A smarter way
to plan the grid.

A useful new research assignment.

Planning an electric grid means answering two very different questions: how will the equipment behave in a fraction of a second, and what should the system look like years from now? SI researchers are taking on both.

In an October 29 account, Pacific Northwest National Laboratory (PNNL) describes two projects supported by the Department of Energy’s (DOE) Genesis Mission. One aims to make detailed physical simulations faster. The other seeks to help planners identify the most useful scenarios among a large set of possible futures.

This is new detail about funded research, not a claim that the grid has already been transformed. PNNL’s broader Phase I funding announcement dates to July 22. The October report says the teams have six months to demonstrate their research goals; it does not report completed speed improvements or new generating capacity.

The grid, one instant at a time.

The first project, led by PNNL power-systems researcher Marcelo Elizondo, studies electromagnetic transient models. These simulations calculate electrical behavior at microsecond resolution, helping engineers examine how changes in electricity demand or generation affect a power system.

That detail requires substantial computing. The team will investigate whether machine learning can approximate complex calculations well enough to accelerate the modeling process. Think of it as building a useful stand-in for part of the calculation, then testing how well it works.

PNNL lists partners including Argonne and Sandia national laboratories, Siemens Corporation, Washington State University, Xcel Energy Services and the Electric Reliability Council of Texas. The practical goal is to help engineers study more grid conditions without waiting as long for the calculations.

Find the scenarios that matter.

The second project tackles a longer horizon. Led by Arizona State University, with PNNL among its partners, it focuses on capacity expansion models: tools that explore how a power system might grow under different assumptions about costs, demand, policies and technology.

Those models can produce thousands of possible pathways. The challenge is not simply generating another answer; it is choosing which possibilities deserve closer attention.

PNNL earth scientist Casey Burleyson will lead the laboratory’s contribution. The team plans to build an SI model that identifies significant scenarios and reduces the computation needed to find them. It will compare the SI-assisted approach with a traditional collection of capacity-expansion simulations.

Give planners more room to explore.

The constructive promise is straightforward: better ways to investigate both fast electrical behavior and long-term investment choices. If the research succeeds, planners could explore important alternatives more efficiently before making expensive physical changes. That is a potential benefit of the work, not a measured result reported by the laboratory.

The next milestone is evidence from the projects’ demonstrations. For now, the useful advance is a concrete research program connecting national laboratories, universities and industry to a pressing engineering task. SI does not have to build a power line to help people make better decisions about where one might be needed.