Mending Our Broken Capacity Markets


The ability to provide reliable capacity is becoming both riskier and more costly to society and investors alike.

Fortnightly Magazine - June 2006

The year was 1902 and Samuel Insull, nominally the father of the electric utility industry, faced a dilemma. The largest generators at his flagship utility, then called Chicago Edison, used 4,000-kW engines. His engineers, led by Fred Sargent (who later founded Sargent & Lundy), thoughtfully explained to Insull that if the reciprocating engines were made any larger, the stress caused by the push-pull action of the recips quickly would destroy the machinery.

Insull suggested they build a new plant with huge steam-electric rotating turbines. Sargent claimed that using a turbine for large steam applications was impossible. Undeterred, Insull went to Charles Coffin, president of General Electric (GE), asking him to build and guarantee a 5,000-kW steam turbine. After several refusals, GE finally agreed to attempt the unthinkable—build the world’s largest steam generating plant without conventional reciprocating technology, but with untried turbine technology. GE made Insull’s utility accept a huge part of the cost and performance risk of the turbine.

Seventeen months later, in September of 1903, the Fisk Street Station, employing this huge new piece of experimental technology, was ready for dedication. At the initial firing ceremony, the engines were started and began to shake violently. Visiting officials were led off to a safe distance and the engines were turned off by Fred Sargent. The trouble was quickly identified, but as Sargent prepared to restart the engines, he noticed that Sam Insull was still on the turbine deck by his side.

“This is a dangerous business. You need to leave,” pleaded Sargent.

“Then why don’t you leave?” asked Insull.

“Look, Mr. Insull, it’s my job to stay here. I have to. But you don’t. Don’t you understand this damned thing might blow up?”

Insull turned toward the new turbine. “Well, if it blows up, I blow up with it anyway. I’ll stay.”

At least on that day, Sam Insull did not blow up. This historical vignette takes place in an era before utilities were assigned the responsibility to provide reliable service to designated geographies by regulating agencies and, in return, were granted monopoly rights to provide central-station service in their assigned jurisdictional territories. Insull’s decision engendered huge risk to himself and his investors—not simply technological risk, but much more substantively, financial risk. The Chicago Edison Co. sold to customers who readily could be picked off by any of a handful of competing central station and wires owners in upstate Illinois.

The struggling utility also was in head-to-head competition with GE and other manufacturers of power stations who were targeting Insull’s core industrial markets for sales of on-site alternating and direct-current machinery. In this environment of commercial uncertainty, Insull’s financing choices were limited and expensive, resulting in a very high cost of capital and the demand by investors for a very accelerated repayment of the initial, high cost of project construction.

Insull’s gamble paid off in this instance, but at high cost to the firm and, as a result, to customers. The resulting high cost of power fully reflected the economic exposure of this risky investment.

Needed: Clearly Defined Responsibilities

Our industrial ancestors a century ago perhaps can be forgiven for the lack of clearly defined load-serving responsibilities and a well-defined marketplace in which capacity investments could be made with relative revenue certainty and thus relatively low cost to society. After all, back then, electricity wasn’t perceived as a societal need but as a luxury. For the better part of the past century, industrial and regulatory evolutions have provided a sound platform upon which the commercial act of providing reliable power is no longer a “you bet your business” gamble. Assigning clear and unambiguous responsibility for reliable grid service and, in return, providing reasonable assurance of a fair, authorized return on investment—classically know as the “regulatory bargain”—worked well to meet the huge growth of the American grid. Supplemented by the Public Utilities Regulatory Policy Act of 1978 and the Energy Policy Act of 1992, the ability of new entrants to compete effectively with established utilities on the basis of capital and operating efficiency in the generation space, still under the watchful eye of regulated, clearly defined utility planning responsibilities, enhanced the options available for ensuring reliable, low-cost capacity.


Over the past decade, the American grid has been devolving into an environment where the ability to provide reliable capacity is becoming both riskier and more costly to society and investors alike.

The reason for this devolution? Poor designs of capacity markets, or, in some cases, lack of any design at all. The three-fold objective of capacity markets should be to:

1. Ensure reliability of the electric grid;

2. Provide this reliability at lowest cost to society; and

3. Supply this reliable capacity with a mix of resources that, in aggregate, are efficient, environmentally acceptable, low-cost producers of energy.

In several regions in this country, we have lost sight of the importance of these three objectives, and have allowed a combination of embedded constituencies and brilliant economic minds to design market structures that do not necessarily portend a healthy future for the industry.

The reason for this negative outlook is that the classically defined responsibilities for providing adequate capacity supplies have been replaced with either overly complex or unduly simplistic market-pricing signals designed to motivate appropriate capital-deployment decisions. This has replaced singular responsibility for taking thoughtful action, under the watchful eye of state and federal economic regulation, with the proverbial “invisible hand of the market.”

