Utilities that are short on capacity and operate in a stable regulatory environment may be able to extract value from interruptible rates.
Regulated utilities have struggled for a long time to determine the value of an interruptible rate. As deregulation developed, many utilities moved away from interruptible rates that are disconnected from energy markets and started offering tariffs with dynamic pricing tied to the markets. Demand bidding and real-time pricing are two examples of the latter approach. The low prices in today's wholesale electric markets have resulted in a reduction in the value of the retail market-based rates for both the utility and the customer.
Utilities that are short on capacity and operate in a stable regulatory environment may still be able to extract some value from interruptible rates. The value of the interruptible rate hinges on delaying the need for capacity at a cost that is less than the cost of adding and operating the supply-side resource. To achieve this goal, the interruptible rate tariff must establish prices for both capacity and energy that reflect market prices.
What We Know So Far
Dr. Hans E. Nissel, in a June 1983 Public Utilities Fortnightly article, argued that interruptible customers do not contribute to a utility's peak demand because they represent no capacity cost. He proposed that the only cost for a utility to provide nonfirm service is energy and customer cost. In the end, he concluded that it was reasonable for interruptible customers to contribute to the capacity cost of the power pool construction and operation for firm service, since it makes nonfirm service possible.
The most basic definition of an interruptible rate is the offering of nonfirm power. Participating customers (mainly large industrial customers) accept decreased reliability in exchange for a reduced charge. The degree to which the utility can interrupt the customer's load impacts the value of the rate to the utility. In addition, utilities must consider how interruptible programs affect the costs of their operations. The provision of nonfirm service at a discount has value to a utility provided that the cost of the nonfirm service is less than cost of serving the load. In other words, the discount for the nonfirm power must be less than the cost of the resources to serve the customers with firm power.
Traditional interruptible rates represent a form of load management or demand response. One form of an interruptible service rate started with the large industrial firm rate and then discounted that rate by about 50 percent to compensate interruptible customers for the inconvenience caused by interruptions. AmerenUE's cancelled interruptible (10M) rate had this structure, as well as restrictions on interruptions and a requirement that the company "endeavor to obtain temporary power (capacity only or both capacity and energy) to meet" its requirements before calling an interruption.
The actual value of the interruptible rate depends on several factors. The biggest determinant of an interruptible rate's value is its intended use. Some examples of an interruptible rate's intended use are usage during times of system reliability issues only, system peak, sales to markets, and high native peak load. Certainly the rate's intended use will significantly affect the utility's usage of the interruptible rate, therefore impacting its value to the utility.
A second factor is the utility's capacity position. Is the utility long or short in capacity? A utility's capacity position changes from year to year. In one year it may be short capacity, then, if a plant comes on line, it could be long the following year. The impact of the utility's capacity position will vary depending on the interruptible rate's objective. For example, if the objective of the rate is only to be used for system reliability issues and the utility is long on capacity, than the interruptible rate has reduced value. However, if the interruptible rate can be utilized to take advantage of favorable market conditions and is priced appropriately, it still has financial value.
A third factor in an interruptible rate's value is the restrictions placed on the utility in calling interruptions. The key parameters that influence restrictions are as follows:
- Frequency of interruptions
- Duration of interruptions
- Total number of interruptions
- Notice period before an interruption
- Assurances of interruption
The value of curtailable and interruptible programs to the utility is the result of a reduced cost to serve load (or revenues from selling the capacity). If a utility can substitute nonfirm load for firm load, the value of the nonfirm load is the marginal cost of the foregone resource additions. As noted above, not all degrees of nonfirm service provide equal value. A kilowatt of curtailable or interruptible load must therefore be evaluated in the context of the restrictions placed on its availability and utilization.
The three factors (intended use, capacity position, and restrictions) combine to determine the value of the interruptible rate for the utility. The full potential of an interruptible rate will be realized when the three factors are optimized to work together (see Figure 1). If the factors are not complementary, the value will be reduced. For example, an interruptible rate that allows unlimited curtailments (restrictions) but only for system reliability reasons (intended use) will have minimum value to a utility. On the other hand, an interruptible rate that allows 400 hours of interruptions (restrictions) for economic reasons (intended use) at a time when the utility is short in capacity (capacity position) will have significant value to the utility.
