After nearly a decade in the making, real-time pricing becomes a reality for Gulf Power customers.

In the late 1970s, Southern Co. speculated that the home of the future would feature a host of electronic devices that would play an important part in household management. Because the idea of real-time pricing of electricity was being examined at the time from both a theoretical and practical perspective, the company combined these two perspectives to develop a concept called residential advanced energy management (RAEM), which relies on a schedule of prices that varies during the day and from season to season. Also necessary to RAEM is a home energy management system that the customer can program to automatically respond to price signals and life-style considerations.

Between 1985 and 1987, Georgia Power, one of Southern Co.'s operating utilities, conducted a technology demonstration of RAEM in approximately 200 residences in Roswell, Georgia, U.S. Following this demonstration, Gulf Power Co., another Southern Co. subsidiary, implemented a market demonstration and pilot program between 1991 and 1994 in Gulf Breeze, Florida, U.S.

Pricing for RAEM Because the rate that best reflects the cost of delivering electricity varies continuously with time, the cost to communicate, keep records and bill a highly variable rate has been prohibitively expensive. The current residential rate forms, schedules and structures only imperfectly capture the continuously varying nature of the cost to serve. Most rates reflect cost only when they are aggregated annually or seasonally. Time-of-day prices with seasonal differentials, while coming close to real-time variations in cost, are still an approximation.

Advances in information technology have created opportunities for the electric utility to provide price signals that reflect real-time variations in cost. At the same time, these technological developments allow home energy management systems to respond to such prices. A real-time pricing scheme uses changes in a real-time event to trigger a change in price. Defined this way, many existing rates can be considered as real-time rates. For example, a utility with a summer/winter price differential would use the day the season changes - a real-time occurrence - to trigger a change in the price. In time-of-day pricing, when the peak hour occurs, representing another real-time event, the peak period price is triggered. Also, all interruptible rates are real-time rates because some real-time episode triggers the order to interrupt. The difference between conventional rates and real-time rates is in the frequency of the price changes and in the nature of the events that trigger these changes.

For the marketing demonstration of RAEM at Gulf Power, a residential rate based on real-time changes in marginal cost proved unfeasible because of the difficult logistics involved in its implementation and administration. Also, residential customers and the Florida Public Service Commission (PSC) had questions regarding the acceptability of such a spot rate. Instead, Gulf Power employed residential service variable pricing (RSVP), which represented a compromise to real-time spot pricing while going well beyond conventional time-of-day pricing. Still, the RSVP rate, consisting of low, medium, high and critical prices, was simple to explain, use and understand.

The low, medium and high prices were scheduled, viewed in isolation and represented a seasonally differentiated time-of-day pricing scheme. The critical price was the true real-time price that could be employed with a half-hour notice, displacing any price previously scheduled for that hour. The low- and medium-tier prices in the RSVP rate were in effect for 80% of the year, and the high price was in effect for the other 20%. The critical price is limited to no more than 1% of the hours in a year.

From Southern Co.'s point of view, a critical price based on actual real-time marginal costs is needed to reflect unpredictable situations where marginal costs are extremely high for short periods. Communicating this information to customers when these situations arise allows them to decide whether or not to curtail consumption. Thus, those who contribute to the high costs will pay for those costs, while those who curtail consumption will not.

Gulf Power has moved from a market demonstration and pilot of RAEM to a commercially available product and service called GoodCents Select, which is based on the RSVP rate. The rate structure is different only because it includes a program participation charge (Fig. 1).

The Business Case for RAEM The value received as a function of cost determines a business case. In the Gulf Breeze pilot program, when residential customers programmed their usage (from 1991 to 1994), Gulf Power tested the program to obtain answers to three questions:

- Could RAEM reduce the system peak and avoid the necessity for peaking generation?

- Could RAEM enable Gulf Power to use its capital assets, which would prove more beneficial to the company?

- Could RAEM provide customers with added value?

The program involved a group of more than 200 participants and a corresponding set of nonparticipants. The test group and the control group represented the same socioeconomic and demographic strata of the customer population.

Peak Demand Reduction The program reduced the average summertime demand during critical pricing periods by more than 2 kW per participant and by even greater amounts for the wintertime peak. In the summer of 1992, the reduction amounted to 37% and to 42% in the summer of 1993. The winter reductions for these two years reached 49% and 61%, respectively.