The market’s invisible hand only can operate effectively if it is wired into a well-functioning market nervous system and brain. The organized markets of North America have manifested suboptimal structures for motivating efficient capacity decisions by attempting to guide long-term capital decisions via much shorter duration price signals. In the energy market, things work rather well because pricing signals are provided in-time units—day-ahead and hour-ahead—that are consistent with unit-dispatch decision making. The energy market gives us signals in the same temporal space as our decisions are made in. What the grid needs is a capacity market that comes close to producing the same cost-effective actions in capacity deployment as it does in energy dispatch.

Bad Behavior

Capacity markets today actually motivate actions antithetical to grid reliability and economy. Markets are intended to govern economic behavior. In capacity markets, the economic behavior being addressed is quite long-term in nature:

1. Decisions on constructing a new generation or transmission asset that will have an economic life of three to five decades and that will not achieve acceptable returns if it only takes advantage of an attractive near-term market that subsequently erodes; and

2. Decisions on maintaining, preserving, and enhancing existing generation that is losing money in the near term and may not be able to survive long enough to participate in the next market uptick.

The nature of these decisions clearly is not well aligned with price signals, which, depending on the region, flick on and off in intervals that are as short as one month or even for periods of up to four years. The decisions being made in capacity markets are multi-decade decisions and are therefore optimally geared to decade-plus pricing signals. For new construction, it takes three years to site, permit, and build a gas-fired plant today—much more for solid fueled and many renewable projects. For existing generation, capital enhancement and environmental retrofit commitments involve similar planning horizons. These kinds of investments are not going to be made without some degree of long-dated revenue visibility. That long-dated revenue visibility is best provided through long-dated, contractually based capacity payment streams.

A Durable Lesson

One might argue that those kinds of long-term “merchant” capital investments were, in the recent past at least, made by firms on the basis of a combination of short-dated market signals and long-term consultant projections of forward markets. The short-lived period of investment euphoria and excess lasting from 1997 to 2001 recklessly burned through scores of billions of dollars of wasted capital, bankrupting many firms and seriously hobbling previously strong enterprises. More important for the future, it taught a durable lesson to the capital markets to be much more guarded and risk-conscious in any new capital being purportedly committed to take advantage of current or impending capacity shortfalls.

Moreover, current capacity market designs do something worse than simply not motivating the right capital investment decision making. They actively discourage the investment society needs most. Because of the structuring of short-dated market signals, market participants are incented to take advantage of, and to perpetuate, constraint. In a tight capacity marketplace, the market price for installed capacity (ICAP)/unforced capacity (UCAP) may be high, but is visible only for a short period and therefore only hedgeable for a short period.

If a market player commits to build something in a constrained zone that actually fixes the constraint, the reward this player will receive is that the market price for capacity in the newly constraint-relieved zone will drop precipitously. So the reward for doing well is to have a very poorly performing investment. Under today’s capacity market structure, one is only rewarded for taking advantage of, perpetuating, and, at best, only incrementally improving the constraint, not for solving or even materially improving the situation.

This market-design flaw is visible in the failure of a variety of otherwise well-conceived projects in constrained market geographies to achieve commercial acceptance and financeability. Perhaps one of the more well-promoted and visible market bellyflops is the Conjunction transmission project. Designed to arbitrage the surplus capacity and energy markets of New York Independent System Operator (NY-ISO) Zone G into the highly constrained downstate Zone J market, the project was well conceived and engineered, and appeared realistic and cost-effective.

The problem was that potential off-takers were worried that it would work too well. Initially sized at between 1,000 and 2,000 MW, the frequently stated concern of open-season bidders was that the project largely would relieve and unconstrain the bottlenecks between the two zones.

One would think this “complaint” should be a good thing, worthy of accolades—but not in a market environment that values the perpetuation, not the elimination, of constraints. Despite the pleas from project sponsors that the market could tolerate that level of incremental capacity infusion without cratering spreads, the project failed to convince bidders. It now resides in the growing waste heap of potentially viable, beneficial projects that languish due to rules that don’t send the right signals and, therefore, don’t achieve optimal outcomes.

Manifestations of the Problem

The lack of forward visibility of capacity values associated with long-dated capital investments makes the financing of any new “merchant” power project more of a gamble than a risk-mitigatable investment. Gambles involve lots of risk, and risky investments may, in certain circumstances, be financeable, but only with lots of expensive equity capital and limited amounts of debt with high embedded spreads. Expensive financing, at best, perpetuates high-cost, risk-laden markets.

But this “at-best” scenario is not playing out, even in the most highly constrained markets. NY-ISO Zones J and K, essentially encompassing New York City, Westchester County, and Long Island, remain constrained despite market-clearing prices in the NY-ISO six-month auction that have been, for years, clearly high enough to cover the capital cost of new generation. Yet, aside from the modest-sized Ravenswood expansion, and despite a number of sited projects, the only material capacity being added is from utility construction (Poletti expansion, East River) or long-term contracts issued in response to utility competitive solicitations (SES Astoria, Neptune).

This summer, the Zone J UCAP clearing price is over $12.00/kW/month. That should be enough to motivate someone to build something, right? Wrong. Market participants have no visible and financeable view of what capacity prices will be when, if, and after their new capacity increment comes on line. As a result, short-term clearing prices for UCAP remain high and would be even higher were it not for the thoughtful actions of the load-serving utilities, Con Edison, LIPA, and NYPA, to address the situation through self-build and long-term contracting.

The situation is similar in California, but compounded by a perplexing and shifting regulatory environment. Again, as in downstate New York, the only limited capacity increments being constructed are the result of utility term contracting and self-build initiatives, as a short-dated energy-only market clearly isn’t capable of supporting new long-dated capacity investments. The ongoing UCAP food fight in the New England Power Pool isn’t resulting in one more needed megawatt being built where and when it is needed. It is focused on dividing up an existing pie rather than driving toward the end point of creating the right level of future reliability at the right cost to society.

A System That Works

There is a simple, three-pronged approach that has worked structurally for decades, and which, with incremental modifications, still can work today. The basics are as follows:

1. Someone has to be responsible for planning to meet capacity reliability commitments in each defined geographic area;

2. A transparent and open system needs to be put in place to acquire and retain capacity over long-term planning periods at the lowest practical cost; and

3. One needs a fair calculus to be established to allocate all of the costs of acquiring and retaining this capacity across the applicable loads.

On the first leg of the stool, when one assigns the responsibility to provide adequate capacity for customers to “the market”—that is, to everyone—no one is responsible. Ever since electricity was defined, in a political and regulatory context, as a public necessity, a single entity within each geographic jurisdiction has been assigned to determine how best to provide reliability. In the blind rush to embrace the motherhood-and-apple-pie principles of competition and markets, the overarching and clear delineation of who is responsible for capacity reliability and who they are responsible to has been sacrificed.

Second, capacity acquisition should be performed in a transparent, open, and sustainable methodology, geared to achieving least cost. The currently applied structures fail not only because they are short-term oriented, but because their Dutch auction format damages societal interests. That is, the current auction processes reward all bidders with the highest price selected in the auction for the last increment of capacity offered. I may have bid $5.00/kW/month and obviously would have been satisfied to receive the price I bid. But just because the last increment of capacity needed was bid at $10, under most current capacity marked protocols, I get this full $10 price, too. Arguably, this makes sense in energy markets, but not in longer-dated, capital-based capacity markets. Auctions for long-term capacity should award capacity contracts to parties based on least-cost parameters and negotiated deals designed to provide the lowest aggregate cost to consumers.

Finally, there must be a sustainable way to figure out who pays for the capacity being bought and built. Utilities in jurisdictions affording retail competition rightfully argue that there is no certainty of a load-serving entity retaining their current customer base over time. The simple answer is to ensure that all costs of acquiring new capacity and contractually retaining existing capacity, when procured in a fair and visible manner per the second step above, are allocated across all applicable loads within a given zone on jurisdiction.

If there is going to be a “capacity market,” it needs to achieve reliability, cost-effectiveness, and resource diversification. That type of capacity is possible if we assign clear and unambiguous responsibility for getting the capacity, combine it with a fair and transparent long-term system for acquiring or rate-basing the capacity, and cap it with a durable approach for spreading the costs across the entire applicable customer base.

Functional Marketplace

Reduced risks for generators can translate into reduced costs for customers. In unorganized or short-term oriented capacity markets, owners of generation can demand high prices for keeping available otherwise uneconomic, but needed, generation. Much of the UCAP and reliability must run (RMR) payments are flowing to old, insufficient, largely capital-starved units, whose management is not incentivized to think in the long term about capital improvements because they have no certainty about their futures. If these generators were offered the opportunity to bid on long-term capacity commitments, the typical winning bid would be well below recent RMR contract levels or UCAP values in constrained zones. Long-dated, secure contracts then would motivate these generation owners—now contractually committed to keeping their projects in service over the long haul—to make the necessary capital, environmental, and operational improvements for a more productive and profitable energy marketplace.

Long-dated capacity contracts additionally enable leveraged, lower-risk, and lower-cost financing options than those available to at-risk merchant generators. Secure and predictable capacity revenue streams facilitate long-dated, low-cost project or corporate financings which, in turn, translate into the ability to bid and accept materially lower rates for both existing and new generation options.

The assurance of reliable capacity and resource adequacy revolves around very long-dated capital-asset deployment and investment decisions. To ensure that the right amount of capacity is made available at the right time and the right place, at the lowest practical cost to society, a planned, organized, long-term capacity acquisition model can create superior results relative to the short-term or non-existent capacity market variants that have emerged over the last decade.