Copyright © TWI Ltd 2008 World Centre for Materials Joining Technology Palisade User Conference April 22 nd & 23 rd London Maximising the Net Present Value.

Copyright © TWI Ltd 2008World Centre for Materials Joining Technology

Palisade User ConferencePalisade User ConferenceApril 22April 22ndnd & 23 & 23rdrd

LondonLondon

Maximising the Net Present Value of Investment in the Maintenance of

Assets

- Ujjwal Bharadwaj, TWI Ltd, Cambridge, UK


Presentation OutlinePresentation Outline• Overview of TWIOverview of TWI• Risk ManagementRisk Management• Risk Based Asset Management Risk Based Asset Management • The Risk Based Approach to Plant The Risk Based Approach to Plant

MaintenanceMaintenance– Qualitative AssessmentQualitative Assessment– Quantitative AnalysesQuantitative Analyses– Risk Based OptimisationRisk Based Optimisation

• Benefits of using the RB methodologyBenefits of using the RB methodology• Some issues Some issues • QuestionsQuestions


Overview of TWIOverview of TWI


What is TWI?What is TWI?

•Independent Research & Independent Research & Technology Organisation for Technology Organisation for welding and joining related welding and joining related technologies technologies •Serves industrial member Serves industrial member companies/government companies/government

•Non-profit distributing and Non-profit distributing and limited by guarantee of limited by guarantee of membersmembers

•Derives from The Welding Derives from The Welding Institute and the British Institute and the British Welding Research AssociationWelding Research Association


Key FiguresKey Figures

• 530 + staff 530 + staff

• 300 graduate status 300 graduate status

• 3500 industrial members 3500 industrial members

fromfrom

• 66 countries world-wide66 countries world-wide

• 4 major UK locations4 major UK locations

• 4 overseas operations 4 overseas operations

and training basesand training bases


Where is TWI HQ?Where is TWI HQ?

On Granta Science Park 8 miles South of Cambridge


TWI – UK LocationsTWI – UK Locations

Cambridge

North East

Yorkshire

Wales


TWI OverseasTWI Overseas

• Operations/training in Malaysia, Brazil, Iran & China Operations/training in Malaysia, Brazil, Iran & China • Associates in Australia, France, Ukraine & USAssociates in Australia, France, Ukraine & US• Agents/presence in US, India, Korea, Japan, Agents/presence in US, India, Korea, Japan,

Indonesia, Italy, Saudi Arabia and the United Arab Indonesia, Italy, Saudi Arabia and the United Arab Emirates, Kazakhstan.Emirates, Kazakhstan.

Training base at Training base at MalaysiaMalaysia


Key Industry Sectors ServedKey Industry Sectors ServedAerospaceAerospace AutomotiveAutomotive ConstructionConstruction

ElectronicsElectronics MedicalMedical Oil & GasOil & Gas PowerPower

Equipment, Equipment, Consumables & MaterialsConsumables & Materials


Risk ManagementRisk Management• Risk:Risk:

– Combination of the probability of an event and its Combination of the probability of an event and its consequences (ISO/IEC Guide 73)consequences (ISO/IEC Guide 73)

• What is Risk Management?What is Risk Management?– Direction and control with regard to riskDirection and control with regard to risk– In a financial settingIn a financial setting

• Concerned with events that pose opportunities for gain Concerned with events that pose opportunities for gain as well as potential for lossas well as potential for loss

– Currency fluctuation, interest rates etcCurrency fluctuation, interest rates etc– In an engineering settingIn an engineering setting

• Risk is a combination of occurrence of harm and the Risk is a combination of occurrence of harm and the severity of that harm ( ISO/IEC Guide 51:1999 )severity of that harm ( ISO/IEC Guide 51:1999 )

• Harm is physical injury or damage to the health of Harm is physical injury or damage to the health of people, or damage to property or the environmentpeople, or damage to property or the environment

