Firms are awash in newly accessible data that promises to make their operations more efficient. The electricity industry is no exception: enter the smart meter.

“Before smart meters, there was only one data point per consumer per month,” explains Ozge Islegen, an assistant professor of managerial economics and decision sciences at the Kellogg School. Every month, utility companies received a single update about a household’s usage—and customers received a single bill. But smart meters, installed in customers’ homes, can monitor usage in intervals of an hour or less.

Learn more about leading with big data and analytics at the Kellogg Executive Education program.

Critically, the rich temporal data allows utility companies to experiment with charging more for electricity used during peak hours. Such pricing strategies have the potential to increase the efficiency of the entire supply chain.

Utility companies believe that giving customers financial incentives to shift their usage from peak hours to off-peak times will flatten the demand curve for electricity—with customers waiting until night to run the dryer, for instance. For utility companies, this means a need for fewer expensive generators, some of which currently run only during peak times. “If the demand curve was flatter and there wasn’t a significant peak, then the investments for the costly peak-load power plants would decrease,” says Islegen.

So, how does time-based pricing affect the electricity supply chain? In recent research, Islegen—along with colleagues Asligul Serasu Duran, a PhD student at the Kellogg School of Management, and Baris Ata of the University of Chicago Booth School—explores the impact of time-based pricing schemes on electricity consumers and electricity providers—as well as on the carbon emissions from electricity generation.

“Utilities promote these strategies by saying they’re environmentally friendly, but we find that’s not necessarily the case.”

What the researchers find is that peak use does in fact go down—though not necessarily overall use. And a time-based pricing scheme need not be complicated to be effective. However, the short-term impact on customers’ bills is minimal, so utility companies and policy makers who want to reap the long-term benefits of a flatter demand curve should consider making time-based pricing an opt-out affair. The environmental impact of these pricing strategies, moreover, depends heavily on local market conditions.

Simple Time-Based Pricing in Ireland Reduces the Peak Load

The lure of a more efficient electricity supply chain is undeniable. The U.S. is scrambling to adopt advanced meters that enable time-based pricing strategies. “As of 2013, electric utilities had over 46 million advanced meters installed for residential consumers,” the researchers report, “which accounts for 36% of all residential consumers.” Movement is similarly frenzied elsewhere in the world. The European Union, for instance, aims to equip 80 percent of households with smart meters by 2020.

In their investigation of time-based pricing, Islegen and her colleagues turned to data from a field experiment in Ireland. “We wanted to know if these pricing schemes—charging a higher price during peak times and lower price in off-peak times—actually change customer behavior and shift demand,” Islegen says.

Data from nearly 3,500 Irish households was collected by the Irish Commission for Energy Regulation. To get a benchmark for electricity use, households initially paid a flat rate of 14.1 cents per kilowatt hour (kWh). Then, homes were divided into groups, each of which was assigned a different time-of-use tariff (with prices ranging from 20 to 38 cents per kWh for peak electricity, and 9 to 14.1 cents per kWh for off-peak electricity). The Commission also included a control group that continued to pay a flat rate of 14.1 cents per kWh. Throughout 2010, the households’ electricity use was monitored every 30 minutes.

Along with this data, the researchers used household demographic data, as well as consumption data, to build a model that identified optimal retail prices for several time-based pricing strategies. The strategies under consideration were real-time pricing and time-of-use pricing. Real-time pricing allows utilities to update electricity prices in response to fluctuations in the wholesale electricity market. With time-of-use pricing, on the other hand, prime-time electricity usage—between the hours of 5 to 7 P.M., say—is set at a predetermined rate that is higher than the rate for off-peak usage.

Islegen and her colleagues found that time-based pricing strategies do indeed reduce peak loads in Ireland, from about 0.5 to more than 11 percent, depending on the season and the severity of the peak electricity surcharge. Time-based pricing did not, however, significantly change overall electricity consumption.

Interestingly, the researchers found that simply charging customers a higher rate for prime-time usage was just as effective at reducing peak loads as adopting the more complicated real-time pricing schemes. This came as a pleasant surprise to Islegen, as the time-of-use tariffs let customers avoid price spikes without having to constantly track prices.

For Ireland, where electricity usage drops dramatically in the summer but spikes during the winter holidays, time-of-use prices should vary depending on the season. “In December, for example, the peak load is really significant—approaching the capacity constraint of the generators—while in the summer, there’s no significant peak load at all,” says Islegen.

This particular pricing scheme is optimized to Ireland’s environment and social demographics. “For example, air conditioning doesn’t play a significant role in Ireland, but 13 percent of the residential electricity consumption in the U.S. comes from space cooling,” Islegen says. But using the framework developed in this research, similar studies could identify the ideal pricing scheme for other markets.

Impact on Customers and the Environment

Although peak load was reduced when time-of-use pricing was adopted, greater efficiency in the supply chain did not necessarily translate to lower bills for customers. For this reason, the researchers are skeptical that customers will voluntarily flock to the new pricing schemes. They suggest instead that utilities and policy makers offer time-of-use pricing as the default model. Indeed, the researchers report that this default time-of-use pricing model has been adopted by Ireland and other electricity markets such as Ontario, Canada, and Italy.

Time-based pricing, the researchers find, is also no panacea for curbing greenhouse gas emissions. In Ireland’s case, emissions remained about the same when time-based pricing was applied. But whether emissions are reduced, remain the same, or even increase will likely vary by region. The environmental impact largely depends on the precise mix of electricity generators that are used for handling base-load versus peak-load production—as well as other factors like the dispatch decisions of these generators, the electricity market structure, and the consumption profile and demographics of the region under study.

“Utilities promote these strategies by saying they’re environmentally friendly, but we find that’s not necessarily the case,” Islegen points out. “There is a possibility that, depending on the energy mix in the region, these pricing schemes can have a negative environmental impact.”

If base-load production is powered by coal but peak load is handled by “greener” power plants like natural gas plants—as is typical in many parts of the U.S.—then a shift in demand from peak load to base load would trigger more emissions.