ModelingandSimulation
TransientAnalysisCaseStudy
i
l i
S d
ofaLiquidsGatheringSystem
of a Liquids Gathering System
Mr.AugustoGarcia
Mr.ThomasMoore
Mr.
SouthwestResearchInstitute
Description&Outline
Numerous failures of a flexible composite pipeline
Liquids gathering system
Transient hydraulic model to evaluate a 4-inch flexible
composite
it pipeline
i li and
d an 8-inch
8 i h steel
t l pipeline
i li
Introduction
DataCollection
ComputationalModeling
Results
Analysis
y
Summary&Conclusions/Questions
GatheringSystemMap
CDP-01, -02, -03, and
Stabilizer plant
CDPs identical in
configuration with 2 triplex
pumps (PDP)
~12.5 Miles from CDP-01
to stabilizer plant
p
Stabilizer plant PCV holds
500 psig back pressure
Check valve half way up
hill
~500
500 ft from CDP-01
CDP 01 to top
of hill
ModelingProcess
Data
Collection
Functional & process
block diagrams
Computational model
d l
development
t
Computational model
validation
Obtain model results
Analysis
Verify data is consistent
Collect additional data and make
corrections as needed
Document assumptions
Verify data is consistent and appropriate
for modeling software
Ensure predictive measures are consistent
and appropriate for intended analysis
Check that the units are correct
Use known operating parameters from
data collection phase if possible
Ensure results are consistent with known
operating conditions
DataCollection
Documentation
P&IDsandisometricdiagrams
Datasheets
Operationalphilosophy
Reports
p
Anecdotaldata
Failuresappearedtoincreaseaftercheckvalve
installed
Flexible Pipe
FlexiblePipe
Structural
Components
FieldSurvey
Verifieddocumentationandcollectedadditional
information
InstrumentedallthreeCDPs
Did not go to stabilizer plant
Didnotgotostabilizerplant
TransientHydraulicModeling
UtilizedtheStonerPipelineSimulator(SPS)
SlightlycompressibleliquidsEOS
Slightly compressible liquids EOS
8-Inches Pipeline System
CDP SP16-18: P-7310 47%, M15-06: P-6310 23%, M14-16: P-4310 78%
Pressure
Elevation
700
7800
7700
600
7600
500
Pressure (psig)
7500
400
7400
7300
300
7200
200
7100
100
7000
6900
0
10000
20000
30000
40000
Pipeline Distance (ft)
50000
60000
70000
8InchPipelineSystemPressurewithPulsation
Max. Pressure
Min. Pressure
Pulsation
700
35
600
30
500
25
400
20
300
15
200
10
100
0
0
10000
20000
30000
40000
Pipeline Distance (ft)
50000
60000
0
70000
Pulsation (psig)
Pressu
ure (psig)
Pressure
SummaryResults
Theresultsdidnotrevealaproblemrelatedtowater
hammer in either the pre existing or existing pipelines
hammerineitherthepreexistingorexistingpipelines
Thepulsationswerealittlehighinthepreexisting
pipeline but were not considered excessive
pipeline,butwerenotconsideredexcessive
Transientsduringpumpstartandstopswere
significant but were well within the limits of the
significantbutwerewellwithinthelimitsofthe
pipeline
Aninvestigationoffailedcompositepipelinesections
An
investigation of failed composite pipeline sections
revealedthefailureoccurredfromtheinsideoutward
Therewasnocreep,andadjacentsectionswereburst
There
was no creep and adjacent sections were burst
testedtogreaterthan2timestheMAOP
WaterHammerAnalysis
Waterhammernotaproblem
Flowvelocitiesarelowandthereforethemomentum
changeislow
Nomechanismfoundforcausinganysignificantwater
hammer issues
hammerissues
410
2-Inch Pipeline Pressure
Averaged Pressure
390
370
Press
sure (psig)
2-in pipe RampRamp-up
from 20%20%-100%
350
330
310
290
0
8
Time (s)
10
12
14
16
TransientResponseAnalysis
Startup&shutdowntransientsnotaproblem
Severaltestcaseswererunwhichresultedin
significanttransientresponsesbutwerenot
g
p
enoughtoleadtoapipelinefailure
Therewasnosignificantresponserelatedtothe
There was no significant response related to the
additionofthecheckvalvehalfwayupthehill
Pulsation while small may play a significant part in
Pulsationwhilesmallmayplayasignificantpartin
theflexiblecompositepipelinefailure
PulsationAnalysis
Pulsationsarestronglycorrelatedfailuresites
Pressure
Max Pressure
Max.
Min Pressure
Min.
Failures
Pulsation
700
30
600
25
500
400
4-Inch Pipeline
Pulsation and
Failure
C
Correlation
l ti
300
200
15
10
0
100
0
0
10000
20000
30000
40000
Pipeline Distance (ft)
50000
60000
0
70000
Pulsation (ps
sig)
20
Pressure (psiig)
FailureAnalysis
Analysisoffailedsegmentsrevealedfailurefrominside
outward
Innerlinereruptedthroughouterlinerwithlittleevidenceof
Inner
waterbetweenlayers
Hydrotestingofadjacentsegmentsmetorexceeded
Hydro
specifications
Gl fib
Glassfibersnearthe
th
innerlinertendedto
havejaggedbreaks
whereasfibersnear
theoutsidetended
to be clean breaks
tobecleanbreaks
FailureAssessment
Thereappearstobenoobviousfailuremechanism
Failuresoccursomewhatrandomlyandassoonasa
failure is fixed another one appears
failureisfixedanotheroneappears
However,thefailuresappeartocorrelateverywellto
th
theareaswiththehighestpulsation
ith th hi h t l ti
Thepipevendorwarnsthatthepipeisnotdesigned
foruseincyclicapplicationabove20%ofrated
pressure(Cyclicisdefinedasapproximatelyonceper
day)
FailureAssessment,continued
Wetheorizethefollowing:
Damage
Damageisincurredtotheinnerfibersduetothesmall
is incurred to the inner fibers due to the small
pulsations,potentiallyfrombendingstressintheglass
This
Thisgeneratesaweaknessinthepipewhichcauses
generates a weakness in the pipe which causes
increasingstressinthatarea
Individualfiberscontinuetobreakfurtherstainingthe
fibers continue to break further staining the
Individual
remainingfibers
y
Thefailureoftheouterfibersoccursuddenlywhenthe
tensilestrengthofthefiberisexceededleavingaclean
break
Astheglasslayerisweakenedtheinnerlinerbulges
outwardbeforeeruptingthroughtheouterliner
FailureAssessment,continued
FromandO&Mstandpoint,thiswouldappeartobe
somewhatlikeoperatingametalpipeabovethe
endurancelimitwithpulsation
Evenasmallpulsationcaneventuallyleadtopipeline
failure
Onceanfailedareaisrepairedthestressesbuildup
elsewhere
Summary&Conclusions
Failureslikelyduetosmallcyclicstressesinthe
glassfibers
Lessonlearned,verifythatyourassumptionsare
correctandreviewalldatainthefieldwhen
providedtheopportunity
Noevidencetosupportwaterhammerasanissue
f
foreitherthe4inchor8inchpipeline
h h
h
h
l
Therevisedoperatingphilosophyisasubstantial
i
improvementoverthatofthepreexistingsystem
h
f h
i i
Pulsationsarenotasignificantthreattothe
i
integrityofthe8inchsteelpipeline,butmaybea
i
f h 8i h
l i li b
b
significantfactorinthefailureofthe4inchpipeline
ModelingandSimulation
g
TransientAnalysisCaseStudyof
aLiquidsGatheringSystem
Li id G th i S t
