"Spark spread" sets value, but as prices diverge from system

lambda, merchant plant buyers will be flying blind.

Many power plants will be bought and sold in the next decade. Some utilities will divest power plants as required by regulators; others will sell for strategic reasons. Most of the plants sold likely will become merchant plants, with no guaranteed market for their electric output. Merchant plant activity is already significant and growing. The value of these plants will depend on how well they can perform in an uncertain market.

Plant owners in a competitive market constantly will re-evaluate their portfolios of generating assets. In so doing, their expectations of plant financial performance may differ, based on market prices, risk-return requirements, tax factors, upgrade plans or operational practices. Indeed, alternative valuations will make many plant sales transactions feasible. How will potential buyers and sellers set a value on merchant plants?

Current or prospective owners likely will use several new measures of a merchant plant's value. Of those, the "spark spread" will prove the most critical (em and the most uncertain (em as wholesale power prices evolve from a simple mirror of short-term variable costs (system lambda) to reflect more long-term fixed costs. Moreover, buyers must time their investment to anticipate a positive spark spread for the specific niche in which the plant operates.

Valuation: Price vs. Fuel Cost

The value of a power plant (like any productive asset) equals the discounted net cash flow it is expected to generate for the owner over its expected financial lifetime. Most merchant plants operate simultaneously in two related and volatile markets: power and fuel. %n1%n Within the spread between those two markets, known as the "spark spread," %n2%n they must produce enough revenue through power sales to cover all operation, maintenance and other fixed costs, service all debt, meet other financial charges (such as taxes) and earn a profit for equity investors. Yet, a spark spread sufficiently positive to do all this is not guaranteed. Indeed, the spark spread could be negative (see Figure 1).

Four considerations determine the value of a power plant: 1) the outlook for an adequately positive spark spread; 2) the likely sales volume at the expected electricity prices; 3) plant-specific cost factors (plant type, fuels, age, condition, etc.); and 4) the uncertainty and risk associated with each of the above considerations. Each of these factors is likely to be highly dynamic, even for an individual plant. As a result, the value of a plant to prospective owners will vary over time.

Spark spread marks the most uncertain component of this valuation and the element over which the plant owner exerts the least control. Fuel prices and wholesale electric prices are inherently volatile (see Figure 2). A company can manage spark spread in two ways: strategic investment in plant assets and by coordinating fuel management with power marketing.

Usually, the cost of new capacity will set an upper limit on the value of existing plants (em companies will not pay more for an existing plant than a new, highly efficient one. And technical innovations have lowered the cost of building these plants. The cost of building a combined-cycle, gas-turbine unit of 250 megawatts, for example, now runs roughly $500/kW, down from about $700/kW in the early '90s. Similarly the cost of a 250-MW coal-fired plant has declined, from about $1,500/kW to about $1,100/kW today.

The Price Component: Greater Uncertainty

Spot markets today set wholesale power prices at the margin. That process should continue. What will change, however, is the definition of the "margin."

Until now, the short-run variable cost (em the system lambda (em has been the key determinant of the wholesale price of electricity. This mechanism has worked because integrated utilities have recovered most of their costs from their retail customers (em a captive franchise. For these utilities, the wholesale market is just a place to gain incremental revenues by boosting capacity factor. This type of market tends to create a positive correlation between the wholesale spot price and the short-term variable costs. And, since fuel costs represent the dominant share of variable cost, the correlation has linked power prices to fuel costs. %n3%n Moreover, this correlation could grow, as fuel suppliers seek to tie their prices to electric prices in an attempt to garner market share. This correlation tends to protect plant owners in that, over time, increases in fuel prices will correspondingly increase electric prices. Rising fuel prices tend to raise revenues for all electric generators.

This market cannot continue, however. Future wholesale power prices are unlikely to look anything like those in today's spot markets. At some point, the market must move beyond short-term marginal pricing to recover some of the longer-term fixed costs, beyond those reflected in the system lambda. Merchant plants will need to cover longer-term variable costs in the wholesale electricity market (em not the retail market. Without covering what are now considered "fixed" O&M costs, operating merchant plants would have to shut down. Thus, the key question for power plant investors is simple: When will the market move to this longer-term equilibrium?

The answer will prove elusive. The outlook is for uncertainty, volatility and cyclical price behavior. Wholesale electric prices already show highly volatile behavior and this tendency well may increase. In the competitive Scandinavian electricity market, for example, average annual prices vary by a factor of four. Cyclical market behavior has yet to come to electricity, but is likely to play a significant role in the industry, one that will prove critical to investment decisions. %n4%n

The transition to competition could prove the most volatile, because the change will occur in increments. The correlation between fuel and power prices likely will weaken as factors other than fuel price (e.g., the need to cover cost currently defined as "fixed") assume more importance in wholesale electric pricing. This change will make it more difficult to keep the spark spread positive. Rising fuel prices no longer will raise all electric revenues.

