Module 7 - Schedule Risk Analysis - Rev 0

The complete training program consists of the following modules: Module 1 - Introduction To Project Controls Module 2 - EPC Schedule Development including P6 user skills Module 3 - Servi

MODULE 7– SCHEDULE RISK ANALYSIS

MODULE 7 - SCHEDULE RISK ANALYSIS

0 Released for Global Implementation Ed Cimic Project Controls

R5 Management Team

Project Controls R5 Management Team

1 Nov 2011

This document has been prepared for the exclusive use of WorleyParsons

Copying this document without the permission of WorleyParsons is not permitted

SPECIAL ACKNOWLEDGEMENTS:

Mark Spanos,

“A pessimist because of intelligence, an optimist because of will”

NOTE: Review of Module 1– Introduction to Project Controls is required prior to studying this module.

This Training Module is part 7 of an 8 modular training program designed to provide the participants with an overall introduction to the skills & knowledge required by Project Controls when executing an EPCM / PMC project

The complete training program consists of the following modules:

Module 1 - Introduction To Project Controls

Module 2 - EPC Schedule Development (including P6 user skills)

Module 3 - Services Management (including InControl V8.0 user skills)

Module 4 - Commercial Performance Management

Module 5 - Introduction to TIC Cost Estimation

Module 6 - TIC Management (including Prediction Plus / InControl V10 user skills)

Module 7 - Schedule Risk Analysis (including Primavera Risk Analysis user skills)

Module 8 - Cost Risk Analysis (including @Risk user skills)

The aim of this document is to provide a hands-on guideline to assist the project controller in obtaining

a basic understanding of the Schedule Risk Analysis process

Upon completion of Module 7, participants will be able to:

 Understand the concept of Schedule Risk Analysis

 Convert an EPC Schedule into a Schedule Risk Analysis Model

 Understand the Risk Ranging Process

 Generate Schedule Risk Analysis Reports / Graphs

 Understand and Interpret Schedule Risk Analysis Reports

 Utilise the basic functions of Primavera Risk Analysis

Module Content:

1.1 Uncertainty and Risk in Cost Estimation 6

2.0 Project Risk Management 8 2.1 Risk Planning 8

2.2 Risk Identification 8

2.3 Qualitative Risk Analysis 10

2.4 Quantitative Risk Analysis 13

2.5 Risk Response Planning 15

2.6 Risk Monitoring & Control 18

3.0 Quantitative Schedule Risk 19 3.1 Schedule Risk Analysis: Key Steps 19

3.2 Basic Principles 21

4.0 Cost Risk Analysis Process 28 4.1 Background 29

4.2 Can we meet the deadline? 30

4.3 Main Execution Contracts 32

4.4 Contract 1- EPCM 32

4.5 Contract 2 - Module Fabrication 32

6.6 Contract 3 - Transport & Installation 33

6.7 Contract 4 - Brownfield Modifications / Tie-ins / HUC 33

6.8 Contract 5 - Accommodation Vessel 33

Training Exercises

Exercise 1 - Develop Cost Risk Analysis Model 34

Exercise 2 - Risk Analysis Process 44

Exercise 3 - Risk Analysis Results & Reports 51

Attachment 1 - Risk Model (042_TRN007_xx.xls)

ules more often than not overrun their targets

Schedule slippages can result in schedule Critical Path changes;

Changes in the Critical Path often impact originally planned project execution strategies

And so on…

So why does all this happen?

Most of the schedule slippages are usually explained by one of the following factors:

imposed unrealistic targets

completeness/correctness of activities logical sequencing

improper use of constraints

inadequate resources loading However, the fundamental rea-son for schedule slippages lies in the very nature of CPM (Critical

CPM technique itself leads to the optimistic result due to “as soon as possible” approach

Project completion date gets predicted, and regarded as a certain, solid commitment against which a number of project performance targets are established

The fact that activity durations (and associated costs) are only estimates, and therefore uncer-tain, means that they may not go

as planned, but may take more

or less time to complete

This is true even for activities repeated many times over a number of projects

Figure 1 on page 7 illustrates that fact by showing the original (likely) activities’ durations as bars, along with triangular shapes at the bars right end points representing possible ranges of optimistic and pessi-mistic activity durations

