Newsletter December 2016

Newsletter April 2016

Newsletter December 2015

Newsletter June 2015

Newsletter December 2014

Newsletter August 2014

Newsletter April 2014

Newsletter November 2013

Newsletter July 2013

Launch Announcement

Newsletters Signup
RiskIntegral Banner
Colin Cropley

Colin Cropley

Managing Director
+61 412 031 161
email Colin
visit Colin on LinkedIn

Matt Dodds

Matthew Dodds

Principal Consultant
+61 433 215 324
email Matt
visit Matt on LinkedIn

Robert Flury

Robert Flury

Principal Programmer
+61 403 134 479
email Robert
visit Robert on LinkedIn

Peter Downie

Peter Downie

External Director
+61 412 994 568
email Peter

July 2013

Welcome to the inaugural issue of “RiskIntegral” - an occasional newsletter about the analysis and management of risk, mainly (but not exclusively) in projects, by Risk Integration Management Pty Ltd (RIMPL).

In this issue:

Why management of risk matters more now!
Balancing Accuracy and Credibility in Schedule Risk Analysis
RIMPL and what we do

Why management of risk matters more now!

Risk is an integral part of almost everything humans and organisations set out to do, because it is about the uncertainty of achieving objectives in the future. In fact, the definition of risk is “effect of uncertainty on objectives” (AS/NZS ISO 31000:2009).

Many of us have experienced first-hand the effect of uncertainty on projects, usually in an adverse way

Australia, where we are located, has experienced an unprecedented sustained boom in investment in Mining and Oil & Gas over the last 10 years or so, thanks largely to the extraordinary growth and development of China.

Unfortunately, there has also been a recurring pattern of failure to achieve project completion dates and budgets. Independent Project Analysis, Inc. (IPA), the pre-eminent international projects bench-marking organisation, has identified failure of Australian project owners to plan projects adequately, including proper assessment and management of risk, as a primary cause of this problem.

Until recently, a persistent business risk saving such projects has been the opportunity presented by continually rising commodity prices. These have more than compensated for cost and schedule overruns. But as the growth of China cools and changes with the maturing of its economy and with the advent of shale gas in the US and elsewhere, commodity prices have started to fall as demand has fallen and supply has increased. In this more challenging business climate, getting risk assessment right has become much more important in assessing and approving projects. This is even more important in already committed projects where major risks must be regularly re-assessed and effective risk treatments devised to ensure project value is protected and project objectives are achieved as far as possible.


It is proper risk assessment and management that is our passion. We have been involved in analysis of risk on major and complex projects in the Oil & Gas, Mining, Petrochemical and Infrastructure sectors around Australia, as well as in training personnel in cutting edge Schedule Risk Analysis (SRA). In recent years, we have developed and delivered SRA training courses for major Australian contractors in response to their mandating of risk analyses and consequent need to up-skill personnel in what is rapidly becoming an industry standard process.

Our Integrated Cost & Schedule Risk Analysis (IRA) methodology and software, developed and refined over a number of years, have been applied to projects ranging from a few million dollars to more than $15 billion.

We welcome enquiries or questions about these or other services and particularly welcome questions about risk analysis and risk management, through our Knowledge Base or through contacting us.

Balancing Accuracy and Credibility in Schedule Risk Analysis

When Schedule Risk Analysis (SRA) based on the Monte Carlo (MC) Method of randomness simulation was first developed, the tools available imposed practical limits on the size of schedule. Schedules were limited to around 100 activities or less. Since then, software and hardware advances have enabled SRA models to become much larger. How big should such models be? And does it matter?

To answer these questions, we need to deal with two effects with opposite implications.

First, we consider a familiar concept in project development, known as “Fast tracking”. This refers to the desire of project owners to get to market as quickly as possible by overlapping the phases of development of projects, such as Design, Procurement and Construction. At a high level, this seems quite feasible, but when detailed planning is performed, the difficulties of carrying this through become more evident. We can see this by taking a simple schedule model:

If we have two identical strings of activities and resources to do them, each with a 50% probability of being finished by the target finish date, what is the probability of both being finished by that date?


This is the same as working out the probability of throwing dice and getting only sixes. The probability of throwing a six for any one dice is 1/6. The probability of throwing multiple sixes is (1/6)n, where n is the number of dice.

When this is translated to project schedules, it is known as the Merge Bias Effect (MBE). It means that where there are logic nodes in a project schedule, where several strands of schedule logic come together in a milestone, the probability of achieving that milestone by the planned date is the product of the probability of each logic strand being completed by the planned date.

As more strands of schedule logic are joined into milestones representing completion of portions and phases of work, these logic nodes become points resisting the schedule finishing earlier because each logic strand would have to finish earlier simultaneously, which rapidly becomes less and less likely as more detail is added to the project schedule and more intermediate milestones are created.