Instead, investors must examine risk in both power and fuel markets simultaneously. The goal will be to keep the spark spread positive. Plant-specific fuel issues must be addressed in evaluating and valuing power plants, including:

• Fuel Position. What fuels are used? How are they purchased? At what price? What commitments will convey to the new owner? When do contracts expire?

• Supply Options. Can new, less-expensive or less-risky suppliers be obtained?

• Competitive Position. A plant may lose its position in the market if competitors improve.

Overall, investors should expect changes in linkage between power prices and fuel-market dynamics. In fact, merchant plant buyers largely will be flying blind over the next few years, as we don't yet have much of the experience or information needed to develop a coherent view of the market (em and won't until the merchant plant market has existed for some time.

Deciding When to Sell

Power plant owners will sell their generating assets when cash is more valuable than continued ownership, although they may also seek other benefits. %n5%n Plant owners may want to sell for one or more of three basic reasons:

• Mandated Divestiture. Massachusetts and California utilities, for example, are auctioning their power plants in response to state-mandated deregulation.

• Mitigation of Loss. Its operation may be uneconomic or inadequate due to expected market conditions or a deficiency.

• A New Business Strategy. Some companies may decide to get out of generation. The cash may be more valuable elsewhere.

Plants may also be sold with the provision made for continued later use of the output but not as owner. For example, Commonwealth Edison sold its 490 MW State Line plant to Southern Electric International and its 1,108 MW Kincaid plant to Dominion Energy. ComEd will supply the fuel and will take all of the electric output from both plants for 15 years.

The tendency of most plant owners (who want to stay in the business) will be to try to sell their least-attractive power plants. The problem with this strategy, of course, is that they are likely also to be the least attractive to a buyer.

Deciding When to Buy

Power plant valuation for evaluating plant acquisition decisions in a competitive market requires five sequential activities:

1. Weigh the Spark Spread. Determine the outlook for a positive spark spread for the specific plant. Both electric and fuel market issues must be considered. Fuel issues include plant, competition and market factors. Electric issues include trends and cycles in the markets of interest.

2. Study the Candidate. Assess plant-specific characteristics. Estimate the expected contributions to set cash flow under a variety of potential market circumstances.

3. Deflect Risk. Determine how well market risks can be managed using the various techniques available to manage them.

4. Balance the Portfolio. Assess the correlation of the proposed investment with other portfolio assets and perform a thorough sensitivity analysis using different assumptions. Diversification is a key issue. Few integrated utilities today are well diversified. %n6%n

5. Test the Price. Determine whether the market will support the price at which the power plant will be bought or sold. Most often, there is not yet sufficient information to make this assessment.

Keeping Spark Spread Positive

Once a power plant changes hands, the owner's ability to manage market risks will affect its value. After the plant is purchased, a favorable spark spread can be "locked in." This requires coordinated activity in both the electricity and fuel markets. Such measures may include:

• Selling in Load Pockets. Many such opportunities today are created by transmission constraints that create load pockets (e.g., Manhattan, Long Island, San Francisco) or larger constrained areas. Such constraints are likely to diminish in the future as companies invest in transmission-line upgrades and new (often peaking) capacity is built within constrained areas or as regulators address the market power issues. As a result, unique opportunities will grow more difficult to find.

• Electric Market-Based Fuel Contracts. Such contracts tie the price of fuel to that of electricity on specific transactions or through indexing. They are now gaining in popularity among electric generators and fuel suppliers seeking new, cooperative ways to stay competitive. Some fuel suppliers, for example, are touting "alliances" with electric generators. How effective this will be in a market where such practices are commonplace remains unknown. Moreover, these practices will affect the pricing dynamics.

• Tolling. Tolling is the practice of operating a power plant for a fee under contract to another party that provides the fuel and sells the output. It shifts market risk for both fuel and power to the other party. The long-run usefulness of tolling as risk management is questionable, because the plant still needs to operate in the underlying markets no matter who provides the fuel or takes the contract.

• Reverse tolling. Reverse tolling of fuel is a way for a power plant to make money for its owner even when high fuel prices make generation uneconomic. In such a situation, the value of fuel is greater in fuel markets than converted to electricity in the electric markets. The generator may sell contract fuel on spot market for profit and buy incremental electricity. In essence, the generator is conducting arbitrage between the electricity and fuel markets. Reverse tolling requires the generator to establish prior relationships with fuel marketers to sell high-priced fuel as the opportunity arises.

Power plant owners also can combine these techniques. Tolling and reverse tolling may be used, for example, in a complementary fashion to maximize return for a specific plant. The plant may be tolled when the spark spread is positive and reverse tolled when it is negative.

In essence, one invests in the use of a plant asset when the spark spread is positive and in a fuel asset when it is negative. This practice (along with indexing fuel investments to the wholesale power market) serves as a real example of "convergence" of the wholesale electric and fuel markets. t

Jeffrey P. Price is president of Resource Dynamics Corp. The author would like to thank Alex

R. Henney, John H. Herbert and James T. Doudiet for their help with this article.

Merchant Plant Sales Activity

New England. In early August, New England Electric System sold its non-nuclear generating assets (about 4,000 MW) to U.S. Generating Co., an affiliate of Pacific Gas & Electric Co. The sale price of $1.59 billion ($312/kW) included about 1,100 MW of purchased-power contracts with wholesale suppliers.

California and Elsewhere. Another plant sold on a merchant basis is Dusquene Power and Light's 288-MW Fort Martin plant, which was sold to AYP Capital for $169 million ($587/kW). Anticipated plant sales by Boston Edison and by California's three major investor-owned utilities (next year) may provide more price points and help the players to calibrate their expectations. These large, unique events may or may not set precedents.

Merchant Plant Investment Activity

Currently Operational:

Indeck Industries, 38 MW cogen, Pepperell, MA

Mid America Power, 53 MW repowering, Stoneman, WI

AYP Energy (Allegheny Power), 288 MW, Ft. Martin, PA

Williams Field Services, 74 MW, Bloomfield, NM

Under Construction:

CSW Energy, 300 MW, GT, Sweeny Texas

Planned or Under Development:

Calpine Power, 240 MW, TX

Berkshire Power, 252 MW, Agawam, MA

LG&E Power, 300 MW cogen, Corpus Christi, TX

US Generating Co, 240 MW, Sumas, WA

US Generating Co., 500 MW, OR

US Generating Co., 400 MW, Charlton, MA

Diamond Energy, 500 MW, Southern CA

Diamond Energy, 240 MW, OR

Duke/Louis Dreyfus & Puget Sound P&L joint venture to develop merchant capacity in Pacific Northwest

CNG Energy Services, 240 MW, Pacific Northwest

Power Development Corp., 272 MW, Westfield, MA

Energy Management, Inc., 170 MW, Dighton, MA

Energy Management, Inc., 250 MW, Tiverton, MA

UtiliCorp United/Air Products joint venture, Midwest

Commonwealth Gas, 300 MW, Chesapeake, VA

Source: Stephan H Watts, Mcguire, Woods Battle & Boothe LLP

1The exceptions are hydro and other renewable energy plants. For these types of units, the major risk factor is the uncertain availability of the energy source (e.g., water, wind) and the correlation of its availability with electricity price.

2The spark spread is defined as the difference between the fuel and electricity price, expressed in equivalent terms. For example, assume that the power price is $20/MWh and the fuel price is $1.50/million Btu for a 10,000 Btu/kWh plant. The spark spread = $20/MWh - [(10 x 106 Btu/MWh) x $1.50/106 Btu)] = $5.00/MWh.

3This correlation depends on the market share of the specific fuel. It is higher for fuels that are used to generate a large share of a region's electricity. It also may vary by season.

4Cyclical price swings in commodity markets reflect lags in investment by producers that lead to periods of surplus or scarcity. While prices may well reflect short-term variable costs during periods of excess capacity, they likely will run much higher when capacity runs short.

5Regulatory authorities may change this incentive. The California auction essentially nullifies this incentive because any proceeds below book value are to be recovered in the competitive transition charge and any revenue above book value will accrue to the benefit of the ratepayers.

6Traditionally, electric utilities considered a power plant as a component of an integrated physical system built to deliver reliable electricity to customers at minimum cost. Investors, however, will view a plan as an asset in a financial portfolio, along with sales and purchases of fuel and power.

Analyzing investment decisions from this new perspective requires estimating the proposed plant's contribution to the risk-and-return characteristic of the total portfolio, including scale, diversification and flexibility of commitment. Portfolio management techniques could be borrowed from other commodity industries (such as gas), but they must account for unique aspects of the electric power industry, including lack of storage and the need for spread management.

Articles found on this page are available to Internet subscribers only. For more information about obtaining a username and password, please call our Customer Service Department at 1-800-368-5001.