Utilizing Pressure Gradients to Verify Machine Learning Wind Strength Forecasting for Grid Resiliency in Utility Operations

The resilience of the power grid is increasingly challenged by extreme weather events, particularly high wind speeds that can cause extensive damage to infrastructure and disrupt energy supply. Accurate forecasting of wind strength is essential for utility companies to mitigate these risks and enhance grid resiliency. Measuring mean sea level pressure differences over various spatial scales can help inform on expected wind speed and direction, offering a tool for utility companies to prepare for potential disruptions. Leveraging a database of historical pressure gradients during high impact weather events resulting in significant outages, a framework was developed to integrate pressure gradient forecasts into the expected outage forecast. During one such event, this tool checked the validity of a machine-learning outage prediction model being utilized in operational and business continuity decisions. The pressure gradient analogs were able to better inform on the expected strength of winds with historical perspective. This enabled the company to understand the risk and make proactive decisions regarding potential outages and emergency response. This approach not only enhances operational efficiency but also contributes to the overall resilience and response of the power grid against extreme weather events. Examples of this includes building into business cases a storm benefit based upon the historical pressure gradient analysis to help inform system hardening like undergrounding. The findings underscore the importance of integrating meteorologists and meteorological data science into utility operations in managing wind-related risks to critical infrastructure.