The spread of the COVID-19 pandemic has prompted state and local governments in the US to put quarantine measures in place. As a result, the pattern in electricity consumption has significantly changed. There were dramatic shifts in energy use between the first and fourth weeks of March 2020. In the US, several non-essential retailers such as salons, clothing shops and dry cleaners were using only 35% as much electricity toward the end of the month as compared to that during the beginning of the month, while several restaurants were using only 28% in the last week compared to the first week. Conversely, the consumption in storage facilities rose sharply, using 142% as much electricity in the fourth week compared to the first.

In this backdrop, artificial intelligence (AI) based load forecasting is assisting utilities in making timely adjustments. AI-based load forecasting gives energy providers the ability to develop informed and data-driven strategies. Without AI tools, utilities would see their forecasts being highly inaccurate. This would place an enormous strain on their operations and may even lead to blackouts and brownouts. All this would eventually drive up costs for businesses and consumers.

Somik Das, Senior Power Analyst at GlobalData, comments: “AI will not resolve all of the utilities’ issues. However, it enables them to better adjust to the shifted daily and weekly power load profiles. A few notable utilities in the US such as Portland General Electric, Gexa Energy, Avangrid, Arizona Public Service Electric, WGL and Mega Energy, use AI platforms to forecast short and long-term loads and variances.

“In the current situation, forecasting the load shift and simultaneously understanding the new stress points in the grid could prevent failures of key electrical components such as reclosers, interrupters and circuit breakers and in turn reduce the chances of power cuts and blackouts.”