Supported by experience and measured evidence from pro-jects, this likely variability of ac-tivities durations, enhanced by complexities of their logical rela-tionships, often impacts major milestone(s) or overall schedule

1.1 CPM (CRITICAL PATH

METHOD) & CERTAINTY

versus UNCERTAINTY

PCDP Module 2 - EPC Schedule

Development described the

pro-cesses, techniques and tools

used by WorleyParsons to build

project schedules

Not much different than with any

other organization engaging in

project delivery: project plans

and schedules are one of the

essential tools of project

man-agement

PMI (Project Management

Insti-tute) suggest in their best

prac-tices documents that all projects

must be managed to their

sched-ules

And yet, despite all knowledge,

sophisticated software and

en-gagement of experienced Project

Delivery teams, project

sched-“When you reach the top, keep climbing”

~ / ~ Zen proverb

Sometimes this impact may

take place as a surprise,

be-tween two reporting periods

and have dramatic

conse-quence, and some other times,

a number of small,

incremen-tal changes over longer

peri-ods may add to a significant

impact on project schedule

With the repeated reference to

the PCDP Module 2, where

aspects of schedule

uncertain-ty are brought up in its sections

5.8 Risks and Opportunities

and 5.13 Schedule Reserve,

this PCDP module 7 is

dedicat-ed to the review of probabilistic

approaches and related

pro-cesses, techniques and tools

The processes of project risk

management will be described

in more detail in the

subse-quent sections

Schedule risk analysis comes

with answers not possible or

not available in CPM planning/

scheduling approach:

 Probabilistic View offering

 Uncertain durations (and costs) are defined by three-

point estimates - optimistic (low or minimum), likely and pessimistic (high or maxi-

mum)

Durations (and costs) are expressed as PDFs (Probability Distribution Func-tions)

The range of expected values (dates, costs) are obtained by simulation

Some of the benefits of schedule risk analysis are:

It provides a range of possible completion dates (range of expected costs) along with corresponding probabilities

Provides the extent of possible overruns and required contin-gency reserve

Identifies the areas of greatest risks, inclusive of analysis of near-critical paths

Provides inputs to risks sponse plans

re-And more…

But before continuing with the more focused and detailed train-ing in quantitative schedule risk analysis, the entire next chapter

is dedicated to the broader ject of project risk management, and as such it is shared / repeat-

sub-ed in the PCDP Module 8 - Cost Risk Analysis

Figure 1 – Activity Duration Uncertainty

Risk management is an essential

and integral part of the

Wor-leyParsons project delivery

pro-cess

Understanding project risks is

key for successful

implementa-tion of cost risk analysis

As per PMI PMBOK Guide – 4th

Edition (ANSI/PMI

99-001-2008):”The objectives of Project

Risk Management are to

in-crease the probability and impact

of positive events and decrease

the probability and impact of

negative events in the project.”

Project Risk Management

in-cludes the processes of

conduct-ing risk management plannconduct-ing,

identification, analysis, response

planning, and monitoring and

control on a project

2.1 RISK PLANNING

Planning for risk management is

the process that defines how to

approach, plan, and execute the

risk management activities for a

project It creates a roadmap for

the remaining risk management

processes

This process is general and high

level in nature and therefore

takes place early on the project

The output of the risk

manage-ment planning is the Risk

Man-agement Plan (RMP)

The RMP details the risk

man-agement activities that will be

undertaken by the project ing the purpose, scope, process, responsibilities and extent of technical risk studies

includ-Another important part of risk management plan is a descrip-tion of how risks will be catego-rized

A tool for creating consistent risk categories is the Risk Break-down Structure (RBS)

In the RBS, the categories of risks are decomposed into fur-ther details An example of RBS showing risk categories is shown

in Figure 3

The risk management plan is plotted out by meeting with all appropriate stakeholders

This is followed up with a further analysis to determine the appro-priate level of risk and the ap-proach warranted on the project

The RMP is either referred to or included in the Project Execution Plan (PEP) as required For more information, please refer to the EMS Task Sheet PAP-9002

2.2 RISK TION

IDENTIFICA-Risk identification is a formalized process that identifies which risks could impact the project and to understand the nature of these risks

