Methodology for the Quantitative Risk Analysis of Wildfires in the Wildland-Urban Interface: Application to Electrical Infrastructure

Abstract

Wildfires have caused significant damage in Chile in the last decade, with electrical infrastructure being particularly vulnerable to extreme wildfire events. Therefore, the objective of this work is to estimate the risk of a wildfire damaging an electrical substation operated by the Chilean company Chilquinta Distribución SA, located in the Wildland-Urban Interface (WUI) of the city of Valparaiso, Chile. A Wildfire Quantitative Risk Analysis (WFQRA) framework is proposed for this purpose, where the quantification of risk results from the product between the probability of a wildfire reaching the WUI and its consequences. The former was determined with an event tree analysis, a tool in which the advance in time of the initiating event (ignition of a wildland fuel) was analyzed through a series of questions related to the area burned, wind direction and ranges of ambient temperature and wind speed, with answers of the type yes or no. This anaylsis gave 0.004 events/year as the probability of a wildfire reaching the WUI. On the other hand, the analysis of consequences requires determining the vulnerability of the object of study to a thermal insult, which is the quantitative relationship between the thermal exposure of the object to the fire and the damage experienced by it. This vulnerability was assessed by expressing it in terms of a probability of failure (ignition) for different doses of thermal radiation, i.e., a dose-response curve. Since these curves are typically of sigmoid shape, a novel probit equation was determined by analyzing experimental data of PMMA, a material that served as a proxy for the actual materials within the electrical substation. Thermal exposure of a target within the substation was calculated with an average fireline intensity (1832 kW/m) and flame length (3.98 m) obtained with FlamMap simulations of fire behavior in a simplified landscape representing the study area. With these results, the vulnerability model for PMMA developed in this work gave PMMA ignition probabilities of 0.029 and 0.269 for two wildfire scenarios considered in the substation. Finally, wildfire risk was estimated as 10-4 to 10-3 events/year, or equivalently, one event every 1,000 to 10,000 years. These results can be used as input to the decision-making process in Chilquinta and public institutions for taking mitigation measures to reduce wildfire risk in the WUI. Future work will be focused on refining this result.