• Risk management process…Risk management process…


Risk Management Strategic Objectives

Risk AnalysisRisk IdentificationRisk DescriptionRisk Estimation

Risk Evaluation

Risk Assessment

Risk Mitigation

Risk Acceptance

Risk Communication

Monitoring

Mod

ificati

on

Form

al

Au

dit


Risk Based Asset ManagementRisk Based Asset Management• Asset Integrity Management Asset Integrity Management

– Ensure integrity of engineered systemsEnsure integrity of engineered systems– ‘‘Fit for purpose’ throughout asset Fit for purpose’ throughout asset

lifecyclelifecycle– Ability of an asset to perform required Ability of an asset to perform required

function effectively and efficiently whilst function effectively and efficiently whilst safeguarding life and the environmentsafeguarding life and the environment

• Objective: Maximize returns, Objective: Maximize returns, minimize risksminimize risks

– Safety, Health and Environmental (SHE) Safety, Health and Environmental (SHE) risksrisks

– Business risksBusiness risks


Risk Based Asset ManagementRisk Based Asset Management

• Attributes of a good RBM for asset Attributes of a good RBM for asset mgmtmgmt

– Consistent, Transparent and AuditableConsistent, Transparent and Auditable– Identify potential and active DMs of Identify potential and active DMs of

plantplant– Plan inspection, such that residual risk Plan inspection, such that residual risk

for each DM is within acceptable limitsfor each DM is within acceptable limits– Increased reliability, safety and Increased reliability, safety and

availabilityavailability– Reduced scope of work for shutdown Reduced scope of work for shutdown

inspectioninspection


Failure rate over asset life Failure rate over asset life cyclecycle

Wear out failures

Infant mortality failures Constant

random failures

Failure

rate

Life of the asset

Infant mortality failures


Failure rate over asset life Failure rate over asset life cyclecycle

Failure

rate

Operating Life

Infant Mortality

Stage

Useful LifeAgeing/ Final Stage


The Ageing Stage in a Plant’s The Ageing Stage in a Plant’s LifeLife

• Accumulated damage e.g.Accumulated damage e.g.– Thinning due to Corrosion/ ErosionThinning due to Corrosion/ Erosion– Fatigue due to Cyclic StressesFatigue due to Cyclic Stresses– Creep due to high temperatureCreep due to high temperature

• Remaining Life Assessment/ Remaining Life Assessment/ EstimatesEstimates

– Prediction for Prediction for replacement/maintenancereplacement/maintenance

– DeterministicDeterministic– Probabilistic Probabilistic

» to capture the uncertainty involvedto capture the uncertainty involved


The Risk Based Approach to Plant The Risk Based Approach to Plant MaintenanceMaintenance

• Risk is a combination of probability of an Risk is a combination of probability of an event and its consequence (API 580)event and its consequence (API 580)

• Step 1: Preliminary Risk Analysis of the Step 1: Preliminary Risk Analysis of the System System – By Qualitative Risk AnalysisBy Qualitative Risk Analysis– Identify high risk componentsIdentify high risk components

• Step 2: Detailed Risk Analysis of Step 2: Detailed Risk Analysis of identified System components identified System components – By Quantitative Risk AnalysisBy Quantitative Risk Analysis– Develop a probabilistic RL model of the Develop a probabilistic RL model of the

degradation mechanism degradation mechanism

• Step 3: OptimisationStep 3: Optimisation– Such that financial benefit is maximisedSuch that financial benefit is maximised


Qualitative Risk AnalysisQualitative Risk Analysis• System AnalysisSystem Analysis

– ETA using @Risk ETA using @Risk Precision tree for Precision tree for all sub componentsall sub components

– System boundaries, System boundaries, failure criteria failure criteria specifiedspecified

– Data usedData used• Historical (local)Historical (local)• Specific dataSpecific data• GenericGeneric• Expert OpinionExpert Opinion


