I recently had the pleasure of attending the Probabilistic Methods Applied to Power Systems (PMAPS) conference in Boise, ID. The conference brought some of the brightest minds together to explore how computer models and mathematics can help keep the lights on.
It was inspiring to see both the great depth of experience and expertise of the researchers and planners, as well as the enthusiasm of students entering the field. It’s clear how important research has been, and how critical it will continue to be, for the future of the electric grid.
Developing the electric grid didn’t seem like a very exciting field 20 years ago, but the rise of renewable generation, storage, and electric vehicles has revitalized the research community. Academic researchers are more connected to the utility industry because of these changes.
While there were plenty of technical talks laden with mathematical equations, some general themes emerged.
There are challenges with the traditional ways the electric grid has been operated and expanded. Kevin Carden, from Astrape, showed an example of observed thermal generator outages that didn’t match model assumptions.
Chris Dent, from the University of Edinburgh, talked about how new generating technologies may complicate the traditional ways reliability of the electric grid is measured. Technology is changing quickly, and planners are striving to adapt.
And yet, traditional approaches still have value. Dimitry Gorinevsky, from Stanford University, talked about using machine learning to transform the impact of wind and solar generation on reliability into a problem that would be recognized by power planners from the 70s. The basic structure was more like an evolution of power system planning rather than starting over from scratch.
Brandon Heath, from the Midcontinent Independent System Operator, presented on adapting traditional planning approaches to measure the impact of increasing levels of renewable generation in the MISO footprint.
The structure of energy markets and methods of operations also present challenges. Ramteen Sioshansi, from Ohio State University, talked about the storage operation problem. He showed that having a price for capacity, not just energy, in the market can increase the reliability of storage. So, with the same battery and different market structures you would have a more or less reliable system. Without a price for capacity in the market, you need to build a battery that can store more energy (MWh) to gain a similar grid reliability.
Power planners struggle with computer models used to help plan the system. Dent noted that complex models lead to functions that are undefined almost everywhere. When looking at thousands of variables that interact, computers can only go so far in exploring possible solutions to these complex problems.
Sioshansi discussed the complexity of modeling state-of-charge for storage projects. The computational load of current computer models for the electric grid came up often.
Computer models also affect the physical grid. Utilities depend on models, planners, and operators to identify where reliability upgrades are needed.
Noha Abdel-Karim, from NERC, noted that the historic model of planning will not fit the grid of the future, challenging those responsible for maintaining and upgrading the grid. NERC recommends expanding the use of probabilistic approaches to power planning. That expansion will depend on the planners of the future having better tools to use in figuring out how to keep the grid reliable in a quickly changing environment.
PMAPS shows a bright and interesting future for power planning. It was exciting to be able to participate in the conference with some of the cutting-edge researchers on power planning here in the Northwest. There were many sessions I was unable to attend, where I’m positive many more interesting topics were covered.
This was my impression of the larger themes of the sessions I attended. If there are errors or omissions, the fault is in my understanding. I hope this gives you a sense of the state of research for planning on the power grid.