Cascading risks

Download a pdf copy of this help file here

Ideally, we would like to be able to capture the complete potential impact of a risk to be able to understand the level of attention it should receive. Maybe some risk occurring in a project has a relatively minor impact directly, but it increases the chances of another, much larger risk. We can think of lots of reasons why: management are focused on handling the effects of small risk A so nobody is paying attention to a looming big risk B; little risk A occurs, blame is passed liberally around, people stop communicating and helping each other, and those who can see big risk B coming along think 'It's not my problem'. The biggest risks in projects are, after all, driven by people issues. I think that the occurrence of a lot of big risks come at the end of a chain of small risks occurring that were perhaps much easier and less costly to deal with.

Let's look therefore at a couple of simple ways of modelling a cascading set of risks in a project. We want to model the probability of a risk occurring, and perhaps the size of its impact too, as being to some degree influenced by whether other risks have occurred. The figure below gives an example:

Example model Cascading_risks - correlated (cascading) risks

The model uses the ModelRisk function VoseRiskEvent.

The interesting parts of this model reside in shaded Cells F6, F7, F10 and G10 where we have used IF statements to change either the probability or impact of a risk depending on whether other risks or combination of risks have occurred: if Risk A occurs it increases the probability of Risks D and E occurring; and if Risks B, C and F all occur then both the probability of occurrence and the size of impact of Risk H increase.

Read on: Including identified risks in a project schedule model