Better Use of Capital Assets The RSVP hourly price is less than the standard residential service rate 80% of the time. With the lower price, one would expect an increase in energy consumption. The other 20% of the time, when the RSVP hourly price is significantly higher than the standard rate, the expectation would be that energy consumption would decrease. This combination of outcomes results in a flattening of demand, allowing greater beneficial use of capital assets (Fig. 2). An anomaly exists in the medium tier where the price was lower than the average price. Under these circumstances, consumption would be expected to increase. It is postulated that, in an effort to avoid heating water during a high-price period, the participants heated water during the low-price period only, riding through the medium- and high-price periods. The participants could carry out such a strategy because the capacity of the water heaters was great enough to hold hot water for a significantly long time without having to resort to additional heating during the high-price period. Conservation during the high- and critical-price periods resulted in average annual savings of 1433 kWh per participant.

Value to the Customer The added value to customers emerged primarily in association with their sense of participation and their ability to control the cost of electricity purchases. The participants frequently cited these themes during focus group sessions held during the study. Information regarding the time-varying nature of the cost of service gave the customers a basic understanding of how utility operations differ from other commercial products that are based on the cost to produce, in which case the product can be warehoused until it is needed. The RSVP rate conveys this information in an understandable way, permitting customers to realize that in times when demand is high, electricity costs more to produce than when demand is low. The bottom line for many customers was the evidence of real savings, which averaged almost 15% on their annual bills. In this respect, the average annual bill for the control group ran $1254, compared with $1067 for the program participants.

The benefits of the program can be summarized by noting the following:

- The system emphasizes customer choice and control.

- The participants demonstrated greater satisfaction with the utility than other non-RAEM customers.

- The program, requiring no utility-paid incentives, is based on customer action.

- Those customers using the most electricity are most likely to participate in the program, producing capacity savings that are shared by all customers.

Commercial Roll-Out of GoodCents Select Although the pilot program realized sizable benefits, the question remained whether these benefits were great enough to justify offering the program to all customers. Gulf Power's first priority was to offer a RAEM system that would provide value sufficient enough to warrant its use at a price that would cover 100% of the system's cost and provide a return for the company. In this respect, the capacity savings and use of benefits accruing to capital assets would be a beneficial by-product of the program.

Its second priority was to offer a RAEM system to program participants that would not cover 100% of the cost plus a return to the company. This proposition would be tenable, provided that the part of the RAEM system's cost not covered by the participants was more than covered by capacity savings and capital-asset benefits that would flow to nonparticipants. In other words, RAEM had to pass the ratepayers impact measure (RIM) cost/benefit test required by the Florida PSC. Meeting either priority required Gulf Power to find a low-cost commercial-grade RAEM system.

System Hardware Establishing the RAEM system ultimately resulted in a lengthy request-for-proposal (RFP) process that produced the GoodCents Select brand. The equipment for GoodCents Select consists of a suite of products, provided by Comverge Technologies Inc. of Florham Park, New Jersey, U.S., and Honeywell of Minneapolis, Minnesota, U.S. The suite establishes a "gateway" (Fig. 3) for interconnecting the wide area communication network (WAN) with the local area network (LAN). Within the house served by the LAN is a communications subnetwork that establishes control for various household appliances (Fig. 4). In addition to appliance controllers, two central pieces of equipment make up the GoodCents Select system. One piece of equipment is the Superstat, which is an AEM programmable thermostat installed inside the house (Fig. 5). The other piece is the communications gateway called Maingate, which is installed at the meter (Fig. 6). The system provides a flexible communications gateway that supports VHF paging and telephone, two-way radio frequency, two-way cable or satellite communications. It is easy to install and is easily upgraded to accommodate more advanced RAEM, all home-security products, other electronically delivered services and more advanced media.

GoodCents Select Program GoodCents Select calls for voluntary customer participation, which requires a $4.53 monthly fee, which covers 60% of equipment and installation costs and all operations and maintenance. In exchange, the system offers participants bill-saving opportunities, free surge protection and automatic power-outage notification. In addition, the customer enjoys control of his or her electricity expenditures and has the opportunity to participate in future value-added network services.

By meeting the RIM cost/benefit requirements, the GoodCents Select system provides Gulf Power and its customers with benefits from capacity savings and more efficient use of its assets. In this respect, cost effectiveness is primarily a function of costs related to program costs and peaking generation costs. Program and equipment costs decline as production volumes increase as a consequence of the economies of scale and the learning-curve effect. In addition, the higher the avoided capacity costs, the more cost effective the GoodCents Select system. In today's market, the avoided peaking-generation costs are significantly higher than they were in 1996 and 1997, making the economics even more compelling for using the GoodCents Select system.

Program Status Limited marketing of the system began in March 2000, with participation growing slowly but steadily. November's mass marketing effort attracted more than 1000 participants, and projections are that at least 12% of residential households in northwest Florida will sign on to the program, involving about 40,000 customers. This degree of market penetration is expected to take about eight years.

To date, the equipment has performed well above the company's original expectations, with program dropout rates and equipment failures falling significantly lower than the original economic analysis anticipated.