Berkeley Lab Study May Help Improve Monitoring of Western Electricity Markets

By Consulting Specifying Engineer Staff February 27, 2007

A new study, announced by the U.S. Dept. of Energy’s Lawrence Berkeley National Laboratory in late January, has evaluated methods for monitoring power markets that should allow regulators to more quickly identify the efforts required to identify unusual fluctuations in the price of power in the Western Interconnection.

“This study is another important step to ensure that the failures in monitoring power markets that contributed to the 2000-01 western electricity crisis are not repeated,” said Joanna Prukop,improve western power markets.”  

WIEB’s board members include policymakers and representatives of utility regulatory bodies from 12 western U.S. states and three western Canadian provinces.

Berkeley Lab’s researchers, who worked with the Analysis Group, found that econometricprice prediction methods provide a benchmark screen for identifying possible instances of uncompetitive pricing in western power markets. The study surveyed analytical methods that can apply to market monitoring in the West, where there are few organized wholesale markets and individual wholesale transactions are often opaque.

Charles Goldman of Berkeley Lab’s environmental energy technologies division, a co-author of the report, explained that “in the electric industry, market monitoring involves the systematic analysis of prices and behavior in power markets to determine whenand whether potentially anti-competitive behavior is occurring. Our study was designed to explore different methods of doing this, using readily available data.”

Ed Kahn, another of the report’s co-authors, adds that “the basic goal is to develop a benchmark for wholesale prices in a well-functioning competitive market. Once you’ve done that, you can compare actual prices to the benchmark and identify ‘outliers’—instances where suppliers may have manipulated prices and further investigation is warranted.”

Researchers reviewed five potentialanalytic methods to test their feasibility for market monitoring. They selected the two most promising methods—econometric analyses and production cost modeling—and tested them using day-ahead market data from two electricity trading hubs in the west:  Palo Verde in Arizona, and theMid-Columbia hub in the Pacific Northwest.

Kahn explains that “production cost models are detailed simulations of the operation of power systems that are often used for resource planning.” He added that “when adapted for market monitoring purposes, their complexity is both ablessing and a curse. They provide a lot of detail that simpler models cannot handle, but their workings are somewhat of a ‘black box’ and they are time-consuming and data-intensive to implement.” In the end, the researchers found that production cost simulations based on several discrete scenarios of future supply/demand conditions in the west did a relatively poor job of predicting seasonal variations in actual wholesale electricity prices at the Palo Verde or Mid-Columbia hubs.

“The results for the econometric analyses were much more promising,” Goldman said. Econometric analyses are well-established statistical methods that can be applied to model the relationship between a set of fundamental price drivers (such as weather and fuel prices) and wholesale power prices.

“For the Palo Verde hub data, we developed a model that predicts over 90% of the observed price variations,” Goldman said. He added, “We were able to link all the outliers we found in the Palo Verde data to well-defined events—such as outages of major coal generation plants—that were not captured by variables in our model.”

However, the report’s authors caution that econometric methods do have some pitfalls. For the mid-Columbia hub, the results were not as compelling. Kahn said, “We suspect that the problem lies in capturing the effects of spring runoff on the hydroelectric system.  If this type of model was actually used for market monitoring screening, it would require more attention to the unique dynamics of hydroelectricity in the Pacific Northwest.”

Overall, the researchers are optimistic that econometric analyses could provide a promising market monitoring screening tool for the west, but they caution that it must be “implemented and interpreted with care.” “A Regional Approach to Market Monitoring in the West” (LBNL-61313), is authored by Matthew Barmack, Edward Kahn and Susan Tierney of the Analysis Group, and Charles Goldman of Berkeley Lab.