Qualitative Risk AnalysisQualitative Risk Analysis

40Very High Low Foundation

48High Medium Tower

36Medium Medium Coupling

36Medium Medium Entire Nacelle

48Medium High Mech Control

36Medium Medium Axle/ Bearing

48Medium High Brakes

36Medium Medium Blades

24Low Medium Grid Connection

32High Low Hydraulic

48High Medium Generator

40Very High Low Entire Turbine

24Medium Low Yaw System

40Very High Low Gear Box

48Medium High Elec Control

Risk Score

Consequence

Likelihood Component

• qualitative analyses

• semi quantitative analyses

Risk Histogram using Risk Priority Measure (RPM)

0

20

40

60

Elec C

ontro

l

Gea

r Box

Yaw S

yste

m

Entire

Tur

bine

Gen

erat

or

Hydra

ulic

Grid

Con

nect

ion

Blade

s

Brake

s

Axle/ B

earin

g

Mec

h Con

trol

Entire

Nac

elle

Coupl

ing

Tower

Found

atio

n

Component number

RP

M


Quantitative Risk Analysis Quantitative Risk Analysis

• Qualitative Analysis highlighted a structural Qualitative Analysis highlighted a structural component as high risk.component as high risk.

• FMEA identified FMEA identified three main damage mechanisms three main damage mechanisms

– Corrosion Corrosion – Scouring Scouring – Fatigue Fatigue

• Corrosion chosen to illustrate Risk Based Corrosion chosen to illustrate Risk Based approach approach

• Probabilistic Corrosion Model gives failure rate Probabilistic Corrosion Model gives failure rate with respect to years of service.with respect to years of service.

Failure Mode Effect Consequence category Consequence Cost (£k) Source/commentMaintenance durationRegulator penaltiesOther turbinesSea vehicleslocal structuresInjuryDeathInstallation of new structureRepair and recommissionInsurance premiumTechnology confidence lossLost productionMaintenance cost

Keep in serviceMaintenance durationRegulator penaltiesOther turbinesSea vehicleslocal structuresInjuryDeathInstallation of new structureRepair and recommissionInsurance premiumTechnology confidence lossLost productionMaintenance cost

Keep in serviceMaintenance durationRegulator penaltiesOther turbinesSea vehicleslocal structuresInjuryDeathInstallation of new structureRepair and recommissionInsurance premiumTechnology confidence lossLost productionMaintenance cost

Keep in service

Personnel

Maintenance costs

Reputation

Repair

Corrosion - Uniform Band in

Splash Zone

Collapse

Find and Assess

Collapse

Scouring

Fatigue - Circumfrential

cracking

Collapse

Find and Assess

Find and Assess

Reputation

Repair

Production loss

Secondary Damage

Personnel

Maintenance costs

Reputation

Repair

Production loss

Secondary Damage

Production loss

Secondary Damage

Personnel

Maintenance costs


Quantitative Risk AnalysisQuantitative Risk Analysis

• Main input is distribution of Main input is distribution of corrosion rate (CR) reflecting the corrosion rate (CR) reflecting the uncertainty involved.uncertainty involved.

• Remaining life model:Remaining life model:• RL= (Tstart - MAT)/ CRRL= (Tstart - MAT)/ CR

Tstart= Starting wall thickness (mm)Tstart= Starting wall thickness (mm)

MAT=Minimum Allowable Thickness to maintain MAT=Minimum Allowable Thickness to maintain integrity (mm)integrity (mm)

CR=Corrosion Rate (mm/yr)CR=Corrosion Rate (mm/yr)

• Probability of failure is P(RL<0)Probability of failure is P(RL<0)



• Corrosion rate estimateCorrosion rate estimate• Curve fitted over historical values Curve fitted over historical values

using @Riskusing @Risk• Expert opinionExpert opinion

Normal(0.40000, 0.1)Trunc(0,+inf)

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

-0

.1

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

>5.0% 90.0%

0.2355 0.5645

Typical Corrosion Rate of Steel, mpy

Mud Line

Quiet Sea Water

Low Tide

High Tide

Splash Zone

Marine Atmosphere



Material Process conditions Current thick ness

MONTE CARLOSIMULATION

Remaining li fe, years

Fre

qu

enc

y


Pro

babi

lity

Material Process conditions Current thick nessMaterialMaterial Process conditionsProcess conditions Current thick nessCurrent thick ness

MONTE CARLOSIMULATION


Fre

qu

enc

y


Fre

qu

enc

y


Pro

babi

lity


Pro

babi

lity

•Repeated sampling of values from input distributions gives PoF based on RL for the considered DM.





