piramid - bureau of safety and environmental piramid a quantitative risk-based approach to integrity...

PIRAMID

A Quantitative Risk-based Approach toIntegrity Maintenance Planning

and Design Optimization forOffshore Pipelines

Mark StephensC-FER Technologies Inc.

Edmonton, Canada

advantage technology

createsC-FERTechnologies Inc.

Background to the PIRAMID Project

l A Multi-year Joint Industry Program

l Sponsorsl BC Gas Utilityl Enbridge Inc. (Interprovincial)l Foothills Pipe Linesl Kinder Morgan (KN Energy)l Southern California Gasl TransCanada Pipelinesl Gas Research Institutel Canadian Geological Surveyl U.S. Minerals Management Service



Program Goal

Develop Models and Software to:

l Make optimal maintenance decisionsl Ensure acceptable risk levelsl At the lowest possible cost

l Explain rationale behind decisionsl Internally within companyl Externally to regulators and the public



Issues To Be Addressed

l What is the operating risk associated withthe pipeline in its present state

l What effect would each candidate maintenancestrategy have on the operating risk

l What is the lowest cost maintenance optionthat meets acceptable safety & environmentalconstraints



PIRAMID Functions

Divide system into segments

Rank segments according to risk level

Determine optimal strategy for targeted segment

Implement optimal actions in order of refined ranking

Repeat for eachtargeted segment

MaintenanceOptimization

RiskRanking

The PIRAMIDThe PIRAMIDFunctionsFunctions



Risk Ranking

For each Segment in System

Identify Hazards

EstimateTotal Operating Risk

Rep

eat

for

All

Haz

ard

s

EstimateProbability of Failure

EstimateConsequences of Failure



Maintenance Optimization

For each Targeted Segment

Identify IntegrityMaintenance Options

Estimate Effect of Maintenance Strategyon the Probability of Failure

Calculate Risk reduction

Select Optimal IntegrityMaintenance Strategy

Rep

eat

for

All

Op

tio

ns



PIRAMID Features

l Quantitative approachl Extensive use of engineering modelsl Calculates total risk

l financiall safetyl environmental

l Validated by real pipeline data



Probability Estimation Approaches

Information Analysis FailureProbabilities

Conditiondata

ReliabilityModels

Failure rates Adjustmentfactors

FailureProbabilities

General

Adjusted historical failure rate method

Structural reliability method

FailureProbabilities



Adjusted Historical Failure Rate Method

ijlkllijkl AFMFRfbRf ==

Failure rate forsection, mode and cause

1. Baseline failurerate per cause

3. Adjustment factorfor section and cause

2. Relative failureprobability per mode



Other

External Corrosion

Mechanical damage

Ground movement

Environmental cracking

Fabrication defects

Internal Corrosion

Baseline Failure Rates by Cause



Failure Rates by Mode

Failure Caus e Mode Factorsmall leak large leak rupture

Metal Lo s sCorro s ion

??? ??? ???

MechanicalDamage

??? ??? ???

GroundMovement

??? ??? ???

Cracks ??? ??? ???

OtherCaus e s

??? ??? ???



Probability Adjustment Factors

( ) CCCTCPSC FFFFTtAfAF ,, =

Adjustmentfactor

Operatingtemperature

Wall thickness

Line age

Soil corrosivityfactor

Cathodic protectionfactor

Coating typefactor

Coating conditionfactor

E.g. - for External Corrosion



Adjusted Failure Rates - Summary

l Simple models utilizing

l Statistical data

l Engineering analysis

l Judgment

l Suitable for segment ranking



Structural Reliability Approach

Load or ResistanceMean

ResistanceMean Load

Resistance DistributionLoad Distribution

Overlap Leadingto Failure Probability



Application to External Corrosion

Outer Diameter

(D)

Wall Thickness

(t)Hoop

Stress (S)

Corroded Length (L)

Maximum Corroded Depth (d)

Failure rate / km = No. Defects per km x Failure probability per defect



Failure Probability per Defect

Data on pipeproperties and

dimensions

Maximumoperatingpressure

InspectionData

Measurementuncertainties

Corrosion modeland test results

Data fromrepetitive

inspections

Failure probabilityas a

function of time

Model results

Test results

xxxx

xx

x

xx

x

x

modeluncertainties

Flaw depth (mm)

Prob. density

corrosioncharacteristics

Prob. density

Growth Rate (mm/yr)

corrosiongrowth rates

Yield stress (MPa)

Frequency

pipeproperties



Probability of Failure Versus Time

0.0001

0.0010

0.0100

0.1000

1 3 5 7 9 11 13 15 17 19

Time (years)

An

nu

al P

rob

abili

ty o

f F

ailu

re



Effect of Maintenance

l Mitigation philosophyl Find and eliminate defects before they reach critical size

l Maintenance optionsl Inspection and repairl Hydro-testing

l Maintenance impactl Reduce number of defects per unit line lengthl Shift defect size distribution toward smaller values



Quantifying Effect of Maintenance

Size oforiginal defects

Size ofundetected defects

Size ofdetected defects

Measured size ofdetected defects

Inspectiondetection power

Size of remainingdetected defects

Inspection sizingaccuracy

Repaircriterion

Size of allremaining defects



Modified Defect Size Distributions

0

0.2

0.4

0.6

0.8

1

0 1 2 3 4 5 6

Defect Depth (mm)

