The successful integration of the growing amount of wind and solar power into the bulk power system will require broader and more detailed weather models, according to a report from the Energy Systems Integration Group.
The report, Weather Dataset Needs for Planning and Analyzing Modern Power Systems, discusses the gaps in existing weather data and modeling approaches and outlines a process for building more robust weather datasets.
“Modern power system models must rigorously account for the impact of weather on supply, demand, and system infrastructure,” Justin Sharp, the lead on the weather data project, said in a statement. “The weather data being used today are not up to this task, and their shortcomings are poorly understood by the sector, which could have serious consequences for reliability.”
Going forward, available generation will increasingly be defined by the weather occurring at the location of every wind or solar plant and variables, the report said, noting that temperature, wind, and solar irradiance, in particular, will increasingly affect the amount of generation possible.
In the past, the impact of temperature on demand was viewed as the most important factor in electric system reliability, but the electrification of transportation and heating are making end-use loads more susceptible to weather extremes, especially in winter, the report said.
To more accurately reflect the changing nature of the electric grid, between 10 and 40 years, or even more, of weather data will need to incorporated into planning scenarios, the report’s authors said.
The authors argue that using available observations and sophisticated computer programs that model the laws governing atmospheric processes, it is possible to fill in many of the data gaps. They also noted, however, that weather model outputs have limitations.
“All too often, synthetic weather data produced by these models are either used in power system modeling as if they are equivalent to high-quality observations, or, on the other end of the scale, model output is rejected in favor of simpler, easier-to-understand observational records that are then extrapolated using statistical methods with dubious scientific basis. Both outcomes lead to study results that have greater uncertainty than is typically advertised and may result in poor downstream decisions when model- synthesized data that ‘seem reasonable’ are assumed to accurately reflect actual present or future conditions,” the authors wrote in the report.
The solution may mean installing new observation stations in places where wind and solar are likely to be deployed, but “the most obvious and cheapest solution” is to make the thousands of observations now available at existing and future wind and solar plants generally available, the authors said. They added that “keeping data proprietary is counterproductive” and must change to enable more accurate system modeling, including a better understanding of the accuracy and uncertainty of synthetic time series of weather data.
“There can be no reliable energy transition without broadly available, consistent, weather datasets for power system studies,” the report said, adding that “the necessary data can be considered a public good, one that is government funded, publicly available, and routinely maintained.”
The Energy Systems Integration Group, previously known as the Utility Wind Integration Group, is a nonprofit educational organization that aims to chart the future of grid transformation and energy systems integration.