A Comprehensive Forecasting, Risk Modelling and Optimization Framework for Electric Grid Hardening and Wildfire Prevention in the US

Historical wildfire patterns have experienced a recent shift in terms of its scale and intensity. Through the continuing advancements in electrical protection technology, statistical forecasting methodologies, availability of meteorological field data, and regional risk-modelling, wildfire management practices can be made more proactive in the United States and around the globe. To create a comprehensive and practical operating framework, an advanced seasonal autoregressive integrated moving average time series modelling technique for wildfire forecasting is explored. These regressive models, due to their mathematical accuracy has been used in many engineering and scientific applications. The study presented here was done using a qualitative investigation approach to wildfire data. Computer automated grid search techniques were developed to determine suitable seasonal regressive model hyper-parameters. With the usage of power transforms to fit skewed statistical models under study, it is found that a much more accurate and computationally efficient model can be generated. Statistical forecasts and regional risk mapping techniques can influence strategic operational practices for regional and local fire authorities. Concepts that can enhance power system protection and electrical grid hardening are explored and practical guidelines to help electrical utilities improve electrical grid operations are provided. Many benefits of using distributed energy resources are discussed and an optimal power flow involving these resources is formulated to help grid operators preserve system stability under these wildfire scenarios