Pro

bab

ility

Den

sity

Detected1.25 MAOP1.5 MAOP



Effect on Probability

0.0001

0.0010

0.0100

0.1000

1 3 5 7 9 11 13 15 17 19

Time to Next Inspection (years)

An

nu

al P

rob

abili

ty o

f F

ailu

re p

er k

m

No repair1.25 MAOP1.5 MAOP



Application to Mechanical Damage

Outer Diameter

(D)

Wall Thickness

(t)Yield Stress (S)

Impact Energy (E)

Failure probability = (No. line hits) x (Failure probability per hit)



Probability of Failure Given Hit

Yield stress (MPa)

Frequency

Model results

Test results

xxxx

xx

x

xx

x

x

Data on steelproperties and

pipe dimensions

Puncture modeland

test results

Data onequipment

size and weight

Failure probabilitygiven hit

modeluncertainties

pipeproperties

Load (kN)

Frequency

outsideforce



Frequency of Line Hits

Fault Tree Model(inductive logic)

Basic event probabilities(a function of line attributes)

Pipeline hitby

dragged object

Seabedcontact with

dragged object

Failure ofon-bottomprotection

Vesselactivity on

sea surface

Draggedobject

deployed

Seabeddisturbance

exceedscover depth

Failure ofmechanicalprotection

Pipeline hitby

dragged object

Seabedcontact with

dragged object

Failure ofon-bottomprotection

Vesselactivity onsea surface

Draggedobject

deployedover

alignment

Seabeddisturbance

exceedscover depth

Failure ofmechanicalprotection

Similar fault tree forvessel hull grounding

Top event



Effect of Maintenance

l Mitigation philosophyl Prevent potential line hits

l Example prevention optionsl Enhance awareness of pipeline locationl Modify cover depth inspection frequencyl Increase pipeline burial depthl Introduce mechanical protection

l Prevention Impactl Modify fault tree basic event probabilitiesl Reduce hit probability



Effect on Failure Probability

An

nu

al p

rob

abili

ty o

f fa

ilure

per

km

Statusquo

Increased cover depthinspection frequency



Structural Reliability Approach - Summary

l Calculate failure probability from

l Structural behaviour models

l Line and ROW information

l Defect information (corrosion / SCC/ cracks / dent-gouges)

l External forces (ground movement / mechanical damage)

l Suitable for maintenance optimization



Consequence Estimation Approach

Category Measure

Financial impact DollarsPublic safety impact Number of people at riskEnvironmental impact Effective residual spill volume

Consequence measures



Consequences of Pipeline Failure

Failure

Hazard Models

Damage Thresholds

Human Impact Thresholds

Effective Shoreline Impact

Volume

Number of People at Risk

Spill Decay &Clean-up Models

Line RepairCost

Lost ProductCost

Service InterruptCost

Financial Cost

ShorelineImpact Model

OffshoreDamage

Cost



Consequence Analysis

Release

Immediate ignition

ignition

Delayed ignition

No ignition

Explosion

No explosion

Fire ball / Jet fire

Vapour cloudexplosion

Vapour cloudfire

Toxicvapour cloud

No immediate

Step 1 - Use event tree analysis (logic model) to estimate relativelikelihood of all conceivable release hazards

Consequences of Acute Release Hazards




Jet FireJet FireToxic VapourToxic VapourCloudCloud

Immediate Ignition

NoIgnition

VapourVapourCloud FireCloud Fire

Vapour CloudVapour CloudExplosionExplosion

DelayedIgnition

Step 2 - Use hazard characterization models to estimate size of affected areas

Consequence of Acute Release Hazards

Step 3 - Estimate offshore damage cost(no. involved structures & vessels x property value)

Estimate number of people at risk(no. involved structures & vessels x crew size)




Long-term Consequences of Product Release Hazard

SpillVolume

ShoreImpactVolume

Spill TrajectoryModel

ShorelineSensitivity Model

Spill DecayModel

OffshoreClean-up Model

ShorelineClean-up Model

Step 1 Step 2 Step 3

Step 4 - Assess clean-up costsEstimate degree of natural resource damage

SpillSpillVolumeVolume

ImpactImpactVolumeVolume

EffectiveEffectiveImpactImpactVolumeVolume



Spill Trajectory Model

D

C

B

A

3

1 2

4

A B C D

1

2

3

4

Shoreline

Lau

nch

Zo

ne

Trajectory Analysis Results

Probability of Impact: 0 to 1

Time to Impact:

days

#

Shoreline

Pipeline

Spill Trajectory Analysis



Software



PIRAMID Structure

SegmentRanking

RiskEvaluation

DecisionAnalysis

Pipeline Definition and SegmentationScenarioDefinition

Risk Estimation

ConsequenceEstimation

ProbabilityEstimation

Adjusted FailureRate Models

StructuralReliability Models



Segment Ranking Chart



Segment Ranking Table



Risk Variations Along a Segment



Failure Rate Versus Time



Decision Analysis - Utility Chart



Summary

l Comprehensive user-friendly approach forrisk-based integrity maintenance planning

l Quantitative and objective methodologyl Historical incident datal Analytical models for

l pipeline failure predictionl release hazard characterization

l Pipeline condition data

l Benefitsl Generates line-specific risk estimatesl Quantifies the impact of maintenance actionsl Identifies minimum cost solutions

piramid - bureau of safety and environmental piramid a quantitative risk-based approach to integrity...

Documents