Risk identification builds the “risk

register”, which is a list of all risks, their causes, and any pos-sible responses to those risks that can be identified at this point

in the project (Please refer to WorleyParsons Risk Manage-ment Software Ver 4.05, its risk register template and guideline) Typically, identifying the risks is the first step in the Cost Risk Analysis process There are various tools available for the risk identification process:

Documentation Reviews

A documentation review is tured review of project documen-tation, including cost estimate and schedule basis, assump-tions, prior project files, and oth-

struc-er information

The documentation is reviewed for completeness, correctness, and consistency

Information Gathering Techniques

There are numerous techniques for gathering information to cre-ate the risk register

“My education was interrupted only by my schooling.”

~/~

Winston Churchill

The techniques most commonly applied in the context of risk are:

Although it may not be tive, this tool provides structure

exhaus-to the Risk Identification process

Assumption Analysis

Assumptions should not only be documented, they should also be analysed and challenged if nec-essary

Diagramming Techniques

Ishikawa diagrams, also called cause-and-effect diagrams and fishbone diagrams, are another way to show how potential caus-

es can lead to risks

Another diagramming method used to identify risks is Influence Diagram This diagram shows how one set of factors may influ-ence another

For instance, late arrival of rial may not be a significant risk

mate-by itself, but it may influence other factors, such as triggering overtime work or causing quality problems later on in the project due to inadequate time to properly perform the work

Figure 3– Example Risk Breakdown Structure (RBS)

“I have a

Finally, flow charts are useful in

identifying risks Flow charts are

graphical representation of

com-plex process flows

They are especially helpful when

used to “sketch” something very

complex into an understandable

diagram

SWOT Analysis

Strengths, Weaknesses,

Oppor-tunities and Threats (SWOT)

analysis is another practical tool

used to identify potential risks

and group them into four chart

quadrants representing strengths

(S), weaknesses (W),

opportuni-ties (O), and threats (T)

Strengths and weaknesses are

usually of internal nature, while

opportunities and threats present

themselves as external risks to

project

SWOT analysis can give another

perspective on risks that will

often help identify the most

sig-nificant project risk factors

WorleyParsons provides a

brain-storming prompt list to aid in the

identification of Threats and

Op-portunities which may arise

Please refer to the EMS

CRP-0013 Risk Brainstorming Prompt

List for additional details

2.3 QUALITATIVE RISK

ANALYSIS

Qualitative risk analysis process

helps to rank and prioritize the

risks so that the right emphasis

is put on the right risks

It helps to ensure that time and resources are spent in the ap-propriate risk areas

Qualitative risk analysis is a risk probability and impact assess-ment process

It takes each risk from the risk register and analyses its proba-

bility of occurring and impact to the project

Risk Event = Risk Probability x

Impact (Value)

Probability = Frequency of

Rel-evant Event / Total Number of Possible Events

Impact = Value of Loss / Gain

Figure 4– SWOT Analysis Chart

“Between two evils, I always choose the one I never tried before”

~/~ Mae West

By using the probability and impact matrix (PIM), a prioritiza-tion and ranking can be created, which is updated on the risk register

Each risk in the risk register is evaluated for its likelihood of occurring (probability) and its potential impact on the project

Each of these two values are given a ranking:

Risk Probability Scale falls tween 0.0 – no probability, and 1.0 – certainty

be-Risk’s Impact Scale can be scriptive, i.e very low, low, me-dium, high, very high; or numeric 0.1/0.3/0.5/0.7/0.9

de-The risk probability and impact are multiplied together to get a risk score

This resulting score is used to set priorities and relative risk ranking WorleyParsons pro-vides guidelines for Likelihood and Consequence scales in risk assessments (For more, please

refer EMS document CRP-0012) Teams are advised to review the categories and determine a

scale that is relevant to their project

“Risk Analysis is no more about risk than astronomy is about telescopes

~/~ Edsger W Dijkstra

Figure 5– Sample Risk Register

The main output from the

Quali-tative Risk Analysis is an

updat-ed Risk Register The next step

is to add more details to the

register including the following:

Relative ranking or priority of

project risks

The urgency of the risks

The categorization of the risks

The Risk treatment plan, and

Any trends that were noticed

in performing the qualitative

risk analysis

An example of the risk register is

shown in Fig 5 on page 14

2.4 QUANTITATIVE RISK

ANALYSIS

Performing Quantitative Risk

Analysis generally follows the

risk identification and qualitative

risk analysis

It is the process of numerically

analysing the effect of identified

risks on project objectives

It is performed on risks that have

been prioritized by the

Qualita-tive Risk Analysis process as

potentially and substantially

impacting the project’s

compet-ing demands

Quantitative Risk Analysis

anal-yses the effect of those risk

events and assigns a numerical rating to those risks individually

or evaluates the aggregate fect of all risks affecting the pro-ject

ef-It also presents a quantitative approach to making decisions in the presence of uncertainty

There are various tools and methods available for quantita-tive risk analysis process:

Data Gathering & tion techniques (Interviewing)

Representa-Interviewing uses a structured interview to ask experts about the likelihood and impact of identified risks

After interviewing several perts, for instance, the project manager might create pessimis-tic, optimistic, and realistic val-ues associated with each risk

ex-Sensitivity Analysis

Sensitivity analysis is used to determine which risks have the most potential impact and the

degree of overall project ity to any of the evaluated risks while all other variables are kept constant

sensitiv-It’s the simplest form of risk ysis, which helps to determine which risks have the most poten-tial impact on the project

anal-It examines the extent to which the uncertainty of each project element affects the objective examined, when all other uncer-tain elements are held at their baseline values

The advantage of this method is that it gives a range of possible outcomes in which critical varia-bles are easily compared in a sensitivity diagram

The weakness of this method is that variables are treated individ-ually, limiting the extent to which combinations of variables can be assessed, and a sensitivity dia-gram gives no indication of antic-ipated probability of occurrence

Probability Analysis

Probability analysis overcomes

the limitations of sensitivity

analy-sis by specifying a probability

distribution for each variable

Probability distributions are

math-ematical representations that

show the probability of an event

occurring

The probability is usually

ex-pressed as a table or graph

Consider flipping a coin as an

example

There would be two possible

outcomes from this event: head

or tail

Now imagine flipping this coin six

times Doing this, the probability

of the coin landing on heads a

given number of times can be

analysed

Probability distributions are very

useful for analysing risks

They consider situations where

any or all of risk variables can be

changed at the same time,

allow-ing the project manager to take a

good look at the probability of an

event occurring and to make a

rational decision about how to

approach the risk

The most typical probability

distri-butions used in quantitative risk

analysis are:

Triangular, if Optimistic (Min), Pessimistic (Max) and Most Likely scenarios are used

Normal or Log Normal, if mean and standard deviation are used

Other common distribution types include: trigen, uniform, beta, pert, etc

The disadvantage of using the probability analysis method is that defining the probability of occurrence for any specific varia-ble may be difficult, as every project is different

Expected Monetary Value Analysis (EMV)

Expected monetary value sis takes uncertain events and assigns a most likely monetary value

analy-It is a statistical concept that calculates the average outcome

of future scenarios

Opportunities are expressed as positive and threats are ex-pressed as negative values

EMV is calculated by multiplying the outcome’s values and proba-bilities, and adding them togeth-

er

EMV is typically calculated by using decision tree analysis

Decision Tree Analysis

Decision trees describe a sion under consideration and the implications of choosing one or another of the available alterna-tives It incorporates probabilities

deci-of risks and the costs or rewards

of each logical path of events and future decisions

“When someone says he’s going to put all his cards on the table, always look up his sleeves”

~ / ~ Lord Leslie Hore-Belisha

Solving the decision tree cates which decision yields the greatest expected value to the decision maker when all the un-certain implications, costs, re-wards, and subsequent decisions are quantified

indi-Modelling and Simulation

Modelling and simulation ods are approaches that trans-late the uncertainties into their potential impact on project objec-tives

meth-There are almost as many types

of simulation as there are jects; however, one technique used for schedule and cost risk analysis is Monte Carlo simula-tion

pro-Cost risk modelling and Monte Carlo simulation using @Risk software is the preferred method used by WorleyParsons Sched-

ule risk modelling and Monte Carlo simulation using Pri-mavera Risk Analysis Expert (formerly Pertmaster) is the pre-ferred method used by Wor-leyParsons