Each of the three factors mentioned above should be considered when valuing the capacity and energy value of an interruptible rate. Since an interruptible rate is a demand-side option and customer demand typically is met on the supply-side (i.e., generating plants), an equivalency needs to be established between the supply and demand sides. To establish the equivalency, resources with similar characteristics are compared. For example, an interruptible rate is a demand-side option for reducing peak demand throughout the year. A comparable supply-side option is a peaking plant. If all operating characteristics (see factors above) are equal, the two options should have an equivalent value to the utility.
A peaking, supply-side option is valued on the capacity (fixed) and energy (variable) costs based on its potential utilization. Similarly, while valuing the capacity (or demand) portion of the rate, the "equivalent" capacity and financial value is determined. The "equivalent" capacity is derived from the demand-side option's reliability and economic dispatch contribution. The majority of a supply-side's variable costs are energy related. The demand-side's energy value can be calculated using the comparable supply-side's energy cost.
Reliability and Economic Dispatch Contribution
As a utility proceeds through its annual resource planning, it establishes a capacity position for the foreseeable future. A utility's capacity position includes forecasted peak load, supply-side capabilities, and demand-side capabilities. AmerenUE reports capacity capabilities according to Mid-America Interconnected Network (MAIN) Guidelines (3B and 4) using the MAIN summer assessment, winter assessment, and EIA (Energy Information Administration) form 411. In addition, MAIN compliance staff perform capacity audits beginning in the early spring of each year.
Resources listed on the capacity assessments have varying value to the utility. This point is worth repeating: All capacity, supply or demand, does not have the same financial value to a utility. The value of any capacity is dependent upon the operating characteristics of that particular resource. One megawatt of coal generation capability is not equal to one megawatt of gas generation capability. Why? The capacity position is the calculation of the amount of reserve megawatts at time of system peak and does not provide an indication of the capacity, or load relief, which is available throughout the entire year to meet customer requirements.
The ideal approach for developing capacity cost-based incentives requires the following steps:
- Determine reliability contribution of nonfirm service.
- Adjust the resource plan according to the difference in reliability from the target level.
- Determine the value of the capacity (or contract) deferred or eliminated.
In a 1993 article, Peter Jackson offered a conceptual basis for thinking about demand-side reliability within the broader framework of the planning process. In the article, he reviewed two methodologies utilities used in Illinois for quantifying the value of DSM reliability. The two methods he compared are the Reserve Margin Adder Method (RMA) and Capacity Equivalence Method (CE). RMA uses a simple and straightforward approach of adding a reserve margin credit to the level of capacity reduction at the point of use. For example, a utility with a 15 percent planning reserve margin would have a 1.15 MW reduction in demand for each 1 MW reduction that is attributed to a DSM resource. CE is less straightforward but reflects the fact that availability of a DSM resource may vary throughout the year. The CE method is a superior methodology for two basic reasons: 1) It reflects a demand-side resources contribution for the entire year; and 2) It establishes an accurate relationship between demand-side and supply-side alternatives.
To determine the reliability contribution of a demand-side resource using CE, a relationship is calculated to a supply-side resource. CE is the true capacity value of a resource (DSM, DR, wind, hydro, etc.) compared to a supply-side option. To determine the capacity equivalence, AmerenUE uses the same calculation MAIN uses to determine reserve margin: Loss-Of-Load Probability (LOLP).
Table 1 is the result of performing a LOLP sensitivity analysis that evaluates several nonfirm service options. The first column of the table indicates the total number of days a nonfirm customer may be interrupted during the year. In the next three columns, the maximum hours per interruption vary. In the final three columns, the notice period varies from one to four hours. Any departure from the ideal nonfirm service derates the system reliability contribution and lowers the value to the utility. For example, assuming a utility is offering nonfirm service with maximum interruptions of 40 days, 4 hours per interruption, and 1 hour notice, the capacity equivalence is .5290 (.5577 x .9486 = .5290).
This approach illustrates that the most desirable nonfirm service option from the utility viewpoint is an interruptible rate with no notice, no restrictions upon hours of interruption per incident, and ability to interrupt frequently during peaking periods. Under these conditions, 1 kW of nonfirm saves 1 kW of capacity, and the value is the full avoided capacity cost (capital or contractual cost) attributed to reliability.
