInstitutemakesnorepresentation,warranty,orguaranteeinconnectionwiththispublicationandherebyexpresslydisclaimsanyliabilityorresponsibilityforlossordamageresultingfromitsuseorfortheviolationofanyauthoritieshavingjurisdictionwithwhichthispublicationmayconflict.

APlpublicationsarepublishedtofacilitatethebroadavailabilityofproven,soundengineeringandoperatingpractices.Thesepublicationsarenotintendedtoobviatetheneedforapplyingsoundengineeringjudgmentregardingwhenandwherethesepublicationsshouldbeused.TheformulationandpublicationofAPlpublicationsisnotintendedinanywaytoinhibitanyonefromusinganyotherpractices.

AnymanufacturermarkingequipmentormaterialsinconformancewiththemarkingrequirementsofanAPIstandardissolelyresponsibleforcomplyingwithalltheapplicablerequirementsofthatstandard.APIdoesnotrepresent,warrant,orguaranteethatsuchproductsdoinfactconformtotheapplicableAPIstandard.

Allrightsreserved.Nopartofthisworkmaybereproduced,translated,storedinaretrievalsystem,ortransmittedbyanymeans,electronic,mechanical,photocopying,recording,orotherwise,withoutpriorwrittenpermissionfromthepublisher.ContactthePublisher,APIPublishingServices,200MassachusettsAvenue,NW,Suite1100,Washington,DC20001-5571.

Copyright◎2022AmericanPetroleumInstitute

ii

CopyriontAmericanPatoeunhantse

CopyriontAmericanPatoeunhantse

Foreword

NothingcontainedinanyAPIpublicationistobeconstruedasgrantinganyright,byimplicationorotherwise,forthemanufacture,sale,oruseofanymethod,apparatus,orproductcoveredbyletterspatent.Neithershouldanythingcontainedinthepublicationbeconstruedasinsuringanyoneagainstliabilityforinfringementofletterspatent.

Theverbalformsusedtoexpresstheprovisionsinthisdocumentareasfollows.

Shall:Asusedinastandard,"shall"denotesaminimumrequirementtoconformtothestandard.

Should:Asusedinastandard,"should"denotesarecommendationorthatwhichisadvisedbutnotrequiredtoconformtothestandard.

May:Asusedinastandard,"may"denotesacourseofactionpermissiblewithinthelimitsofastandard.

Can:Asusedinastandard,"can"denotesastatementofpossibilityorcapability.

ThisdocumentwasproducedunderAPlstandardizationproceduresthatensureappropriatenotificationandparticipationinthedevelopmentalprocessandisdesignatedasanAPIstandard.QuestionsconcerningtheinterpretationofthecontentofthispublicationorcommentsandquestionsconcerningtheproceduresunderwhichthispublicationwasdevelopedshouldbedirectedinwritingtotheDirectorofStandards,AmericanPetroleumInstitute,200MassachusettsAvenue,NW,Suite1100,Washington,DC20001.Requestsforpermissiontoreproduceortranslatealloranypartofthematerialpublishedhereinshouldalsobeaddressedtothedirector.

Generally,APIstandardsarereviewedandrevised,reaffirmed,orwithdrawnatleasteveryfiveyears.Aone-timeextensionofuptotwoyearsmaybeaddedtothisreviewcycle.StatusofthepublicationcanbeascertainedfromtheAPIStandardsDepartment,telephone(202)682-8000.AcatalogofAPlpublicationsandmaterialsispublishedannuallybyAPI,200MassachusettsAvenue,NW,Suite1100,Washington,DC20001.

SuggestedrevisionsareinvitedandshouldbesubmittedtotheStandardsDepartment,API,200MassachusettsAvenue,NW,Suite1100,Washington,DC20001,standards@.