2.5 RISK RESPONSE PLANNING

Earlier, in the process of Plan Risk Management, we created a general approach to risk ( the risk management plan)

Then, in risk identification cess, we created a list of risks and started our risk register

Then we qualitatively and tatively analysed the risks, and now we are ready to create a detailed plan for managing the risks

quanti-This is exactly what Risk sponse Planning does; it creates

Re-a plRe-an for how eRe-ach risk will be handled The resulting plan is actionable, meaning that it as-signs specific tasks and responsi-bilities to specific team members

Remember that risk can be a positive (opportunity) or negative (threat) event

Figure 6 – Sample Decision Tree

EMV of the Chance Node $ 41.5M

EMV of the Decision

$ 49

Build or Upgrade?

Build New Plant

Upgrade Existing Plant

Strong Demand

Weak Demand

Strong Demand Weak Demand

“If an expert says it can’t

be done, get another expert”

~ / ~ David Ben- Gurion

Therefore, careful consideration

must be given to each risk,

whether the impact of that risk is

positive or negative

Avoidance

Avoidance is changing project

plan to eliminate risk, to isolate

the project objectives form the

risk’s impact, or to relax the

ob-jective that is in jeopardy

Examples of this method are

extending the schedule, reducing

scope to avoid high-risk

activi-ties, adopting familiar approach

instead of innovative one, and

avoiding an unfamiliar

subcon-tractor

Transference

Transferring a risk is shifting the

negative impact of a threat, along

with the ownership of the

re-sponse, to a third party

Contractual agreements,

warran-ties, and insurance are common

ways to transfer risks

Mitigation

Mitigating a risk is the reduction

in the probability and/ or impact

of an adverse risk event to an

acceptable threshold

For instance, if you were

con-cerned about the risk of winter

weather damage to a

construc-tion project, a mitigaconstruc-tion plan can

be to construct the building side of the winter season

out-As stressed earlier, risks can be positive or negative, and where positive risks are concerned, the project manager wants to take steps to make them more likely

The following are specific gies taken to capitalize on the positive risks

strate-Exploit

Exploit means eliminating the uncertainty associated with an upside risk by making the oppor-tunity definitely happen

For instance, if a positive risk of finishing the project early is iden-tified, then adding more talented resources to ensure that the project is completed early would

be an example of exploiting the risk

Share

In order to share a positive risk, the project seeks to improve their chance of risk occurring by work-ing with another party

For example, if a contractor tifies a positive risk of getting a large order, they may determine that sharing that positive risk by partnering with another contrac-tor would be an acceptable strat-egy

iden-Enhance

Enhancing is modifying the “size”

of an opportunity by increasing probability and/or positive im-pacts, and by identifying and maximizing key drivers of these positive-impact risks

Enhancing a positive risk first requires understanding the un-derlying causes of the risk

By working to influence the derlying risk triggers, you can increase the likelihood of the risk occurring

un-For example, an airline might add flights to a popular route during holidays to enhance traffic and profitability during heavy travel times

that didn't tell

you all about

When accepting a risk you are simply acknowledging that the best strategy may not be to avoid, transfer, mitigate, share,

or enhance

Instead, the best strategy may be simply to accept it and continue with the project

There are two kinds of ceptance strategies:

ac-Passive Acceptance: requires

no action, leaving project team to deal with the threats or opportu-nities as they occur

Active Acceptance: establishing

a contingency reserve, including amounts of time, money, or re-sources to handle known, or sometimes potential, unknown threats or opportunities

Acceptance may be the best strategy if the cost or impact of the other strategies is too great

Contingent Response Strategy

A contingent response strategy is one where the project team may make one decision related to risk, but make that decision con-tingent upon certain conditions

It is a response designed for use only if certain events occur, or predefined conditions take place

For example, a project team may decide to mitigate a technology risk by hiring an outside firm with expertise in that technology, but that decision might be contingent upon the outside firm meeting intermediate milestones related

devel-In Risk Response Planning cess, appropriate responses are chosen, agreed-upon, and in-cluded in the risk register

pro-Components of the risk register

at this point can include:

Identified risks, their tions, their causes (eg RBS element), and how they may affect project objectives

descrip-Risk owners and assigned responsibilities

Outputs from the qualitative and quantitative risk analysis, including prioritized risks and probabilistic analysis of the project

