Climate change is increasing the frequency with which the UK infrastructure is threatened by extreme weather events. To explore the potential impact of future climate conditions, the CReDo project is working to develop a digital twin of key infrastructure networks. This digital twin can be used to help make decisions to better protect the networks in advance of extreme weather events, and ultimately to help inform a real-time response to extreme weather events. The novel feature of this tool is that it will provide the collaborating asset owners- and also crisis management teams- with not only assessments concerning the impact of a weather-induced flooding incident in a future climate on the infrastructure and networks monitored by the individual asset owners, but also the operability of assets owned by other companies- where the failure of these assets impinges on the functionality of their own. The highly interdependent nature of these infrastructure networks, such as telephone lines relying on power supplies being operational, mean that reliably modelling the impact of an extreme weather event requires accounting for such connections. It is planned that the shared appreciation of the mutual threats described by the digital twin across the different actors will encourage further coordination between the companies in their strategic plans to mitigate these increasing threats.
This report outlines just one component of this development. We demonstrate how it is possible to elicit from asset owners the probabilities that each of their assets might fail, in a particular future flood scenario that makes consideration of the impact of climate changes on extreme weather patterns. Taking these unfolding events, and through working with teams of domain experts drawn from asset owners associated with the local power, water and telecommunication companies, our team demonstrate how it is possible to elicit probability distributions of the failure of each asset and their connections within the network. This information would then be fed to operational researchers who can calculate the knock-on effect on the whole network of each simulated future incident. From a decision-analytic perspective, the digital twin would thus consist of connected digital twins representing hydrology, the failure modes of assets, and the system in which the assets sit, with a decision support layer sitting above this.