Probability of Failure (PoF) Vs Time

0

20

40

60

80

100

1990 2000 2010 2020 2030

Year

Po

F (

%)

Probability of Failure VsTime

%age Remaining Life consumed Vs Year of Operation

0

20

40

60

80

100

1995 2000 2005 2010 2015

Years of Operation

%ag

e R

emai

nin

g L

ife


Risk Based Optimisation (1/3)Risk Based Optimisation (1/3)• Risk expressed in Expected Values (EV)Risk expressed in Expected Values (EV)• EV= PoF x CoFEV= PoF x CoF• Consequences of failure mainly lost Consequences of failure mainly lost

production production • The Optimisation model finds the time of The Optimisation model finds the time of

replacement of the plant when the Net replacement of the plant when the Net Present Value (NPV) is the maximum Present Value (NPV) is the maximum over planning periodover planning period

• The optimisation weighs the EV of The optimisation weighs the EV of replacement with the EV of not replacing replacement with the EV of not replacing to identify the optimum year of to identify the optimum year of replacement. replacement.


Risk Based Optimisation (2/3)Risk Based Optimisation (2/3)

Probability of Failure (PoF), Action NPV Vs Years of Operation

0

10

20

30

40

50

60

70

80

90

2000 2005 2010 2015 2020

Years of Operation

Pro

ba

bility o

f F

ailu

re

(Po

F)

%

012345678910

Actio

n N

PV

Probability of Failure (PoF) % Vs Years of Operation

Action NPV Vs Years of Operation


Risk Based OptimisationRisk Based Optimisation

Optimized Action Schedule (w/o budgetary constraints)

Description Action Year Capital Cost NPV

Str 1 2010 -4.072237 -0.548586Str 2 2009 -3.878321 -0.62253Str 3 2005 -3.190704 -0.283258Str 4 2009 -3.878321 0.15813

Total NPV -1.296244

(3/3)(3/3)

Optimized Action Schedule (with budgetary constraints)

Description Action Year Capital Cost NPV

Str 1 2010 -4.072237 -0.548586Str 2 2006 -3.350239 -0.647286Str 3 2003 -2.894063 -0.32284Str 4 2009 -3.878321 0.15813

Total NPV -1.360582


Potential Benefits from using the Potential Benefits from using the Risk Based MethodologyRisk Based Methodology

• Target SHE and business risksTarget SHE and business risks• Maximize return on investment in Maximize return on investment in

O&M by risk prioritisingO&M by risk prioritising• Better understand DMsBetter understand DMs• Better control and prevent Better control and prevent

unexpected system outagesunexpected system outages• Identify and eliminate gaps in Identify and eliminate gaps in

existing integrity mgmt process at a existing integrity mgmt process at a sitesite

• Provide an auditable path for Provide an auditable path for integrity mgmtintegrity mgmt


Some issuesSome issues• Limitations of the optimisation toolLimitations of the optimisation tool

For complex systems, non-linear optimisation For complex systems, non-linear optimisation tools required.tools required.

Increasing dependencies require more computing Increasing dependencies require more computing power.power.

• Limitations of the methodologyLimitations of the methodology Used mainly in the Ageing phase when time Used mainly in the Ageing phase when time

dependent damage has accumulated. Needs to dependent damage has accumulated. Needs to be used in conjunction with an overall strategy for be used in conjunction with an overall strategy for plant/equipment life management.plant/equipment life management.

More suitable to business critical systems as More suitable to business critical systems as opposed to safety critical systems.opposed to safety critical systems.

• Availability and quality of dataAvailability and quality of data


QuestionsQuestions

Copyright © TWI Ltd 2008 World Centre for Materials Joining Technology Palisade User Conference April 22 nd & 23 rd London Maximising the Net Present Value.

Documents