Agreed-upon response gies

strate-Specific actions to implement the chosen response strategy

Symptoms and warning signs

of risks’ occurrence

Budget (or schedule) activities required to implement the cho-sen responses

Contingency reserves signed to provide for stake-holders’ risk tolerances

de-Contingency plans and triggers that call for their execution

Fallback plans for use as a reaction to a risk that has oc-curred, and the primary re-sponse proves to be inade-quate

Residual risks that are pected to remain after planned responses have been taken, as well as those that have been deliberately accepted

ex-Secondary risks that arise as a direct outcome of implementing

a risk response

Contingency reserves that are calculated based on the quanti-tative cost risk analysis

“If opportunity doesn’t knock, build a door”

~ / ~ Milton Berle

2.6 RISK MONITORING &

CONTROL

Plans have to be re-assessed

and re-evaluated

Where risk is concerned, we’ve

done quite a bit of planning,

iden-tifying, analysing, and predicting,

but the process of Risk

Monitor-ing & Control takes a look back

to evaluate how all of that

plan-ning is liplan-ning up with reality

Monitor and control risks is a

process that is performed almost

continually throughout the

pro-ject

Risk Assessment

As you perform a project, your

information about risks changes

You should assess this

infor-mation as often as necessary in

order to make sure that the risk

needs of the project are current

and accurate

Risk Audits

Risk audits are focused on

over-all risk management In other

words, they are more about the

top-down process that are about

the individual risks

Periodic risk audits evaluate how

the risk management plan and

the risk response plan are

work-ing as the project progresses and

also whether or not the risks that were identified and prioritized are actually occurring

Variance and Trend Analysis

Variance analysis focuses on the difference between what was planned and what was executed

Trend analysis shows how formance is trending

per-The reason trend analysis is important is that a one-time snapshot of cost may not cause concern, but a trend showing worsening cost performance may indicate that things are steadily worsening and may indicate that

a problem is imminent

Technical Performance Measurement

Performance can take on many flavors In the risk context, tech-nical performance measurement focuses on functionality, looking

at how the project has met its goals for delivering the scope over time

re-Status Meetings

This particular technique is not necessarily suggesting that spe-cially called status meetings related to risk are called

Instead, it is suggesting to create

a project culture where bringing

up items related to risk is always acceptable and risk is discussed regularly

3.1 SCHEDULE RISK ANALYSIS: KEY STEPS

In Section 1 we briefly addressed the subject of CPM (Critical Path Method) and its deterministic nature in relation to uncertainty and probabilistic approach

It is important to understand that the schedule risk analysis does not replace the CPM but it takes

it beyond its reach

The CPM remains the key ing block in the whole process,

build-as the quality of the CPM ule is directly related to the quali-

sched-ty of risk analysis outputs

In other words, the results of the schedule risk analysis (and the same applies to cost risk analysis for cost estimates), do not change the project schedule (or cost estimate) but may directly or

indirectly help to improve them based on the informed decisions made in conjunction with these results

In practice, we also see more of

“risk adjusted schedules”, but more as a result of informed decisions made

The major steps in schedule risk analysis are:

CPM Schedule Development

Develop a quality CPM ule that reflects the project scope and execution strategy

sched-The schedule should show the important project structure and clearly define parallel paths and their meeting points

In the case of large number of activities, focus on those on critical and near critical paths

first, then expand if necessary

— use of “risk banding” or QRTs (Quick Risk Templates) may be an option for large numbers of activities

Avoid open ends in the ule logic, they compromise the schedule integrity

sched-Date constraints should be removed, or minimal

Risk Inputs and Activity tion Ranges

Dura-Risk inputs into the model can be performed in a number of ways, depending on which tool is cho-sen for it: Primavera P6 or Pri-mavera Risk Analysis (previously known as Pertmaster), and using one or combination of techniques explained in Section 2:

Assess the activity duration

ranges as three point

esti-mates: optimistic, likely and

pessimistic

Keep in mind that in the CPM

schedule, original activity

dura-tions may not be the most

likely ones

Choose / Determine

appropri-ate PDFs (Probability

Distribu-tion FuncDistribu-tion)

