techniques may imply trade-offs in service

quality.By Scott L. Englander, John E. Flory,

Leslie K. Norford, and Richard D. TaborsAs facility manager for a large hotel, you browse your energy vendor's web site to view tomorrow's hourly prices. But it seems your computer (pc) has already done some browsing of its own. Since it's connected to your energy management system, your pc has already looked up the weather forecast and has logged on to the hotel's main computer to find out what rooms will be used. Using this information, your computer is now churning through several billion operating scenarios to pick the least expensive energy strategy for the following day.

A while later, an incoming e-mail message describes the operating schedules for your facility's mechanical systems. You see that while prices will skyrocket to over $1 per kilowatt-hour (Kwh) for a couple of hours in the afternoon, your pc has planned an operating scenario that will take advantage of bargain basement prices during the early morning hours to shift certain loads, and will run your auxiliary generators when the high prices kick in. A comparison shows you that your total bill for the month will undercut what you would have paid under your previous conventional electricity rate. You leave for the day, knowing that your energy bills will once more come in under budget.

A fantasy? Perhaps, but not far from today's reality. At the core of this scenario lies the concept of "real-time" pricing, or RTP: that electricity can be bought, sold, and traded in a way that takes into account variations in cost related to time and location.1

In today's market, traditional flat or time-of-use (TOU) rates still predominate. They reflect the supplier's marginal costs only in an average sense. As such, they skew consumer behavior and create subsidies within and between customer classes. By contrast, RTP offers a clearer economic signal, motivating customers to adjust patterns of use to match the utility's marginal costs.

During the past few years, electricity providers have begun offering real-time prices, even while heavily regulated. RTP or variable pricing rates are currently available at over one-third of the utilities that serve the 60 largest cities in the United States.2 Initially, such rates were used for load management (especially for local transmission and distribution constraints). More recently, RTP is seen as a tool to attract or retain customers.

In a competitive retail market, RTP will likely become the dominant foundation of electricity market transactions. Wholesale spot prices will in all likelihood serve as the basis for the energy component of retail RTP.

The California Public Utilities Commission (CPUC) envisions retail customers having the option to receive hourly prices that reflect the changing wholesale prices of the Western Power Exchange. The CPUC also envisions retail customers making use of contracts for differences, forward contracts and other risk management instruments similar to those available to wholesale customers.

Thus, RTP will retain a role as the market evolves. The basis of prices will shift from the production costs of the local utility to those of the regional wholesale market. As the CPUC has noted, the opportunities for improving asset utilization and cost control in this capital-intensive industry still remain.3

Attributes: Real Time vs. Time of Use

The details vary widely, but RTP plans share several common factors. Real-time retail electric prices typi-cally do not change frequently in real time, but tend to remain constant for periods that range from a half-hour to five hours (hourly being typical). The customer usually receives energy prices a day ahead, though forecasted prices may be available earlier. Hourly meters record loads, and the utility bills the customer monthly. Consumers receive notice of prices via fax, modem, e-mail, and more recently, the World Wide Web. Hourly prices typically fall below average for much of the time, shooting up to high levels (em perhaps a dollar or more per kilowatt-hour (em only for short periods, which may represent only 1 or 2 percent of all hours each year.

Most current RTP rates are designed so that if the customer does not deviate from "baseline" (i.e., pre-RTP) operation, he or she will see no change in total electricity cost. In other words, from the utility's perspective, the rate stays revenue-neutral if the customer (or class of customers) does not alter consumption patterns. Deviations from the baseline load profile will increase or decrease the bill, depending on the hourly price at the time they occur.

So why don't traditional TOU rates offer enough of a price advantage?

Experience with traditional TOU rates has demonstrated that customers see little incentive to shift or curtail demand at critical times if the swing in prices is too small. Likewise, the customer becomes less able to respond when high prices remain in effect for too many hours during the week. Fixed-demand charges based on the customer's monthly peak provide crude price signals at best; they can severely penalize short-term operation mistakes or equipment failures, whether these occur at critical times or not. Curiously, such demand charges persist in many real-time rates. Further time-differentiation in prices is needed for customers to effectively change their behavior.

Commercial/Industrial Uses: Some Trade-offs

To control mechanical systems in response to price, operators of energy-using facilities can take advantage of RTP to control mechanical systems in response to price. Common price-control techniques fall under several general categories:

s Storage. Shifting energy purchases from high-priced periods. Relying instead, for example, on central thermal storage for cooling, or (for industrial facilities) on storing intermediate product.

s Fuel Switching. Switching between an electric chiller or a chiller driven by a gas-fired engine, for example.

s On-site Generation. Using auxiliary power when (marginal) electricity prices exceed the operating cost of the generator.

Other response strategies may involve a trade-off between cost mitigation and service quality, especially when energy managers must decide which loads are noncritical, or choose among resources that offer different levels of service quality to meet the same need:

s Curtailment. Cutting back on noncritical uses (e.g., lighting) when prices exceed a given threshold.

s Distributed Storage. Precooling a building's thermal mass or frozen product, for example.

An energy manager might use some of these same approaches to minimize energy costs for a group of buildings or facilities (em perhaps by scheduling production in different plants at

various times, while weighing time-specific electricity costs at each location, as well as transportation costs. An oil pipeline, for example, could dispatch its pumping stations in different utility territories to minimize cost. A producer of industrial gases could plan production in the same manner.

However, the flexibility and innovation implied by these techniques comes at the expense of a demand for increased vigilance, knowledge, and computational skills on the part of the facility operator. Computer-assisted tools can help fill that gap, allowing operators to inventory loads and assess control strategies, as well as perform the necessary numerical computations. In all likelihood, as RTP proliferates, energy infrastructure equipment more and more will arrive equipped with intelligent control systems that can incorporate electric price changes in their operation.

The Residential Class: Enough Load to Matter?

Undoubtedly, some customers may receive little benefit from RTP, or find such pricing burdensome. They may even constitute a significant market. Other customers may prove willing to "play the market," but only if they are guaranteed some level of price certainty.

Energy service companies might well serve such markets by offering consumers a range of packaging options that enable them to choose the level of price volatility (or supply interruptibility) that best fits their needs. TOU rates will likely become more refined: In general, on-to offpeak price spreads will increase; the onpeak period will be defined more precisely; critical days for the utility will be specially priced.

Improved technology in telecommunications and metering is already expanding the market to which RTP can be offered cost-effectively. Home management

systems will soon include energy management under RTP along with other functions such as security, video (TV), voice (telephone), and data (Internet, electronic payment). To provide such services, a number of electric utilities (e.g., Pacific Gas & Electric, Public Service Electric & Gas, Florida Power, Entergy, CSW, Detroit Edison, Wisconsin Electric, American Electric Power, and UtiliCorp) have entered into joint ventures with telecommunications firms (e.g., TCI Cablevision, Microsoft, AT&T, IBM, BellSouth, Ameritech). Residential customers (or their energy service providers) may well take advantage of RTP within the next decade. t

Scott L. Englander is a senior analyst with Tabors Caramanis & Associates, an energy consulting firm with offices in Cambridge, MA, and Davis, CA. John E. Flory is principal-in-charge of TCA's Davis office. Leslie K. Norford is a principal of TCA and associate professor at the Massachusetts Institute of Technology. Richard Tabors is president and founder of TCA, and senior research engineer and senior lecturer at MIT. Tabors Caramanis & Associates has conducted real-time marginal pricing studies for more than 20 U.S. utilities, and has developed software to help customers take advantage of RTP.

Articles found on this page are available to Internet subscribers only. For more information about obtaining a username and password, please call our Customer Service Department at 1-800-368-5001.