BY DAVID QUINCEY
The understanding of risk in projects has never been as important as it is today.
PROBABILISTIC MODELLING – RISK BASED CONTINGENCY
MONTE CARLO SIMULATION OF CONTINGENCY – P80 CERTAINTY
These are all terms used to describe the calculation of a value using the statistical technique of probabilistic modelling, most often in the context of contingency.
In most estimates of the cost of construction the one of the biggest single line items is contingency. In some infrastructure heavy rail projects in the early stages of design this can be 60% of the total of assessed costs. Putting a value for contingency as a simple percentage of the total of the assessed costs – a deterministic contingency – is the traditional way of operation, and many organisations produce guidelines and ‘norms’ around this. However many large government and non-government bodies have recognised that in this day and age this simplistic approach is outdated and are turning to probabilistic evaluation of contingency in order to evaluate the project holistically and refine their budgets from early on.
The probabilistic evaluation of contingency, at the highest level, involves reviewing the uncertainties within the estimate that has been prepared, considering risks that may or may not occur and which could impact the project, and applying values and probabilities to those uncertainties and risks. These then form the inputs to the statistical risk modelling, the outcome of which is a range of values from P0 to P100, the most commonly used of which are P50 and P80.
These values give levels of certainty to the projected final outturn cost of the project. At the dollar value given by the modelling for P80, for example, there is a 80% probability that the final outturn cost of the project will be at or below that $ figure; or conversely there is a 20% probability of the final outturn cost of the project exceeding that figure.
For project sponsors, funding bodies, commercial decision makers, business case writers etc this data provides context and meaning to a contingency. If a deterministic contingency is applied there is no knowledge of whether the bottom-line number produced is a most likely final cost, or worst case, or optimistic. With the outcomes of the probabilistic assessment the range of risk is defined and known, and then the decision maker has a solid base from which to proceed.
This process is now required by most federal and state government departments, and the bodies overseen by those departments, and it is also being applied by many major developers, utility companies, and others who have significant capital expenditure projects or programs of works.
One common usage of the data is in the delegation of authority, such as the project manager being given the authority to spend up to the P50 level, the capital works manager being authorised to spend up to the P80 level but Board authority being required if the spend exceeds that level. Organisations who have significant programs of works can also apply the same modelling technique to manage program contingency and not ‘tie-up’ significant capital sums in each of their individual projects.
At WT we have been at the forefront of probabilistic modelling for many of our clients and continue to develop our skills and models. Our processes meet the standards required by all Federal and State governments and bodies, as well as the ISO and the Risk Engineering Society.
The risk workshop is the key part of the process. Run by a WT facilitator, they should bring all key stakeholders together, and the outcomes recorded and forming the inputs to the mathematical modelling. The WT facilitator is able to use our high level of experience on complex projects to do a deep-dive into the project with the workshop participants and stimulate discussion on all manner of risk topics.
For greater insight into how risk modelling can benefit your project please contact your local office.