i

Contents

Page

1Scope 1

2NormativeReferences 1

3Terms,Definitions,Acronyms,andAbbreviations 2

3.1TermsandDefinitions 2

3.2AbbreviationsandAcronyms 6

4TechnicalOverview 7

4.1SelectionConsideration 7

4.2CPMSystemFeatures 7

5InfrastructureSupportsforCPM 9

5.1FieldInstrumentationandMeasurement 9

5.2Communications 11

5.3SCADA 12

5.4IntegrationofCPMandSCADA 13

5.5DataHistorian 14

6CPMOperation,Maintenance,andTesting 14

6.1CPMOperations 14

6.2SystemTesting 16

6.3OperatingIssues 18

6.4CPMSystemDataRetention 19

6.5CPMDocumentation 19

6.6CPMControllerTraining 20

AnnexA(informative)DiscussionofCPMThresholds 2

AnnexB(informative)DescriptionofTypesofInternalBasedCPMSystems 24

AnnexC(informative)MetricsandOtherPertinentTextfromAPIPublication1155 27

Bibliography 36

Figures

A.1CPMReleasesandTechniques 22

B.1CPMSystems 24

C.1GeneralizedExampleoftheSoftware-basedLeakDetectionProcess 28

C.2ExamplesofSensitivityCurvesBasedonDifferentOperatingThresholds.TheseExamplesareTypical

ofSystemsthatOperateonAccumulatedParameterErrors(e.g.VolumeBalance) 31

C.3ExamplesofSensitivityCurvesTypicalofEventOrientedSystems.SuchSystemsMightEmployPattern

RecognitionTechniquestoldentifytheOnsetofaLeak 31

C.4TabularFormatfortheRankingoftheLevelofImportanceforEachPerformanceMetric,andanOptional

TableforQualitativeorQuantitativeSpecificationofPerformanceCriteriaRelatedtoEachMetric………35

V

CopyriontAmericanPatoeunhante

ComputationalPipelineMonitoringforLiquids

1Scope

Thisrecommendedpractice(RP)focusesonthedesign,implementation,testing,andoperationofCPMsystemsthatuseanalgorithmicapproachtodetecthydraulicanomaliesinliquidpipelines.TheprimarypurposeofthesesystemsistoprovidetoolsthatassistPipelineControllersindetectingcommodityreleasesthatarewithinthesensitivityofanalgorithm.ItisintendedthattheCPMsystemprovideanalarmanddisplayotherrelateddatatothePipelineControllerstoaidindecision-making.ThePipelineControllerswouldundertakeanimmediateinvestigation,confirmthereasonforthealarmandinitiateanoperationalresponsetothehydraulicanomalywhenthealarmrepresentsanirregularoperatingconditionorabnormaloperatingconditionoracommodityrelease.

ThepurposeofthisrecommendedpracticeistoassistthePipelineOperatorinidentifyingissuesrelevanttothedesign,implementation,testing,andoperationofaCPMsystem.ThisRPisintendedforpipelinecontrollersandoperators,CPMsystemdevelopersandengineers,andothersinterestedinCPMsystemdesign,implementation,andoperation.

ThisRPincludesdefinitions,sourceandreferencedocuments,conceptsofdataacquisition,discussionofdesignandoperationofapipelineasrelatedtoCPM,fieldinstrumentationforCPMpurposes,alarmcredibility,PipelineControllerresponse,incidentanalysis,recordsretention,maintenance,systemtesting,training,considerationsforsettingalarmlimits,trending,andrecommendationsfordatapresentation.TherelationshipbetweenthePipelineControllerandtheCPMsystemisalsodiscussed.

Thisrecommendedpracticeiswritenforliquidonshoreoroffshoretrunklinesystems.CPMsystemshavetypicallybeenappliedtosteelpipelinesystems.CPMapplicabilityandperformancemaybelimitedbythecharacteristicsofnon-steelpipelines.

Thisrecommendedpracticewaswrittenconsideringsinglephase,liquidpipelines.Manyoftheprinciplesapplytoliquidpipelinesinintermittentslacklinefloworliquidpipelinesthatmayhavepermanentslacklineflow.ThisRPmaynotapplytothespecialcaseofdeterminingleaksduringshut-inconditionsthatoccurwhenthelineisshutdown(sometimescalledstaticconditions)unlessshut-inleakdetectionispartofthedeployedCPMsolution.

ItisrecognizedthatnosingleCPMmethodologyortechnologyissuitableforallpipelinesbecauseeachpipelinesystemisuniqueindesignandoperation.

Thisrecommendedpracticecomplementsbutdoesnotreplaceotherproceduresformonitoringtheintegrityoftheline.CPMsystemsareonepartofanoperator'sleakdetectionprogram.Forfurtherinformationonleakdetectionprograms,seeAPIRP1175.