Economic dispatch is the second factor considered in deriving the equivalent capacity. If the number of hours that a supply-side option can be called is limited, the economic dispatch availability could be limited. It is unlikely that the load dispatchers will be able to determine the optimum usage of an interruptible rate that limits the number of hours of nonfirm service severely. For example, if an interruptible rate is limited to 100 hours of nonfirm service and the utility faces an unusually warm summer, it could exhaust its hours of interruptions by calling too many interruptions early. Another situation under the same scenario is that the utility holds hours in reserve in case things get worse, with the result that misses opportunities. One method of determining the equivalent values for various economic dispatch scenarios is to perform LOLP sensitivity analysis based on the number of allowable interruptible hours and potentially missed opportunities.
The Financial Value of Capacity
Utilities face markets that recently have gone through a boom-and-bust cycle. The nature of these markets makes long-term price commitments risky. In addition, the utility's capacity position changes yearly, which contributes to the dynamic value of capacity. Even with these constraints, the nature of an interruptible tariff requires a utility to establish a standing price for nonfirm power.
The avoided capacity cost is the change in a utility's capacity cost at the generation level due to an increment of capacity supplied by a qualifying resource. The comparable resource for nonfirm power is peaking capacity. EPRI's 2001 Demand Trading Toolkit notes that "the primary indicator of the value can be assessed through displaced agreements into which the energy company would have entered." In AmerenUE's situation, contracts or physical plants for supplying peaking capacity would be approximately "equivalent." The interruptible rate must be at least as reliable as the power from the utility's resource. If it is not, the "equivalence" factor discussed above should be used for discounting.
Below are several supply-side, peaking options that can be used as financial benchmarks for an interruptible rate.
- Forward Contract for Regulatory Capacity
- Description: A contract that meets the requirements for MAIN-accredited capacity usually purchased for peaking season only. Any energy purchased under this contract is at market rates.
- Period: Monthly or seasonal
- Price: 15,500 MW-season or $3.88 KW-mo
- Long-term Tolling Contract
- Description: A contract similar to the lease of a plant (or portion of a plant's capacity) in which the purchaser supplies the fuel and receives the energy at cost. The seller continues to operate the plant.
- Period: Multi-year contract (, five years)
- Price: $60,000 MW-year or $5.00 KW-mo
- Overnight Cost of Building or Purchasing a Plant
- Description: Utility builds or purchases a built plant.
- Period: Life of plant
- Price: $500-$350 kW or $6.25-$4.38 kW-mo
From the list above, the value of peaking capacity is in the range of $6.25-$3.88 kW-mo. These capacity values are subject to changes in the market. Also, remember that the utility's need for capacity (capacity position) changes every year. Therefore, the incremental value to the utility of peaking capacity will change every year. Due to both of these issues, an interruptible rate needs to allow re-evaluation of the capacity value periodically.
The Energy Value
The majority of a peaking plant's energy cost is fuel. A utility has several options in structuring this portion of an interruptible tariff. The energy cost could be calculated on a yearly, seasonal, monthly, or daily interval. The longer the interval, the more risk the utility will assume. If the energy costs are based on a yearly average of the forward gas price curve and actual gas prices swing low during the interruptible season, the utility would be required to pay above- market energy cost to the interruptible customers or avoid interruptions. The reverse situation could occur and the utility would receive "above market" value by calling an interruption. To minimize this risk, the energy portion of the tariff should be based on the day-ahead price.
If the utility refunds the full energy cost to the customer, the interruption would end up costing the utility more than the cost of operating a CTG. For example (scenario 1a in Figure 2), if a utility receives $20/MWh from the customer for energy customer usage, fuel prices are $40/MWh, and incremental system costs are $40/MWh, in an interruption, the utility would lose the $20/MWh from the customer and pay the customer $40/MWh for curtailing. The result is that the utility's costs are -$40/MWh. If the utility decides not to curtail (scenario 1a), it receives the $20/MWh from the customer and pays $40/MWh for fuel, with the result of -$20/MWh. Under these conditions, the utility would be better off not curtailing the customer's load.
