Mixed Signals: Resource Adequacy and Capacity Markets

By Ron Moe, R. W. Beck, Inc.

New capacity markets offer to reduce uncertainty and provide an equitable, stable environment for generators

Why Capacity Markets Exist

The capacity markets are intended to help complete wholesale markets and provide total compensation to suppliers, to achieve the lowest cost consistent with adequate and reliable electric service. The capacity markets are also intended to support the development and continued operation of infrastructure needed to maintain reliability. Due to price caps in the energy market, all generation units that currently provide energy or reserves during times of shortage are being deprived of revenues, otherwise received in an uncapped market. Under the new capacity markets, all such units will have an opportunity to recover those lost revenues.

Like locational marginal pricing (LMP) for energy, the newer capacity markets similarly have a locational element that yields different prevailing capacity pricing in different sub-regions, resulting in higher capacity prices within import-constrained zones. This is intended to more properly value capacity by its location.

The new capacity markets also improve the prior capacity markets by replacing the highly volatile vertical demand curve with a downward-sloping demand curve. As a result, more predictability is introduced into capacity pricing, thus providing better incentives to encourage both new investment in supply and higher reliability of both new and existing generators.

The new capacity market demand curves are designed to yield capacity prices that, together with revenues provided by the energy and ancillary markets, will be adequate to attract new investment in supply. On average, and over the long term, the total of all market revenues received are intended to approximate, but not exceed, the full costs—including an appropriate risk-adjusted rate of return of the lowest-cost new supply entry, that is, a new combustion turbine peaking unit.

The First Generation of Capacity Markets

Initially, separate bid-based markets were established for energy, capacity, and ancillary services. The expectation was that market forces would induce appropriate levels of investment in new supply resources, and competition among suppliers would result in price efficiency to serve the regional demand.

The market's initial design was imperfect. During intervals when supply became tight, stakeholders wielding market power could drive the market clearing price for energy to excessive levels. As a result, administratively set price caps were put in place for the energy market in order to protect consumers. This form of price mitigation has removed, from the overall market, a crucial element of compensation needed by suppliers to recover their costs.

The initial capacity markets implemented were also problematic, consisting of a bid-based auction market. However, the design of the market resulted in a vertical demand curve that produced "bipolar" pricing. When the total regional installed capacity level was less than or equal to the amount of installed capacity needed each year to cover peak load plus required reserves to satisfy the reliability standard, the clearing price settled near the administrative deficiency charge which served as a price cap. Conversely, even slight levels of surplus supply resulted in capacity prices near zero. The capacity auction markets typically settled near zero until the occurrence of a single high price event.

Overall, it was found that the resulting price signals from the market alone were insufficient to attract new investment. Market revenues were also not sufficient to support the ongoing operation of some of the existing generation sources. In some regions, to assure the continued operation of various generators that are considered key to maintaining system-wide and/or localized reliability, it was necessary to enter into reliability must run (RMR) agreements with certain suppliers that allow these generators to operate "out-of-market" (that is, when the market clearing price is below their operating cost). The RMR agreements provide price supports to cover the respective shortfall in generator operating costs.

New England, New York, and PJM all had reserve margins that load-serving entities were required to satisfy, but since the settlement was on a short-term basis (daily or monthly), the resulting equilibrium prices were very volatile. The prices were too volatile to take into account in assessing the feasibility of a new plant, and especially in evaluating the risks associated in financing such a plant.

The Second Generation of Capacity Markets

The new capacity markets should provide longer-term price signals, allowing for the early years of a new financing to be based upon a more stable revenue stream and provide customers with the ability to ensure longer-term resource adequacy. Additionally, the new capacity markets should provide regulators with a more transparent mechanism to monitor resource adequacy in the region.

The main attributes of the new capacity markets that should lead to more stable prices include multiple year forward auctions, appropriate locational signals, implementation of a downward sloping demand curve, flexibility in self-scheduling, and demand response by the load-serving entities.

PJM's soon-to-be implemented reliability pricing model (RPM) uses a four-year forward auction to ensure adequate supply in the near to medium term. Additionally, incremental markets are held throughout the timeframe to true-up requirements, and load-serving entities are able to continually pursue bilateral contracts to meet needs outside of the capacity market (Figure 1).

1. PJM's reliability pricing model timeline. This chart demonstrates the long-term pricing mechanism, as well as the ability for load-serving entities to enter into bilateral agreements throughout the process. Source: PJM Interconnection

PJM also produces a variable resource requirement curve for the forward capacity auction year (Figure 2). The slope of the demand curve—from Point 1 to Point 2—is designed to be steep. The steeper-sloped portion of the demand curve is intended to signal the need for further capacity investment any time that declining installed capacity begins to approach the variable resource requirement, and thereby reduce the risk that capacity will fall below the resource requirement. Finally, PJM proposes to model six locational constraints, acknowledging the limited ability to import capacity in certain areas due to transmission, voltage, or stability limitations.

2. PJM filed variable resource requirement curve. This figure demonstrates the steepness of the curve as reserve margins tighten, sending the signal for new investments. Source: PJM Interconnection

The ISO-New England (ISO-NE) proposed locational installed capacity (LICAP) model was similar to the PJM model in that priced capacity was based on need within five geographic regions and faced a similar demand curve. However, there was no auction, and prices were set on an annual basis through a prescribed formula. Currently, at press time, the structure of the proposed LICAP model was in negotiation. One of the main changes proposed is the implementation of a three-year forward capacity auction. Under the proposed forward procurement model (FPM), ISO-NE will forecast needs three years in advance of the auction, and new and existing power, alternative energy, and demand-response assets can bid into the market.

Effects of Changes in Capacity Markets

The primary effect of the changes in capacity markets described is to reduce the uncertainty plant owners (and load-serving entities) face regarding future earnings from the markets. In particular, the utilization of administrative demand curves to set the capacity price will keep the price from oscillating between nearly zero in surplus periods and very high values in deficit periods.

The reduction in capacity price uncertainty will have several important effects. Capacity market revenue streams should be predictable enough that financial institutions would recognize them in assessing the viability of a proposed merchant plant. Over time, this may allow merchant plants, without power purchase agreements, to be financed. In particular, peaker merchant plants may become financeable as the revenues they earn become easier to predict.

In addition, the changes to these markets should result in less pronounced boom-bust cycles in the construction of power plants. The four-year ahead markets provide important information to market participants about the likely state of the markets, four years out. Developers and financial institutions should be able to use this information to make better-informed decisions about going forward on specific proposed plants, that is, canceling plans if the four-year ahead price is below Point 2 in Figure 2 and quickly developing plans for additional plants if the price is above Point 1.

There are also some unintended consequences of which to be aware. New capacity markets are still theoretical and have not been around long enough to determine if these send the right signals for new build. A mandate for reserve margins still exists, not allowing the consumer to pay for the reserve margin they select. Motivating demand response may be a better option than building excess peakers. Additionally, a disconnect still exists between the decision to build power plants versus the decision to upgrade or build transmission lines.

While locational capacity and energy markets indicate where additional resources are needed, without cost recovery on transmission lines, there is no correct market signal for when it is more economical to build a transmission line than a power plant.