Load Forecasting, Solar Eclipses, & More

On Monday, April 8, parts of the U.S. and surrounding areas experienced a total solar eclipse that lasted for 4:28. Utilities in the path of totality experienced a 40 GWh dip in solar output. Fortunately, total eclipses are an infrequent occurrence, happening every 1-3 years often at the Earth’s poles or in the ocean, with the next set to occur in the U.S. in August 2044, two full decades from now. Still, from total eclipses to increases in extreme weather events, load forecasting is a powerful tool to not only plan for but learn from this natural phenomenon to better prepare for the next time. So how can load forecasting help?

It Can’t Make That Much of a Difference… Can It?

While many were experiencing the natural splendor of the total eclipse in the path of totality, the National Renewable Energy Laboratory (NREL) monitored the impact of solar production in real time, a powerful indicator of how the natural world influences the energy sector. In the video shared by NREL, you can see how dramatic solar production drops, with peaks corresponding with the literal coverage of the sun. This was such a dramatic shift in solar production, that ERCOT reported a massive spike in real-time costs, jumping up to $870 per kWh during the eclipse. These figures are something that load planning can help account for.

As of this writing, it takes 2.469 million photovoltaic solar panels to produce 1 Gigawatt or 98.76 million to produce the approximately 40 GWh lost during the most recent total eclipse. To meet U.S. decarbonization goals by 2035, 30 GWh of solar photovoltaics must be installed between now and 2025 and 60 GWh installed between 2025 and 2030, which is approximately 43b more PV installations. Unfortunately, the more reliant we are on solar, a relatively dependable resource, the more impact subsequent eclipse events will have on energy production. Fortunately, load forecasting offers a long-term solution to plan and prepare for natural phenomena like eclipses and increased volatile weather.

Climate Change & Extreme Weather Events

The increasingly erratic weather patterns caused by climate change have cost the world around $391m per day over the last two decades, a figure that will increase with some experts anticipating an annual cost of $1.7b by 2050 for utility companies. These temperature and weather extremes are anticipated to increase even more, further complicating grid resiliency and energy security initiatives, especially in the face of electrification efforts and the increase in distributed energy resource (DER) market penetration. While mitigating the impact of climate change presents a protracted challenge for utility providers (and everyone else), load forecasting provides the tools necessary to learn from these instances and plan for the next.

Real-time AND Historical Data

The people managing the grid, whether that’s demand flexibility program managers or grid operators, are just that: people. Human memory is a strange thing that is often misleading. Because of that, remembering specific events like an eclipse that happened two decades prior may not be the easiest to recall in the appropriate detail. What that means for things like solar eclipses is a distinct potential to lose the lessons learned from solar eclipses and other extreme weather or natural events. Fortunately, load forecasting technology relies on real-time and historical data, allowing those same program managers or grid operators to refer to the date of the last solar eclipse to obtain information about how much solar generation might have been lost during those events.

Load forecasting is the most effective the larger the dataset. For utilities today, that means decades of potential data before the next solar eclipse, although far less time between hurricane season and other extreme weather events that can and often do disrupt service and result in costly infrastructure upgrades. As such, load forecasting informs demand flexibility strategies like demand response, EV managed charging, or virtual power plants by assessing when demand may exceed supply based on weather or other events.

Load Shifting

As the name implies, load shifting is simply shifting load to off-peak periods of demand. Historically, that load shape has remained relatively consistent with demand peaking in the early evening hours to correspond with seasonal temperature extremes and more consumers at home. Load forecasting is an effective and reliable tool to assess when load-shifting strategies might be necessary, based on real-time and historical data. For example, utilities that currently utilize load forecasting technologies have the opportunity to learn specifics about how this solar eclipse and anything else like it might affect their energy output in the future. This is powerful data to look back on, especially given internal turnover and the fallibility of human memory, informing not just users, but future utility program managers and operations to come.

Examples of load shifting include the aforementioned demand response and EV managed charging. These programs shift customer demand through device control schemes that connect utilities to end users, ceding control of homeowner devices during specified event windows for willing participants. This allows utilities to shift demand to cheaper, less energy-intensive hours of usage, lowering customers’ costs while increasing energy security.

Generation

Virtual power plants are more than a buzzword, but an increasingly prevalent reality. Virtual power plants can function in many ways, including by aggregating the very same community solar assets impacted by solar eclipses and cloudy, inclement weather. On sunny days, one application for virtual power plants is to aggregate the excess solar of participating customers, redistributing it to the grid. This application of demand flexibility is increasingly common and promises to become more so with the proliferation of virtual power plant technology. Load forecasting can guide these initiatives by informing utilities on the optimal times for demand flexibility programs or energy purchases.

Load Forecasting, Solar Eclipses, & More Conclusion

While the next solar eclipse in the U.S. is decades away, that doesn’t mean that the lessons learned on April 8, 2024, won’t continue to yield interesting and informative insights. Especially considering the anticipated increases in volatile weather events and temperature extremes, planning for tomorrow remains crucial. Load forecasting offers an analytical approach to load planning that easily informs the demand events and energy needs that you can expect in the future.