The goal of valuing the interruptible rate is to make demand-side and supply-side options equivalent. As shown above, any energy credit based on fuel cost needs to be discounted to maintain equivalency. Applying this principle and using the conditions of the previous example, in an interruption the utility would lose the $20/MWh from the customer and pay the customer $20/MWh for curtailing. The result is that the utility's costs are -$20/MWh. If the utility does not interrupt, it receives the $20/MWh from the customer and pays $40/MWh for fuel with the result of -$20/MWh. The utility would be neutral to interrupting or running the CTG. Therefore, the options are equivalent in energy costs.
Value of an Interruptible Rate
The first step in the proposed method of valuing an interruptible rate is to establish the capacity discount. Fundamentally, LOLP analysis can be run for various sensitivities mentioned earlier. One difficulty is that the parameters in the calculation are constantly changing. An example that has been mentioned several times is the utility's capacity position. When a utility is adding supply-side resources, such as peaking plants, it can be short one year and long the next. LOLP analysis for each year will result in different discount values. Changing the value of the capacity equivalence each year would not be acceptable to most customers. When purchasing nonfirm power, customer's need to incorporate the curtailments into their business plans. They may even purchase equipment to allow them to curtail. To evaluate these investments, a customer needs a certain amount of rate stability. General discount values should be established using analysis and experience to provide the consistency needed by the utility and customer.
Before deriving the capacity equivalence values, the operating guidelines for the interruptible tariff need to be outlined. The assumptions used in creating the "Capacity Equivalence Values" table are as follows:
- Tariff has some kind of LD or buy-through
- Customers agree to long-term agreements (i.e., 3 years into the future)
- No limitations on reasons for interrupting customers
- No limitation on frequency of interruptions
Other assumptions will result in different discount values. For example, a rate that limits the interruptions to system reliability reasons only will have smaller capacity equivalence value. The basis for this method of interruptible rate valuation is equivalence-finding the balance in value of the supply-side and demand-side resource.
AmerenUE's Interruptible Rate: A History
In 1999, AmerenUE filed a stipulation and agreement (Case No. EO-96-15) that provided for the elimination of AmerenUE's Interruptible Power Rate (10M), among other things. It also provided that AmerenUE would implement a new tariff, the Voluntary Curtailment Rider. In 2000, a group of AmerenUE's industrial customers filed a pleading requesting that the commission open a case to investigate the establishment of an additional alternative rate option for AmerenUE's interruptible customers. The commission staff and AmerenUE filed pleadings opposing the application. The commission denied the industrial customers' pleading. At the time the rate was terminated, AmerenUE had four Missouri customers utilizing the rate. The four customers had a total nonfirm load of 47MW.
AmerenUE's and the commission staff's main objection to the 10M interruptible rate was the cost. Savings for customers on 10M amounted to $5/Kw-mo. At the time of the case, the average interruptible credit paid by other utilities in Missouri was $2.01/Kw-mo. By offering the Voluntary Curtailment Rider, AmerenUE was able to maintain much of the reliability benefit realized from the 10M rate.
The concept of the Voluntary Curtailment Rider is to provide customers the means to reduce their energy costs by bidding their ability to change their electricity demand characteristics. The generic name for this type of offering is demand bidding. The demand bidding approach offers customers a proactive role in energy markets and provides a greater ability to affect the price of electricity. Customers are offered price signals to reduce demand in return for a share of the benefits derived from that reduction. The result is that customers have the ability to trade their demand response capability and still have flexibility.
Another market based approach is real-time pricing (RTP). In May of 2000, a Retail Services Report article noted that "NERC suggest many users may be willing to absorb more price uncertainty, and even more supply uncertainty, in exchange for lower overall rates." A two-part RTP rate allows a portion of customers' load to be protected from price volatility while offering customers the opportunity to save money. In a two-part RTP rate, customers are refunded/charged at markets prices for deviations to their baseline.
A drawback to the market approach is that depressed energy markets result in fewer opportunities for reducing demand while a regulated utility may be adding supply-side capacity to meet its resource needs. If designed and priced appropriately, an interruptible rate could reduce the amount of new supply-side resources needed and move interruptible resources from one of the last resources in the dispatch stack to a fully competitive alternative to supply. Along with demand bidding and real-time pricing, the proposed method has a "pay for performance" component. -M.W.
Articles found on this page are available to subscribers only. For more information about obtaining a username and password, please call our Customer Service Department at 1-800-368-5001.