If a detailed project schedule represents the way that the project will be executed, summarising it will remove details including intermediate milestones that have to be achieved. This will remove logical resistance points to earlier completion such that a MC simulation model built on a summary schedule is likely to produce a falsely optimistic forecast of the likelihood of finishing the project by the planned date.

Second, producing realistic results from MC simulations with progressively larger schedules is increasingly difficult.

To explain why we need to understand how MC SRA simulations work. The foundation of SRA is the replacement of single durations for activities with duration ranges represented by probability distributions, such as shown on the left, representing a task with a duration somewhere between 58 and 77 days long. This recognises that for most tasks we can only say with confidence that they will not be shorter than some Optimistic duration, nor longer than some Pessimistic duration, with a Most Likely duration somewhere in between, where the probability is at a maximum (64 days in this case).

triangle distribution

Simulations consist of hundreds or even thousands of iterations of critical path calculations, randomly substituting sampled durations from each task’s duration probability distribution. The software selects the duration for each iteration so that it matches the probability distribution defined for that task. So more samples are taken around the Most Likely values than toward the extremities of the ranges.

An inherent assumption of MC simulation software is that each element in the model is completely independent of every other. This assumption is relatively safe when using a highly summarised model but progressively breaks down as the number of activities increases. So one task representing all earthworks will be relatively independent of another task representing all civil works and a further task for structural steel installation, etc.

But if there are 50 tasks describing in detail earthworks to create, for example, access roads, the ROM Pad and the building of tailings dam walls, some, perhaps all will be related to each other to varying degrees. We tell the software the extent to which tasks are related to each other through duration correlation: 100% represents a perfect positive relationship, 0% represents complete independence and -100% represents a perfect inverse relationship. Normally the relationship is positive and somewhere between weak (say 10-30%), moderate (say 30-60%) and strong (say 60-90%).

If we did not tell the software the extent to which tasks are related, it would choose short durations in combination with long ones and vice-versa, such that they would tend to cancel each other out over the iterations and lead to the calculation of very narrow distributions (e.g., a 2 week range between the P10 (10% probable) project finish date and the P90 date in a 2 year project (less than a 2% range)).

This effect is due to the rising impact of the Central Limit Theorem (CLT). As the number of activities increases in the SRA model and the average task duration decreases, the distribution produced tends to a normal distribution and narrows in its spread. The narrowness of the distribution is called “kurtosis”*.

Countering this to produce realistically spread distributions becomes increasingly difficult as the size of the schedule increases. Strategies for countering the CLT tendency include the application of risk factors and increasingly sophisticated correlation models.

To produce the most realistic SRA modelling requires the use of the largest practical SRA schedule model, preferably based on a Level 3 Master Control Schedule for the project and the application of comprehensive correlated risk factors and carefully developed duration correlation models. Achieving this is much harder than producing summarised schedule models, but the results are much more likely to reveal useful information on what is driving the schedule and how best to manage the drivers to maximise project objectives.

Using Oracle’s Primavera Risk Analysis (PRA) and a fast PC, it is practical to analyse a schedule with up to about 5,000 activities in it. Around 1,000 activities is more practical to handle. Provided true critical paths can be identified and the schedule realistically represents the planned project strategy and scope, this size schedule should produce good and useful results.


*Kurtosis describes the “peakedness” of a distribution. The diagram, below right, shows the different types of kurtosis.

Skewness is also important. Most SRA output distributions are skewed in a positive direction (because there are more ways for a project to be late than early!). Positive and negative skewness are illustrated here.

triangle distribution
triangle distribution

Skewing is important because it is in the low probability skewed “tail” that significant low probability, high impact risk events tend to be concentrated. These are often the cause of schedule blowouts of major projects.

The images left and right come from a website offering statistical analysis software:

There is an Australian way of thinking about kurtosis in this context. In SRA output distributions, we wish to follow the platypus rather than the boxing kangaroos.

triangle distribution
(the image comes from

Risk Integration Management Pty Ltd (RIMPL) & What we do


RIMPL is an innovative risk management and project controls services business. We provide risk management and analysis consulting services as well as skilled project controls and strategic project advisory services and personnel. RIMPL continues services formerly provided by Crescent PSS Pty Ltd (1996 - 2008) and Hyder Consulting Pty Ltd's Project Management Group (2008 - 2012). Through seven years of software methodology development, the risk management team at RIMPL offers the following services:

• The most sophisticated and realistic Schedule Risk Analysis (SRA) modelling available in Australia

• The only true Integrated Cost & Schedule Risk Analysis (IRA) modelling developed in Australia, to assess project contingencies, profitability and viability

• Licensing of the methodology and software suite to approved clients

• Training in SRA using Oracle's Primavera Risk Analysis (PRA) and optionally also using RIMPL's risk analysis software to enhance the modelling realism of PRA

• Sale of and training in selected RIMPL qualitative and quantitative risk analysis applications

Independent of our risk management services, we also offer development of database applications to client requirements.