HomeTop StoriesTime-sensitive energy pricing holds long-term benefits

Time-sensitive energy pricing holds long-term benefits

The western United States’ energy turmoil in 2000 and 2001 left many irate people wondering who controls the region’s electricity markets. One thing is certain – consumers, the people who should be in control, who deserve to be in control, are not.

Pick almost any product or service – a shirt, a car, a haircut – and consumer demand is the principal driver of quality, availability and price. Not so with most retail electricity service. Complex laws and regulations have long steered America’s utility industry, prescribing who buys electricity from whom, and at what price. True, US policymakers have dabbled with utility deregulation in recent years, but the results have been modest at best. At worst (think of California) they’ve been disastrous.

As the utility business settles into its second century, it is fair to ask why the benefits of competition are not yet available to customers. Or better, to ask if utility customers will ever be able to have a notable influence on the price of their service.

The crux of this issue is a lack of information that would empower consumers to influence both the price and the supply of electricity. Utilities, with rare exception, don’t give their customers timely and accurate price signals before electricity is used. Consumers can’t see the wholesale price their utility pays for a kilowatt-hour before they in turn buy it. They can’t knowledgeably say: "No, I won’t pay that much. I think I’ll hold off awhile and see if the price comes down."


In America’s Northwest, most utilities in early 2000 were selling their regulated, flat-rate power to customers for about 5 to 6 cents per kilowatt-hour. Suddenly, many of those utilities found themselves having to buy wholesale power for customers at 10 to 20 times the fixed retail price. Only much later – months after the electricity was consumed and utilities had racked up millions of dollars in power-supply cost overruns – did consumers finally see the true price of their energy. The bill arrived, after the fact, with eye-popping rate hikes.

What would these customers have done if their utility had quickly passed through the soaring power supply costs, along with usage information for gauging the size of future bills? Would they have reduced consumption? Would they have shifted some of their electricity use to lower-cost, off-peak hours? Would they have flexed enough consumer muscle to affect wholesale markets and, ultimately, drive down their own power costs?

Puget Sound Energy (PSE) decided it was time to look for answers to such questions. In 1995 we began to explore the merits of AMR. Our initial aim was to see if AMR could increase our operational efficiency, reduce costs and enhance service reliability. In 1997, we employed Cellnet to install 700,000 AMR electricity and gas meters. Eventually, we installed wireless AMR technology in more than 95% of our meters. At the time we were building our AMR system, we were also designing, installing, and fine-tuning a new customer information system we called CLX.

The two integrated systems provided a quantum jump in our customer service capabilities. PSE customer service agents were able to obtain on-demand meter readings instantly and provide near up-to-the-minute usage figures, to resolve high-bill complaints, to eliminate on-site check reads for starting or stopping service, and to know at once when a customer’s power is out of service.

But improved service wasn’t the only constructive aspect of our new system. Just as the Western US energy crisis was firmly taking hold in 2001, PSE realised it had gained the technological wherewithal to take a ground-breaking step in energy demand-side management.


In late 2000 PSE launched the largest time-of-use energy programme in US history. We set out, initially, simply to conduct a customer education campaign. Through bill inserts, advertising, news stories, web site information and speaking engagements, we began to inform customers about the price difference that existed at times between peak and off-peak wholesale power – and what that difference means to them. We told them how the timing of their energy usage, with even small shifts from peak to off-peak usage, can lower their power costs. And with our AMR technology, we gave customers a tool to see precisely how much energy they were using at given times of the day.

We selected over 400,000 residential and business customers to participate in the pilot. We began tracking the timing of their electricity consumption, segregating it on their bills within four time blocks of the day: the morning and late-afternoon periods of peak demand; the moderate-demand midday hours; and the low-demand nighttime and weekend periods.

Customers in the pilot could see their personal, time-sensitive consumption data every day by logging on to PSE’s web site. All participants received a report with their monthly utility bill detailing their individual time-of-use consumption. They could also go to PSE’s web site and use a simple ‘calculator’ to compare their bill under time-of-use rates to a traditional flat-rate billing. 

