A data model for heat-related rail buckling: implications for operations, maintenance and long-term adaptation

Heat-related rail buckling is a significant operational and safety issue for the railways of Great Britain (GB). Although continuously-welded rail is pre-stressed to a stress-free temperature of 27°C, degradation and local topographic and microclimatic factors can lead to failures occurring at lower temperatures. These buckle events can cause widespread knock-on delays. As a common risk mitigation approach, speed restrictions are imposed for forecasted heat events, with these measures also being associated with non-trivial delays and disruption. The recently published UKCP18 climate projections indicate that the frequency, duration and magnitude of heat wave events will increase in the coming decades. This paper presents a data model to explain the occurrence of heat-related disruption incidents on GB's rail network. This model is built on historical delay data from Network Rail - the owner of GB's railway infrastructure, given explanatory variables for important meteorological and infrastructure features from published research in the literature. The model is implemented at two scales. Firstly at the national scale, including all Network Rail operational routes, and secondly for the South East of England (climatologically the warmest part of the country and hence at perceived greater risk) including the operational routes of Anglia, Wessex and South East.