Cost-Causation-Based Tariffs for Wind Ancillary Service Impacts

As the use of wind energy increases in the United States, there has been significant interest in assessing the integration cost of wind. With higher penetration, these costs have been shown to increase, but are generally modest relative to the price of wind energy. Although specific tariffs for wind generation for ancillary services are uncommon, we anticipate that balancing authorities (control areas) and other entities will move towards such tariffs. Tariffs for regulation and imbalance services should be costbased, recognize the relevant time scales that correspond with utility operational cycles, and properly allocate those costs to those entities that cause the balancing authority to incur the costs. In this paper we present methods for separating wind’s impact into regulation and load following (imbalance) time scales. We show that approximating these impacts with simpler methods can significantly distort cost causation, and even cause confusion between the relevant time scales. Correctly calculating and allocating cost impacts provides a market signal that can encourage economic efficiency, and avoids subsidies. We present results from NREL’s wind data collection program to illustrate some of the dangers of linearly scaling wind resource data from small wind plants to approximate the wind resource data from large wind plants. Errors in scaling can cause a significant over-estimate of wind impacts in the regulation and load following time frames and result in cross-subsidies in the resulting tariff. Finally, we provide a framework for developing regulation and imbalance tariffs, we outline methods to begin examining contingency reserve requirements for wind plants, we provide guidance on the important characteristics to consider, and we provide hypothetical cases that the tariff can be tested against to see if the results are what are desired. Introduction In this paper we discuss general principles of tariffs, with the goal of pointing out how tariffs can address pricing for services based on cost-causation. Because the use of wind energy has expanded significantly over the past few years, some Balancing Authorities (BA) are considering how to develop their own tariffs for wind plant integration. If care is not used to develop tariffs that address the specific costs for regulation and imbalance, for example, then these tariffs may either over-charge or under-charge wind for its impacts. Good Tariff Design Should Be Based On Operating Principles Modern power systems are complex networks of machines. To operate effectively, system operation personnel have developed various procedures that assure the reliable operation of the power system. The fundamental task is to maintain balance between system loads and generation. Although this is simple in concept, the practice of power system operation is quite complicated. Electric loads vary over all time scales, from a few cycles up through longer periods such as days, weeks, or months. The task of maintaining system balance is divided according to the time scale. Figure 1 is a typical representation of the critical time scales. From the operational perspective, the unit commitment time scale, which can range from several hours to a few days, represents the time required to ensure adequate slow-start generating units are available to provide power. Performing the commitment function correctly requires knowledge of future loads, along with other relevant variables like the availability of other generating plants. The Regulation Time Frame Once the unit commitment process is completed, the operator has the ability to control this online generation, subject to the various physical and electrical constraints of the generators and the balance of system. Some of this control involves automatic generator control (AGC) in response to the changing load-resource mix, performed automatically by computer. This process typically occurs over several seconds to several minutes, depending on the system. AGC units respond in the regulation time scale, and this ancillary service is typically quite expensive. In the minute-to-minute regulation time frame loads move in uncorrelated patterns, and the generators that perform the regulation service do not need to match each individual load change—only the aggregate system balance needs to be chased by the regulating units. The Load Following Time Frame The next time frame is load following. Changes in load in this time frame are generally correlated across electric customers, although there is typically a regulation “wiggle” that is superimposed on the correlated movements. The correlation typically occurs during the morning load pickup, when many customers increase their electrical demand, and in the evening when electrical demand falls off. The system operator must follow load with generation that is either already online, or generation that can be brought online quickly. In either case, the load following unit(s) must have the capability of changing output in response to operator instruction, and must be able to follow the instruction fairly closely though they do not necessarily have to be on AGC. Time (hour of day) 0 4 8 12 16 20 24 S ys te m L oa d (M W )