Overbye awarded BPA grant to verify, validate power grid

4/30/2013 11:15:00 AM ITI Staff

When the loss of a single power line in Arizona caused the September 2011 blackouts in Southern California and the Southwest, nearly seven million people were left without power for over 11 hours. The outage caused major traffic congestion and sewage pumping station failures, as well as resulting in millions of dollars of losses to restaurants and grocery stores.

Written by ITI Staff

When the loss of a single power line in Arizona caused the September 2011 blackouts in Southern California and the Southwest, nearly seven million people were left without power for over 11 hours. The outage caused major traffic congestion and sewage pumping station failures, as well as resulting in millions of dollars of losses to restaurants and grocery stores.

In hopes of preventing future catastrophes like the Great Blackout of 2011, ITI researcher Thomas Overbye is working to help create automated tools to perform transient stability software verification for Bonneville Power Administration (BPA), a federal nonprofit agency and part of the U.S. Department of Energy. Overbye, who was recently elected to membership in the National Academy of Engineering, has been awarded a two-year, $611,724 grant to conduct research on his Verification and Validation of Transient Stability Models and Results project.

This figure shows the difference in the simulation results of a generator model obtained from two software packages. The cause of the discrepancy was found to be a modeling issue in Software B, which was identified and fixed. The figure reflects the verified results.
This figure shows the difference in the simulation results of a generator model obtained from two software packages. The cause of the discrepancy was found to be a modeling issue in Software B, which was identified and fixed. The figure reflects the verified results.
This figure shows the difference in the simulation results of a generator model obtained from two software packages. The cause of the discrepancy was found to be a modeling issue in Software B, which was identified and fixed. The figure reflects the verified results.

Overbye, along with ITI researcher Komal Shetye and Washington State University Electrical Engineering and Computer Science Professor Mani Venkatasubramanian, will be testing software packages that BPA currently uses to run their dynamic simulation planning studies.

The electric grid is our most critical infrastructure, said Overbye, the Fox Family Professor in Electrical and Computer Engineering. In order to operate the grid efficiently we need to have simulation results that match the actual grid behavior. While we as an industry have certainly made progress over the last several years, the September 2011 blackout highlights the need for additional research in this area.

Power companies perform planning studies to determine how they are going to operate the power grid. According to Shetye, to perform planning studies, simulations are run that measure the response of the grid model to various system contingencies, such as what would happen if a key generator failed. The goal of this research is to provide BPA with tools to facilitate better planning studies on their power system model.

The grid is supposed to handle some level of stress and continue operating safely, Shetye said. In transient stability simulations, we apply some stress, or a contingency, to a power system model and see if it can recover. If the model shows it will recover, but in reality it doesn't recover, you need to improve your model. Better planning studies and better models are critical to the safe and optimal operation of the grid.

All the major power companies in the U.S. use standard software packages to run these simulations and do studies. Overbye and Shetye want to verify both that these software packages that are commonly used agree with each other, and that they agree with the actual grid. If there are differences in the results, it can affect decisions people make when operating their power grid.

While ultimately we will be comparing the software packages to the performance of the actual grid, initially we'll be comparing the software packages with each other, Overbye said. Comparing the software simulations to actual grid behavior is a much more challenging task since obviously we can't deliberately apply severe contingencies to the actual grid to see if it will fail.

After the models are tested, Overbye and Shetye will identify the discrepancies and determine why they are happening. The end result is to create tools for BPA to continue this software verification and to update their power system model to ensure performance accuracy.

We want to give the BPA the tools to verify their software packages they use and we're doing the validation of the models with actual system measurements, so that we can improve the models themselves, Shetye said. They will come out of this with better tools and better models to perform their studies.

This research is a continuation of an earlier project, Validation and Accreditation of Transient Stability Results with the Power Systems Engineering Research Center, done by Overbye and Venkatasubramanian, which was published in September 2011.


Share this story

This story was published April 30, 2013.