The quantification of risk should only be attempted after the reference project and raw public sector comparator costing exercise have been completed. In the first instance, risks can be usefully quantified using workshop techniques. This should occur after the risk experts have been given sufficient time to individually consider the likely probabilities and impacts of the risks.
Considerable analysis has been undertaken on the psychological factors that give rise to error in the quantification of uncertainty, both in and out of workshop situations. When quantifying risks in a workshop situation, some of the dynamics to be considered include:
| Term | Description |
| Conformity | When a group of people estimates a risk, group members tend to gain unwarranted confidence from each other's estimates. This results in groups giving a narrower range of estimates than when done independently. |
| Bias | More senior individuals at a session are likely to influence others merely by their presence. Most people will have a bias in a particular direction, but such biases tend to converge when a dominant person is present. |
| Personality | Where the loudest voice dominates, and the quietest is not heard. |
To avoid such issues in a workshop environment, a preferred technique is to present preliminary analysis data to experts before the session, and use the session, not for the actual quantification of risk, but rather for discussion on the level and logical structure of uncertainty and correlations between risks. The aim of the brainstorming session is to ensure that all participants leave with a common perception of the uncertainties of the problem, and that risk experts understand what is required of them in the next stages.
Depending on the stage of development of the project and time constraints, it may be possible to complete both the risk identification and risk quantification workshops in the same day.
To accelerate the workshop process it may be useful to set some parameters for subjective descriptions of probability and impact. For example:
| Boundary | High | Medium | Low |
| Probability of occurrence | greater than 60% | From 30% to 60% | Less than 30% |
| Impact on Capital Cost | $0.5m to | $0.25m to | $50K to $0.24m |
| Impact on Recurrent cost | $50K to | $25K to | $10k to $24K |
Note that the ranges above should not be used without consideration of their applicability to the proposed project. They are examples only and would, for example, be inappropriate for a $1 billion project.
Workshop participants are only required to agree whether the risk should be classified as 'high', 'medium' or 'low' in terms of probability and cost impact, and whether the cost impacts on capital expenditure, recurrent expenditure or both. These parameters provide a useful guide for the risk experts to carry out the actual risk quantification in the following phase.
The major outputs of the workshop are an understanding of the project risks and a general indication of where the risks would lie, i.e. in the high, medium, or low category. A roundtable discussion brings together the experience of the risk participants such as engineers, architects and core service operators, who will assess the risk from the perspective of their own expertise. The workshop will aid in a better understanding of the risks for those involved in the project and act as a guide for the risk expert.
Risks that are difficult to quantify due to the high level of uncertainty in variables or the inherent nature of the risk, (for example, risk due to changes in the private party's key personnel) should be classified as qualitative in nature, and recorded as such.
Each risk is also analysed with regard to any correlations with other risks , i.e. to determine the relationships (if any) between identified risks.