Emerging capacity auctions offer limited but valuable risk-management tools for asset owners.
Mark Griffith is senior vice president at Ventyx Advisors and head of its asset valuation practice. He can be reached at email@example.com.
NOTE: Ventyx Advisors is part of the Ventyx Energy Group formed by the acquisition and merger of Global Energy Decisions and New Energy Associates by Ventyx.
For much of the history of the electric power industry, power generation plants were built by utilities, paid for by ratepayers and incorporated into the utility ratebase as a result of approvals from state regulators.
The past decade has witnessed the development of a large fleet of unregulated, merchant-power generation resources in the United States (see Figure 1). More than one-third of all generation no longer is owned under the traditional vertically integrated electric utility model,1 and these unregulated assets face various markets and a wide range of risks. The revenue expectations for these assets often are divided into two categories: energy and capacity. These concepts frequently are used without a precise definition.
Unlike most major industries, the electric power industry has historically made a clear distinction between the value of energy produced from a power plant and the value of “capacity,” which is the ability of a power plant to generate electric energy.2 While all industries have both units of production and the capability to produce output, when discussing value one typically looks at these concepts separately. Simply stated, one either can purchase a factory, or purchase the factory’s output. When one purchases the factory’s output, there implicitly is some value of the factory embedded in the price, but it is rare to pay a separate price for the factory’s productive capability.
In contrast, the U.S. electric power industry developed largely as a government-regulated monopoly. One of the legacies of this regulation is that energy, when valued at the wholesale level, was often valued at its variable cost of production, with no allowance for the cost of building the power plant. In order to compensate power-plant owners for their investment in generating assets, a separate “demand” or “capacity” charge was paid by customers.
Fast forward to today’s partially deregulated electric power markets. Wholesale electric energy often is traded in various central markets, as well as among individuals in bilateral transactions. Wholesale electric energy prices largely are deregulated, and clearly, over the past decade, market participants have become adept at routinely charging much more than their variable production costs. This “rent extraction,” as economists commonly call it, can take various forms, and while the mechanism for achieving it can be complicated, the evidence is quite clear that today’s wholesale electricity prices typically are higher than the variable costs of most or even all suppliers.
Even though energy prices are much higher than production costs in this competitive marketplace, the prices often are not high enough to induce investors to build new power plants. As a result, various mechanisms have evolved since the early 1990s to induce new construction (and prevent retirement of old power plants) by providing additional revenue to supplement the energy revenue. This supplemental revenue takes various forms, but typically is called capacity revenue. In practice, this term is used as a name for several different concepts. But whatever the name used, capacity auctions are becoming an important feature of the U.S. power-market landscape.
Conceptually, there typically are two types of payments for capacity from a power plant. Essential to differentiating these two types of payments is indentifying who receives the benefit of the power plant’s gross margin. 3
In the first type, the capacity payment gives the buyer the right to the energy produced by the power plant at the energy’s variable cost of production. Purchasing the capacity and energy as a bundle like this sometimes is called a power purchase, as evidenced in the commonly used industry term “purchased-power agreement” (PPA). Such an arrangement is analogous to a call option on the energy price (or more precisely a call option on the spread between the plant’s fuel price plus VOM and the energy price), and is even sometimes called a “physical option.” A variant on this concept is a “tolling” payment, in which the buyer not only owns the energy output, but has the obligation to provide the power plant’s fuel. In either case, the gross margin between the power plant’s variable costs and the market price of energy is captured by the capacity buyer.
The opposite approach to defining capacity is a form in which the buyer has the right to the energy output of the power plant, but does not control the price at which it receives the energy output. Depending on the specific terms of the transaction, the buyer typically will have the right to call on the capacity of the plant at any time, but will pay the higher of variable production costs or the market price of the energy,4 so the positive difference between variable cost and market price is retained by the power-plant owner. This type of capacity payment is sometimes referred to as the “reliability” or “regulatory” value of the capacity or, more colorfully, the “naked capacity” value of the power plant. The retention of the gross margin by the power-plant owner is typical of capacity transactions seen today in the administratively managed capacity markets in PJM, New York and New England. Just to add to the confusion, some markets also make “reliability must-run” (RMR) payments for capacity to ensure reliability, but such arrangements can assign the gross margin value to either the capacity buyer or the power-plant owner. RMR payments often are made for short periods of time (such as one year) and are subject to significant changes due to the development of new power plants or transmission lines.
Capacity payments made under PPA or tolling arrangements are higher than capacity payments made for the reliability capacity of the plant. This is a natural result of a PPA or tolling agreement giving the purchaser an opportunity to buy energy at a price lower than the market value of the energy. In the reliability capacity arrangement, the purchaser never receives energy at a price lower than the current value of the energy. The capacity payment made under a PPA/tolling agreement is a function of the efficiency of the power plant. Buyers of PPA/tolling capacity must be willing to make a relatively higher capacity payment for units with good heat rates as opposed to units with poor heat rates. The good heat rate results in more hours when the plant is “in the money,” and hence creates a higher level of operating profit.
A plant owner that is considering selling its capacity can choose to sell either PPA/tolling capacity or reliability capacity. If the product sold is reliability capacity, then the owner of the plant will receive lower capacity revenue, but will retain the profits that result from any energy produced at a cost lower than the current energy price.
Administrative Capacity Markets
In most of the United States, revenue from energy and ancillary services usually are insufficient to cover the production costs, fixed O&M, and capital charges for new generation. For new generation to enter the market, generators rely on capacity payments to supplement their energy and ancillary- services revenue streams.
Increasingly, capacity markets are becoming an important market driver, and there is optimism by many market participants and regulators that these markets will provide sufficient incentives to induce entry of new capacity to secure future reliability. The capacity markets in PJM, NYISO, and ISO-NE still are evolving, and further modifications can be expected. However, they do have common elements (see table, “Northeastern U.S. Capacity Markets”).
In each of these three markets, auctions are held under a prescribed set of rules to arrive at capacity prices (often referred to as installed capacity, or “ICAP,” prices). The auction rules and the period of time to which the capacity payments pertain vary widely by market. Also, the Northeast capacity markets often can be circumvented by either: 1) bilateral contracting between power plant owners and load serving entities (LSEs); or 2) LSE self-generation. A common problem in today’s ICAP markets is that the period of payment is much shorter than the life of a new power plant, creating a significant source of financial risk for the power-plant owner.
The distribution of loads and resources isn’t uniform within each of these markets, and due to transmission limitations, there are geographically varying values for capacity. The three ICAP markets generally acknowledge this and are attempting to incorporate geographic issues into the capacity markets.
All three capacity markets use a mechanism to set a ceiling on capacity prices in an attempt to limit the potential for market manipulation. In each case, the ceiling is a function of the cost of new entry (CONE), which has proven to be a much-debated value. The CONE is set administratively through a contentious administrative process and is subject to various political and economic pressures. The mechanism by which the CONE is translated into the ceiling price varies widely by market, but a common theme is that the CONE gets reduced by some allowance for profits from the energy market.
Each of these capacity markets is much more complicated than a quick summary can explain, and are in a near-constant state of flux. Also, while the process may be wrapped in science, there are ample opportunities to manipulate the outcome.
In energy-only markets, that is, markets without administrative capacity mechanisms, owners of unregulated power plants usually have two choices, namely,: 1) sell only energy; or 2) sell some combination of energy and capacity. Rarely would a new power-plant owner receive enough revenue in the “selling energy only” mode to produce an adequate return on investment, so that owner likely would make some capacity sales. This can take the form of either a PPA/tolling sale, which provides the energy to the buyer at cost, or a sale of reliability or regulatory capacity, which allows the power plant owner to realize any positive difference between production costs and the market price of energy.
In many of these energy only markets, there are state-level reliability mandates (sometimes referred to as “resource adequacy” requirements) that compel all LSEs to demonstrate an ability to serve their customers under peak demand conditions. 5 That is, there is a reliability requirement. In some markets, such requirements do not yet exist, but are likely to exist in response to the national mandatory reliability requirements in the Energy Policy Act of 2005. This typically can be met by self-owned generation or contracted generation. Such reliability requirements make it reasonable to assume that at least some power-plant owners will be able to receive contractual capacity payments in future markets.
This can be characterized as a “potential capacity” revenue source, which signifies that although there are logical reasons that some power plants will receive such revenue, there is no assurance that all qualified power plants will receive it, or receive it every year. Although the revenue stream is uncertain, forecasting potential capacity is a relatively simple process. The most economic source of new capacity typically is one with a relatively low capital cost, which is usually a natural gas-fired technology such as a combustion turbine (CT) or a combined-cycle (CC) power plant. Potential future energy market gross margins for these two technologies can be forecasted based on a dispatch analysis using an energy and fuel price forecast. The shortfall between each technologies’ levelized revenue requirements6 and the forecasted energy gross margin is in essence a forecast of the capacity revenue each technology would require to enter the market. The smaller of the two shortfalls represents the cheapest method for an LSE to acquire reliability capacity, and hence becomes a forecast of potential capacity prices.
The unregulated merchant generator faces many risks to its expected revenues. Risk related to energy prices is relatively obvious; risk related to capacity prices is less so.
A common assumption is that capacity revenues are in some way less risky than energy revenues. In reality, capacity revenues can be even riskier than energy revenues, depending on circumstances. The specifics of these risks differ according to their causes. Namely:
• Energy: Energy revenue is subject to volatility risks for both the price and volume of production. Price volatility can be measured based on historical observation. Volumetric risks take into account the overall market demand for electric energy and the relative competitive position of each power plant. Energy revenue ultimately translates into a gross margin (energy revenue minus production cost), and at that point the volatility on underlying fuel prices also becomes a major risk factor.
• Potential Capacity: There are no ICAP markets in the United States outside the PJM, ISO-NE and NY ISO markets. While some regions’ administrators are discussing potential future capacity markets, such discussions inherently are too speculative to rely on at this point. In these non-ICAP markets, payments for capacity typically are realized either through bilateral contracts or Reliability Must Run (RMR) payments from a central market authority, e.g., the California ISO. In the case of bilateral contracting, there is significant competitive risk. This can arise from:
1) Supply and Demand: When a market’s installed capacity exceeds the installed capacity requirement (peak demand plus reserve margin) by even a small amount, the value of capacity can fall to very low levels. When power-plant owners have to compete to serve a market smaller than the available supply, competition quickly drives down the PPA value of capacity from full replacement cost to its option value in the energy market. In this scenario the reliability value of capacity is near zero.
2) “Falling off the cliff:” Once all reliability needs are met for a specific period of demand, there can be a world of winners and losers. The winners have secured their PPAs, while the losers (those who failed to secure contracts) have essentially a zero reliability capacity value until the next round of capacity purchasing.
• Administrative Capacity: Admittedly, a market’s willingness to pay ICAP to power plants demonstrates some revenue certainty compared to markets without ICAP. And the demand-curve constructs in PJM and NYISO were designed to mitigate the sudden value drop-off seen when capacity exceeds installed-capacity requirements. Still, those U.S. regions with existing ICAP markets are experiencing many changes in their administration. And these markets set ICAP prices for a relatively short time frame in comparison to the expected life of a new asset. So even though these are existing markets, future ICAP revenue streams are speculative as to the details of how they will be administered and how prices will be set, all of which creates risk for investors.
So, contrary to much conventional wisdom, capacity revenues are not the great risk reducers some would like to believe they are, and in non-ICAP markets capacity revenues even are less certain than the energy revenues. The counterpoint to these risks is that, in the long run, there must be some mechanism for allowing new power-plant investors to have a reasonable expectation of receiving their investment costs and making an acceptable return on their investments. This can be done with an ICAP market (of which there can be many different designs), reliance on bilateral contracts, enforcement of resource adequacy standards, and so forth.
While the market therefore has an incentive to keep the lights on, any one asset or asset portfolio will be at risk as to just how the capacity payment mechanisms translate into revenue received.
1. About 351,000 MW out of a total of 918,000 MW, based on U.S. DOE summer 2006 capacity estimates (Energy Information Administration, Electric Power Annual 2006, Table 2.3.
2. Energy production is measured in MWh and typically priced as $/MWh; capacity is measured in MW and typically priced in $/kW-month or $/kW-year.
3. For this article, gross margin is the difference between the market price of energy and the power plant’s variable production costs, including fuel.
4. Typically the price of energy would be tied to a mutually agreed-upon price index.
5. For example, in California, utilities are required to contract for an 18-percent reserve margin.
6. Levelized revenue requirements typically include return of, and on, capital and can include items such as insurance, real estate taxes and fixed O&M.