Bayes Theory as a Methodological Approach to Assess the Impact of Location Variables of Hyperscale Data Centres: Testing a Concept

Abstract

The theme of ’The Impact of Engineering Practices on a Sustainable Built Environment’ emphasises the importance of considering various dimensions of resilient infrastructure. Selecting the location for a Hyperscale Data Centre is a crucial process that involves assessing the impact of various location variables. To determine the viability of a location, it is essential to identify the potential risks associated with each variable. This paper presents a proprietary methodological approach that includes a Delphi study to identify risks, a Likert scoring system to assess prior probabilities, and a Bayesian theory-based decision tree to assess the impact through risk prediction. The paper's contributions are significant, and the proposed methodology makes it possible to predict the risk level of each location variable by identifying the appropriate contingency percentage. The study's findings indicate that the paper's proposed approach is an effective way to mitigate the risks associated with selecting a location for a Hyperscale Data Centre. Embracing this knowledge allows us to align research and practise with the conference’s call to studying the resilience of buildings and infrastructure to natural disasters and climate change, and developing strategies for adaptation and mitigation, ensuring that these practises become integral to shaping the future of Data Centres