Today at Berkeley Lab nameplate Berkeley Lab
Friday, July 30, 2004


















Berkeley Lab Scopes Out Demand Response

Successful demand response requires more than just subjecting customers to the vagaries of wholesale power prices, says a new study by researchers at Lawrence Berkeley National Laboratory in California.

The LBL researchers looked closed at Niagara Mohawk s use of real-time pricing to shift and limit loads in response to hourly wholesale market price changes. The program was moderately successful, the LBL team concluded, but commercial and industrial customer behavior is considerably more complex than many would have assumed. The work was funded by the California Energy Commission, which is looking into various models for demand response programs.

Conventional wisdom holds that industrial customers will be the most responsive to price signals. But the LBL study of NiMo and the New York Independent System Operator programs found that government and educational customers were the most responsive to market price signals. Commercial customers were least able to respond to real-time prices. The researchers conclude, because commercial customers (e.g., office buildings) have similar physical characteristics and end-use load (e.g., space conditioning and lighting) to government and education facilities, response from this sector is at least technically feasible. If institutional and other barriers can be overcome, the commercial sector may provide a rich source of price response.

The NiMo study, said CEC Commissioner Arthur Rosenfeld, provides useful insights to California policymakers about the building blocks for a successful program. These include transparent wholesale market prices, a broad array of options for those customers that want to hedge and mitigate price volatility, and technical assistance and information tools that help customer manage their facility loads in order to respond effectively to wholesale market price signals.

In the fall of 1998, NiMo offered real-time pricing as part of its default service to customers with demand in excess of 2 MW. Under the program, NiMo charges customers the hourly day-ahead location-based market price established by NYISO. NiMo posts the following day's hourly prices by 4 p.m. so customers can plan the next day's usage.

About half of the 130 customers in the study were government and education facilities, with industrial customers representing another third, and commercial customers the rest. The customers had an average monthly peak demand of 3.4 MW. During 1998-2003, about 55 percent bought commodity power from competitive suppliers. Peak prices were fairly volatile, increasing by 15-30 percent in various regional zones.

Generally, said Charles Goldman, the report's lead author, customers are satisfied with real-time pricing as the default tariff, although some said that price hedging options are not attractively priced relative to perceived risks. Some 35 percent of survey participants were hedged, either through fixed-rate contracts or financial hedging products. The rest were fully exposed to varying prices in the wholesale market. Overall price response, the study concluded, was modest but encouraging, and individual customer responses was extremely variable.

The Niagara Mohawk experience, the report concludes, shows that large customers are likely to provide a moderate amount of demand response when real-time pricing is their default service tariff, even if some customers hedge against price volatility. However, subjecting customers to wholesale market variability is not sufficient to realize their full demand response potential.


Back to "Today at Berkeley Lab"

DOE logo Office of Science logo UC logo