The timing of PSE’s pilot proved nearly as significant as the timing of customers’ electricity usage. Within a few weeks, wholesale power prices in the American West leaped to astonishing heights. Spot-market prices topped $3,000 per MWh in December 2000 – roughly 100 times the typical pre-crisis rate. PSE and the Washington Utilities and Transportation Commission decided the time was right to test whether time-of-use energy pricing could provide customers– and the region – with some relief.

Starting in May 2001, roughly 300,000 PSE residential customers were converted from the information-only phase of the pilot to time-of-use billing. All were notified by letter that they could opt out of the new billing pilot at any time. Based on negotiations with state regulators and other parties, the difference between the pilot’s peak and off-peak rates was set at 30%.


During the first year of time-of-use billing, PSE customers showed overwhelming support for the experiment. In a survey of about 900 participants, 9 out of 10 had shifted some of their electricity usage to off-peak hours. Half said they had reduced their overall usage; 85% said they would recommend time-of-use billing to a friend. Less than 1% reverted to flat-rate billing. Just as important, over the first 13 months of time-of-use billing, customers paying time-sensitive rates used about 5% less energy during peak-demand hours than customers who got personal time-of-use consumption data but remained on flat-rate billing.

To PSE, the evidence seemed clear. Time-of-use rates, paired with an aggressive campaign to educate customers about efficient energy use, could generate positive results.

By mid-2002, however, the region’s energy crisis had subsided, and differences between on- and off-peak wholesale prices largely disappeared. Regional power supplies, for the near-term at least, appeared adequate. As a result, state regulators, PSE, and other stakeholders agreed on a lower price spread between the pilot’s peak and off-peak rates, effective July 2002. In addition, a new $1 monthly fee was levied to help cover the cost of time-of-use information and data management.

Those two changes, we soon learned, had eliminated the bill savings most customers experienced during the pilot’s first 13 months. In October 2002 PSE mailed all customers in the pilot a comparison of their time-of-use bill over the prior quarter with what they would have paid under our flat rate. The results, both for PSE and our customers, were disappointing. 94% of our customers had paid more – about 80 cents per month, on average – with time-of-use rates.

Within a few weeks about 36,000 customers dropped out of the pilot. Many were angry, saying they used less electricity in peak-demand hours but were not being rewarded for their effort. Some felt deceived. With the reversal in customer sentiment, the only thing for PSE to do was suspend the pilot and re-evaluate our options.


Can a demand-management programme based on time-sensitive retail pricing still prove worthwhile? Without hesitation, I believe the answer is “yes”.

First of all, there is no denying that steep spikes in power demand create a less efficient, more costly energy system. More power plants and power-delivery facilities (all built at great expense) are required to meet those demand spikes. As PSE’s time-of-use pilot proved, consumers can and will reduce their peak-period energy if they believe they are doing “the right thing” – even if their bill savings are small. (A majority of PSE customers paying time-of-use rates saved 59 cents per month, on average, during the pilot’s first year.)

But the single greatest benefit of time-sensitive energy pricing is the downward pressure consumers can apply to wholesale energy prices when demand – and price – begin to increase. According to a McKinsey & Company study, if all American consumers shifted just 10% of their peak-period energy use to off-peak times, they would save up to $15 billion annually in power costs.

The real question is: When exactly does time-sensitive pricing make sense? From PSE’s experience, periods of market volatility and high power costs provide the greatest potential for customer reward. On the flip side, when wholesale markets are relatively flat and utilities’ power supply costs are comparably low, the economic basis for high customer incentives (a wide spread between peak and off-peak rates) simply does not exist.

Although PSE suspended its pilot ten months ahead of schedule, we are continuing to study the customer usage data we compiled. Later this year, a report will be issued on the program’s costs and benefits for customers, from both an economic and an environmental standpoint.

In the end, I believe we’ll find that time-sensitive energy pricing can be shown unequivocally to have long-term benefits for utility customers. Yes, there may be periods when energy-market economics do not fully justify customer incentives in a time-based demand management programme. The same can be said about energy conservation programmes. But over the long haul, conservation has proven itself to be cost-effective. If we resist the temptation to demand immediate results, if we look instead five or ten years into the future and design demand-management programmes accordingly, I’m confident they, like conservation programmes, will be a valuable part of our energy future.