Simulation, Reporting,

Deci-sion Making

Simulation is performed using the

Primavera Risk Analysis

(Pertmaster), a Monte Carlo

Method based schedule and cost

analytics tool

The results of simulation provide

the answers to a number of

criti-cal questions like:

Are the major milestones dates

and project completion date

feasible or achievable

How likely those dates are,

therefore whether the overall

project duration is the most

likely one or not

How much time or how many

days of contingency might be

needed to bring the completion

date(s) to an acceptable risk

tolerance levels

Schedule Risk Analysis and

Flow Diagram

The following flowchart illustrates

the major steps and processes

involved:

Review Project ule and Aspects of each Project Stage

Sched-Develop Schedule Risk Model in Primavera P6

or in Primavera Risk Analysis

Conduct Risk Ranging Session

Run Simulation

Generate Analysis Graphs

Prepare Schedule Risk Analysis Report

Schedule Basis Document Critical & Near-Critical Paths, Logic, Constraints, Calendars,

Key Schedule Drivers

Risk Inputs Layout

Appropriate Participants Risks Register

Choice of PDFs

Monte Carlo Simulation using Primavera Risk Analysis

Probability Distribution Sensitivities, Index ‘Tornado’

Graphs, P10 / P90 Windows and Mean Dates

Detailed narrative of Schedule Development

Rationale behind Key Schedule Drivers Conclusion / Recommendation Risk Analysis Outputs and Graphs

“The truth is like sunlight, people used

to think it was good for you”

~ /~ Nancy Gribble

3.2 BASIC PRINCIPLES

Single Logical Path Schedule

To demonstrate the basic

princi-ples of what was described on

previous pages, we will start with

the simple schedule shown in

Figure 3.2.1 - Single Logical Path

Schedule

The activities in the CPM

sched-ule are sequential, with all FS

(finish-to-start) relationships,

(with the exception of project

finish milestone having FF,

finish-to-finish relationship)

The activity durations are fixed

and because of no overlap

be-tween them, the total project

duration is the sum of all activity

durations

The CPM also predicts the finish date But how likely is the pre-dicted project finish date?

As the Figure 3.2.2 shows, the introduction of schedule probabil-istic view, where activities dura-tions are not certain but can take any value between defined mini-

mum and maximum, a number of project finish dates are obviously possible

Figure 3.2.3 shows an example of

a triangular PDF for one activity minimum, likely and maximum (Activity A1010 - Design Line #1)

Figure 3.2.1 - Deterministic Single Logical Path Schedule

Figure 3.2.2 - Probabilistic Single Logical Path Schedule

Figure 3.2.3 - Triangle PDF (for A1010)

Figure 3.2.4 shows the result of

simulation with the range of

pos-sible project finish dates

Given the defined risk inputs

(Figure 3.2.2), the risk outputs

show that the original

(deterministic) project finish is

unlikely - only 15% (P15)

A number of days of contingency

would be needed to bring the

project completion date to P50 or

higher

Parallel Logical Paths

Schedule

Most of our project schedules are

not as simple as the one used on

the previous page

Typical scenario is that there are many more activities, with more complex logical relationships, with multiple logical paths, etc

To prove that the more complex schedules are more risky, we expand the single logical path schedule to two parallel logical paths schedule

Also, the second logical path added is identical to the first one

(See Figure 3.2.5 on page 23)

The activity durations, length of both paths and project finish date are the same

It is also assumed that the tainties around both paths’ activi-ties are the same

uncer-Performing the simulation based

on these risk inputs generates the output report with the range

of different expected project ish dates

fin-More importantly, the output shows that the deterministic pro-ject finish is now only 2% likely, a product of two independent logi-cal paths probability of 15% each(15% X 15% = 2.25% or P2!)