2NormativeReferences

Thefollowingreferenceddocumentsareindispensablefortheapplicationofthisdocument.Fordatedreferences,onlytheeditioncitedapplies.Forundatedreferences,thelatesteditionofthereferenceddocumentapplies(includinganyaddenda/errata).

APIRP551,ProcessMeasurementInstrumentation

APITR1149,PipelineVariableUncertaintiesandTheirEffectsonLeakDetectability

APIRP1175,PipelineLeakDetection—ProgramManagement

CodeofFederalRegulations,49CFRPart195,TransportationofHazardousLiquidsbyPipeline

1

CopyiohtAmericanPatroeumIhabte

2APLRECOMMENDEDPRACTICE1130

3Terms,Definitions,Acronyms,andAbbreviations

3.1TermsandDefinitions

Forthepurposesofthisdocument,thefollowingdefinitionsapply.

3.1.1

abnormaloperatingcondition

AconditionidentifiedbytheOperatorthatmayindicateamalfunctionofacomponentordeviationfromnormaloperationsthatmay:

a)Indicateaconditionexceedingdesignlimits;or

b)Resultinahazard(s)topersons,property,ortheenvironment.

3.1.2

accumulatordata

ASCADAdatavaluethatrepresentsanaccumulatedquantity,usuallyvolumeinliquidpipelineservice.

3.1.3

accumulatorfreeze

AfeatureofsomeSCADAprotocolsthatallowallvolumetricdatatobecapturedatvirtuallythesametime.

3.1.4

alarm

Avisibleand/oraudiblemeansofindicatingtothecontrolleranequipmentmalfunction,ananalogoraccumulationprocessdeviation,orotherconditionrequiringacontroller'sresponse.

3.1.5

alertalgorithm

ApartofaCPMsystemthatevaluatestheinferredmeasurements,comparesagainstthethresholds,andissuesaCPMalarm.

3.1.6

analogdata

ASCADAdatavaluethatrepresentssomemeasuredquantity,suchastemperatureorpressure.

3.1.7

analogdeadband

Aparameterthatdefinestheincrementofchangeinavaluethatissignificant.

NOTESee5.2.3forfurtherinformation.

3.1.8

calibration

Forthisdocument,theactivitiesinwhichinstrumentationandmeasurementareproved.

3.1.9

communicationfailure

AninterruptindataexchangebetweentheCPMsystemandtheRTU,PLC,orFlowComputer.

3.1.10

commodityrelease

leak

Anunintendedlossoffluidfromthepipeline.

CopyiohtAmericanPatroeumIhabtte

CoMPUTATIONALPIPELNEMoNITORINGFORLoUDs3

CopyiohtAmericanPatroeumIhabte

3.1.11

computationalpipelinemonitoring

CPM

AnalgorithmicmonitoringtoolthatalertsthePipelineControllertorespondtoadetectablepipelinehydraulicanomaly(perhapsbothwhilethepipelineisoperatingorshut-in)whichmaybeindicativeofacommodityreleaseandincludesanInferenceEngineandanAlertAlgorithm.

3.1.12

conservationofmass

Theprincipleasappliedtoliquidflowinpipelines,thatstatesthatthetimerateofmassinflowtoapipesegmentminusthetimerateofmassoutlowequalsthetimerateofmassincrease(decreaseisconsideredasanegativeincrease)inthepipesegment.

3.1.13

dataarchiving

ASCADAsystemfeaturethatrecordsdatainanhistoricaldatabaseundersomepre-defineddatamanagementprocess.

3.1.14

dataquality

ASCADAsystemfeaturethatcreatesstatusbitsthatareattachedtoreflectthevalidityofprocessdata

3.1.15

dragreductionagent

DRA

Anadditiveusedinliquidpipelinestoreducefrictionloss.

3.1.16

eventlog

ASCADAsystemfeaturethatcreatesapermanentrecordofchangestothepipelineandthesystem'sstateinchronologicalorder.

3.1.17

falsealarm

Acommodityreleasealarmwhich,afterinvestigation,wasnotcausedbyanactualcommodityrelease.

3.1.18

filter

Adeviceoralgorithmtoremoveunwantedcomponentsfromaprocesssignal.Alsocalledsignalconditioning.

3.1.19

fluidproperties

Thecharacteristicsofthefuidthatdescribeitshydraulicbehaviorincludingdensity;viscosity,compressibility(orbulkmodulus);coefficientofthermalexpansion;thermalcapacity.

3.1.20

historicaldata

Datathathavebeenretainedforlaterretrieval,typicallymaintainedbyaSCADAsystem'sdataarchivalsubsystem.

3.1.21

hydraulicanomaly

Anirregularoperatingconditiononthepipelineorabnormaloperatingconditionthatisexplainablethroughthesystemshydraulics.

3.1.22

inferenceengine

ApartoftheCPMsystemthataccumulatesdata,performscalculation,andprovidesoutputstothealertalgorithm.

4APIRECONMENDEDPRACTICE1130

CopyiohtAmericanPatroeumIhabtte

NOTEAdditionaldescriptionisofferedinAnnexB.

3.1.23

irregularoperatingcondition

Aninfrequenteventduringwhichthepipelinemaybeoperatedinawaythatisnotnormalandmayrequirere-tuningtheCPM.

3.1.24

leakdeclaration

ThedeclarationthatismadeifaPipelineControllerhasreasonstosuspectthatacommodityreleaseisoccurringonthepipeline.

3.1.25

linebalance

Comparisonofthemeasuredvolumeormassenteringthesystemtothemeasuredvolumeormassexitingthesystemormeter-to-meterreadingcomparisonusingconservationprinciples.

NOTE1Certaintypesoflinebalancearecommonlyreferredtoasmassbalanceormaterialbalance.

NOTE2AdditionaldescriptionisofferedinAnnexB.

3.1.26

manualdataoverride

Whenmanualentriesareinputinlieuofactualfielddatavalues.

3.1.27

negativepressurewaveCPM

ACPMsystemthatsensesthepressurewavesignalthatoccurswhenthepipewalliscompromisedandtheproductescapesthroughtheholeinthepipe.

3.1.28

noise

Anunwantedcomponentinaprocesssignal.

3.1.29

pipelinecontroller

controller

Apersonwhoisresponsibleforthemonitoringormonitoringanddirectcontrolofapipeline.

3.1.30

pressure/flowmonitoringCPM

ACPMsystemwhichexaminestherelationshipbetweenpressureandorflowchangesandappliesanalgorithmtodetermineiftheyindicateananomaly.

NOTEAdditionaldescriptionisofferedinAnnexB.

3.1.31

protocol

ThespecificationsofthemessagestructurebetweenRTUorPLCandControlCenterComputerarecollectivelyreferredtoasthecommunicationsprotocol.

3.1.32

rateofchange

ROC

Acalculatedvaluethatreflectsthechangeinananalogdatavalueperunittime.

CoMPUTATIONALPiPELINEMoNITORINGFORLouDs5

CopyiohtAmericanPatroeumIhabtte

3.1.33

realtimetransientmodel(RTTM)CPM

ACPMsystemwhichmonitorsinstrumentdataandphysicalcharacteristicsofthepipelineandfluidstransported,thenemployshydrauliccalculationstodeterminethein/outbalance,inventory,andinstantaneousfloworpressureinsegmentsofthepipeline,orboth.

NOTEAdditionaldescriptionisofferedinAnnexB.

3.1.34

returntonormal

Thetransitionfromalarmtonormalstatethatsignifiesthatanalarmconditionhasended.

3.1.35

remoteterminalunit

RTU

ASCADAsystemcomponent,typicallyinstalledatafieldsite,thatgathersprocessdatafromsensorsfortransmissiontotheControlCenterComputer.

3.1.36

report-by-exception

AfeatureofsomeSCADAcommunicationprotocolsthatintendstoimprovecommunicationefficiencybyreportingonlythedatathathaschangedsincethepreviouspoll.

3.1.37

supervisorycontrolanddataacquisition

SCADA

Thetechnologythatmakesitpossibletoremotelymonitorandcontrolpipelinefacilities.

3.1.38

scantime

ThetimeintervalbetweentwoconsecutivepollstoDataAcquisitionDevicesonaSCADAcommunicationchannel.

3.1.39

segments

Ashorterpartofapipelinesectionoftenboundedbyinstrumentationorotherphysicalfeaturesofthepipeline.

3.1.40

sensitivity

AcompositemeasureofthesizeofaleakthataCPMsystemiscapableofdetectingandthetimerequiredforthesystemtoissueanalarmintheeventthatacommodityreleaseofthatsizeshouldoccur.

NOTEThistermisfullydefinedanddiscussedinAnnexC.

3.1.41

shut-in

Thepipelinehydraulicconditionthatexistswhenfluidisnotenteringorleavingthepipebutmaybecontainedwithin.

3.1.42

singlephase

Afluidstate,eitherliquidorgaseous,baseduponcommodity,vaporpressure,pipelinepressureandtemperature.

3.1.43

slackline

Theconditionwhereapipelinesegmentisnotentirelyfiledwithliquidorispartlyvoid.Mayalsobecalledcolumnseparation.

6APLRECOMMENDEDPRACTICE1130

CopyiohtAmericanPatroeumIhabtte

3.1.44

statisticalanalysisCPM

AmathematicalmethodofhandlingtheCPMrelatedinstrumentoutputsfromthepipeline.StatisticalanalysisCPMsystemscanuseeitherConservationofMasstechniquesorSignatureRecognitiontechniquesorbothtechniques.

NOTEAdditionaldescriptionisofferedinAnnexB.

3.1.45

statusdata

ASCADAdatavaluethatrepresentstheoperationalstateofanitemoffieldequipment

3.1.46

steadystateconditions

Thepipelinehydraulicconditionthatexistswhenallthepipelineoperatingparametersremainnearlyconstantoveraperiod.

3.1.47

system

Anentireentitysuchasacompletepipeline.Segmentsareasubsetofasystem.

3.1.48

threshold

Anupperorlowerestablishedvalueforaparameterwhichmaybefixedordynamic.

3.1.49

timeskew

ThevariationinreportingtimesfromoneDataAcquisitionDevicestoanotherinapolledSCADAcommunicationsprotocol.

3.1.50

timetag

ASCADAsystemfeaturethatrecordsthetimethatameasurementoreventoccursalongwiththedata

3.1.51

transientconditions

transient

Thepipelinehydraulicconditionthatexistswhenpipelineoperatingparameterschangemeaningfullyoveraperiod.

3.2AbbreviationsandAcronyms

Forthepurposesofthisdocument,thefollowingabbreviationsandacronymsapplyCPMComputationalPipelineMonitoring

DRADragReductionAgent

PLCProgrammableLogicControllers

ROCRateofChange

RTTMReal-timeTransientModel

RTURemoteTerminalUnit

SCADASupervisoryControlandDataAcquisition

CopyiohtAmericanPatroeumIhabtte

CoMPUTATIONALPiPELINEMoNITORINGFORLouDs

7

4TechnicalOverview

ThissectiondiscussesthegenerictypesofCPMtechnologiesandapplications,providesalistofdesirableCPMfeatures,andmentionsimportantissuesconcerningthefluidstransported.

4.1SelectionConsideration

EachCPMmethodologycontainsdifferentcombinationsoffeatureswithvaryingdegreesofcapabilityandsophistication.Underappropriatecircumstances,commodityreleasedetectioncanbenefitbyemployingmultipleCPMtechniquesorapplicationsforvalidationorredundancy.Theindependenceoftechniquesusedinsomemethodologiespotentiallyallowsforindependentvalidationorredundancy.

CPMsystemsareinstalledtodetectpipelineleaks.BeforeinstallingaCPMsystem,thefollowingshouldbeconsidered:

—instrumentationcapabilities—communicationsreliability—pipelineoperatingcondition

—pipelinephysicalconstraintsandcharacteristics

-producttype

—commodityfluid/thermodynamicproperties—technicalmaturitylevel

4.2CPMSystemFeatures

ThefollowingisalistofdesirableCPMfeaturesandfunctionalitiesthatimproveperformanceoraddutility,orboth.

TheCPMfeatureslistedbelowarenotinanyordernoristhereanyattempttoweighttheimportanceofeach.Noonemethodologyorapplicationpossessesallthesefeaturesandcertainfeaturesmaybemoreappropriateforspecificpipelinesystems.

TheCPMsystemmay:

-BeabletoperformitsCPMfunctionswithexistingsensorsandinstruments(ordoesnothavespecialoradditionalrequirementsforinstrumentation).

—Beminimallyimpactedbycommunicationoutagesorbydatafailures,however,providealarmingbasedonadegradedmodeofoperationandduringanabnormaloperatingcondition(see6.1.1).

—Accommodatecomplexoperatingconditionsandbeconfigurabletocomplexpipelinenetworks—Beavailableandreliableduringtransients.

—Performanimbalancecalculationonmetersoveraconfigurablesetoftimeconstants.

-Possessdynamicalarmthresholds.

—Possessdynamicliquidpack.

8APLRECOMMENDEDPRACTICE1130

CopyiohtAmericanPatroeumIhabtte

—Accommodatecommodityblending.

—Accountforheattransfer.

—Providethepipelinesystem'srealtimehydraulicpressureprofile,recognizingMAOPandelevationviolations.—Beabletohandleslacklineconditions.

—Beavailableduringshut-ins

-Accommodateallliquidhydrocarbons.

-Canidentifythelocationofarelease.

—Candisplaypressuretrends.

—Allowprovisionstosubstitutemanualoveridetospecificstatusorvaluesduringperiodswheninputdatamaybeunavailable(municationoutage,measurementfailure,maintenance,etc.).

—Providedataattributesassociatedwithsupportingfieldinputsandcalculateddata.

—Identifytheleakrate.

-Accommodatecommoditymeasurementandinventorycompensationforvariouscorrectionfactors(temperature,pressure,density,meterfactor).

-Providebatchtrackingincludinginterfaceandanomalymarkers.Computebulkmodulusandperforminventorycompensation.

-Validatecommodityreleasealarmsusingredundantanalysiswithinthesamemethodorredundantanalysisbetweenmethods,orboth.

—Accountforeffectsofdragreducingadditive.

-Providethenecessarydocumentationandtrainingtosupportself-guidednavigation,maintenance,alarmdefinition,andtheoryofoperation.

-Vendor-providedtechnicalsupport

—Providetoolstosupportacontroller'sdecision-making.

-Provideaninterfacetoallowalarmingofastand-aloneCPMapplicationtobeintegratedintothePipelineController'sprimaryalarmprocessingsystem.

—ProvideaudittrailsofCPMactionstakenbyPipelineControllersandsystemdevelopers.

—Allowsself-testingwithoutaffectingperformancewhilethetestisunderway.

—Allowsforsystemhealthmonitoring.

DifferentCPMsystemshavedifferentfeatures,notallofwhicharelistedhere.Operatorsareadvisedtodoresearchandcomparesystemsbasedontheirneeds.

4.2.1PerformanceMetrics

SelectionofaCPMsystemforagivenpipelineinvolvesevaluationoftherequiredandexpected(orestimated)performanceofthesystem.Otheraspectssuchascommercialconsiderations(e.g.useofacommonsystemin

CoMPUTATIONALPIPELNEMoNITORINGFORLoUDs9

aControlCenterwithmultiplepipelines)oreconomiccriteria(e.g.capitalandoperatingcostoftheCPM)maybeconsideredbutthesearenotdiscussedherein.

Thefollowingcategorizesanddescribesperformancemetricsforselectionconsideration.Duringtheselection,theoperatormayplacemoreweightorimportanceononemetricoranother.Asystemshouldachieveasatisfactorybalancebetweenallfouroftheseperformancemetrics.ForamorecompletedescriptionofthesemetricspleaserefertoAnnexC.

-Reliability—ThemeasureoftheCPM'sabilitytorenderaccuratedecisionsaboutthepossibleexistenceofaleakonthepipelinewhileoperatingwithinanenvelopeestablishedbytheCPMdesign.Asystemisconsideredmorereliableifitconsistentlydetectsactualleakswithoutgeneratingfalsealarmsasdefinedbytheoperator'salarmmanagementplan/program.

-Sensitivity—Thecompositemeasureofthesizeofaleakthatasystemcandetect,andthetimerequiredforthesystemtoissueanalarmifaleakofthatsizeshouldoccur.

-Accuracy—Thevalidityofleakparameterestimatessuchaseakflowrate,totalvolumelost,typeoffluidlost,andleaklocationareindicationsofCPMaccuracy.

-Robustness—ThemeasureoftheCPM'sabilitytocontinuetofunctionandprovideusefulinformationevenunderchangingconditionsofthepipeline(i.e.transients)orinconditionswheredataarelostorsuspect.Asystemisconsideredrobustifitcontinuestofunctionunderlessthanidealconditions.

5InfrastructureSupportsforCPM

ACPMisnotastand-alonesystem.Itdependsuponfieldinstrumentation,communications,andmaybedependentonSupervisoryControl