Figure 3.2.4 - Single Path Range of Project Finish Dates

“Being on a tightrope is living, everything else

is waiting”

~/~

Karl Wallenda

The comparison of results shows

that:

Two-paths schedule is more

risky because each logical

path may delay the project

The risk is increased at the

logical paths converging points

(that can be said for any

inter-im project milestone where two or more logical paths merge)

Two-paths schedule is more risky at the optimistic and mean ranges of dates than at the pessimistic ranges of dates

As the range of expected finish dates is different, in this scenario

we would need to add more time (days) in contingency to bring the project finish to P50 or more This is illustrated by Fig 3.2.6

Figure 3.2.5 - Probabilistic Parallel Logical Paths Schedule

Figure 3.2.6 - Parallel Paths—Range of Project Finish Dates

Criticality Index

The CPM schedule critical path

analysis is useful for a number of

reasons:

It represents the longest

logi-cal path determining the

earli-est project finish date

Because of that, the delay on

the critical path delays the

project

It is usually the path that

re-quires most attention

Figure 3.2.7 shows a slightly

modified schedule from the

previ-ous page example, only one

activity has changed as follows:

A0120 - Supply, Prefab &

In-stall Line #1 is now 45 instead

of 50 days

The logic for Piping Line #1

has different inputs for

mini-mum and maximini-mum ranges

The shorter likely duration

“promoted” the second logical

path to be critical (Logic Path

Piping Line #2)

If we were to follow the CPM

schedule, the logical course of

action would be to focus agement attention to the activi-ties on the critical path Pipeline

Therefore, the CPM approach can potentially miss the risks by focusing on the critical path ac-tivities instead on those that are more uncertain and risky Criticality Index represents the percentage of iterations for each activity that was on a critical path during the simulation

Criticality index of 100% means that regardless of how the task durations varied, the critical path always included that activity

Activities with the high criticality index are more likely to cause delay on projects

Criticality Index is typically ted in the form of “Tornado Dia-gram”

plot-Probabilistic Branching

When the schedule was created the most likely path and activities have been created

Unlike with CPM approach, the schedule risk modeling allows for probabilistic branching in situa-tions when it is not clear what the outcome of an activity may be

The successor activities may be

on different logical paths (branches), depending on the probability of the predecessor outcomes

Figure 3.2.9 shows that two logical paths are possible after the lines are declared ready for testing:

Pass the test (70% chance) and go to commissioning, or

Fail the test, for which 30%

chance of occurrence is given, which would then require lines

to be fixed and re-tested, thus delaying the system commis-sioning

“Look twice before you leap”

~/~ Charlotte Bronte

Figure 3.2.7 - CPM Technique Can Hide Risks

For every iteration of the tion one action is chosen ran-domly using the assigned proba-bility and the other outcomes are ignored

simula-The results of risk analysis, as in Figure 3.2.10, show that:

Frequency distribution results (histogram) get a shape of distinct humps that represent

branched probabilities

Cumulative probability tion gets a “shoulder” shape at the branching probabilities

distribu-Conditional Branching

Similar to probabilistic branching, the conditional branching models special project conditions and their consequences

It is helpful in what-if scenarios where triggers like missed dates

or excessive costs reach beyond preset thresholds

The outcomes and logical paths beyond these triggers are then different than originally planned

Figure 3.2.8 - Criticality Index

Figure 3.2.9 - Probabilistic Branching

Figure 3.2.10 - Probabilistic Branching Result

Correlation

In broad statistical terms,

correla-tion represents a relacorrela-tionship

between two or more random

variables…

Correlation can be positive or

negative, and is expressed as a

range between 1 (perfect

in-creasing correlation) and –1

(perfect decreasing correlation)

Correlation value of 0 (zero)

indi-cates that variables are

uncorre-lated

Using correlation in schedule risk

modeling is useful and

recom-mended, as it prevents

unrealis-tic situations during the

simula-tion

The random sampling gets

“driven” in a sensible way to reflect the degree of correlation between variables

For example, if the foundation size increases, the excavation requirement will likely increase which in turn will proportionally increase the time required for back filling and compacting

In this case of positive tion, a single iteration random sampling will choose the values from the variables PDFs reflect-ing the degree of correlation

correla-PDF Selection

Probability Distribution Functions (PDFs) were already mentioned earlier

In this sub-section just a few additional details are added

It is clear that PDFs are used to model the range of activity possi-ble durations (or costs)

There is a number of available PDFs to be chosen, each meant

to provide the best tion of possible values and corre-sponding probabilities for a varia-ble, in this case activity duration

representa-Using the example of two PDFs, used quite often, we want to demonstrate the importance of choosing the appropriate PDF for activities risk inputs modeling: