Nuclear and Coal: Rebirth on the Horizon?

An analysis of the business opportunities behind coal and nuclear plant expansion.
Fortnightly Magazine - November 1 2001


An analysis of the business opportunities behind coal and nuclear plant expansion.

Electric power industry trade publications and the popular media have noted a growing interest in the rebirth of both nuclear power and coal-fired generation. These technologies would be a supplement to, or an alternative to, the natural gas fired generation that appears to be the predominant fuel and technology for new power generation facilities in the coming decade.

Nuclear power enthusiasts express the notion that this country's need for several hundred thousand megawatts of new generating capacity over the next 10 to 20 years, the interests of fuel diversity, clean air issues, and energy security should be enough to drive this rebirth. Coal enthusiasts stress that new coal-fired generators can be much cleaner than existing units, and clean enough to meet anticipated stringent environmental requirements. We have not seen a strong advocacy on the basis of competitive financial opportunity, although the growing number of new coal-fired project announcements suggests that some developers see such opportunity.

With regard to nuclear, there appears to be a divergence of views, with advocates sensing that energy policy issues will drive rebirth, and the financial community sensing that a realistic promise of sound economics and profitability in the competitive generation environment is an essential prerequisite to any rebirth.

According to Douglas Kimmelman, a chief energy analyst with Goldman Sachs, "It really comes down to the fully loaded capital cost. Unfortunately, investors won't make decisions based on national energy policy."1 We agree with the investment community, which lead us to undertake this brief study of the prospective economic and financial drivers of new nuclear and coal fired generation.

Regional Diversity: Are Nuclear and Coal the Answer Everywhere?

Approximately 45,000 megawatts of new coal-fired generation projects have been "announced"; some appear to be under serious development.2 About 12,000 megawatts are proposed for the Western Systems Coordinating Council (WSCC). In the East, the NERC regions SPP, ECAR, MAIN, and SERC appear to be most active. While several leading nuclear plant operators have expressed interest in developing new nuclear plants, we know of no firm development plans at this time. Natural gas fired simple cycle combustion turbine and combined cycle units very heavily dominate the new generation outlook in all markets.

Our study focuses on several key questions (see sidebar, ). It considers new coal plants brought into service in 2006, and new nuclear plants brought into service in either 2006 or 2011, in some or all of nine selected market areas within the Eastern Interconnection.

The key figure of merit is the all-inclusive nuclear or coal plant initial capital cost, expressed in year 2001 dollars per kilowatt, that yields an internal rate of return (IRR) value that is comparable to the value projected for contemporary (in time and location) natural gas-fired combined cycle units. The latter are selected as the reference because of their energy producer role, even though in some markets, notably ECAR, simple cycle combustion turbine peaking units may be financially preferable to combined cycle units in certain time frames. These "parity capital costs" for coal and nuclear may be viewed as benchmarks, subject to the conditions and limitations of this study, for judgments regarding the economic and financial competitiveness of new projects.

A high "parity capital cost" indicates strong market receptivity to new coal or nuclear, relatively high competitiveness with combined cycle, and relatively high financial margins. In other words, new projects having relatively high capital costs can be economically competitive and financially attractive. Comparisons with anticipated new plant costs can provide insights as to whether projects in general or specific projects are likely to be financially attractive, but a cautionary note is in order when comparing specific projects to the results of this study. Individual developers and projects may have very different views of the future and/or financial parameters than those that underlie this study.

Results are presented for three scenarios (see Table 1, ). Scenario 1 assumes substantial installed capacity or ICAP markets everywhere, and substantial revenues therefrom for all generators. This scenario utilizes the reference gas price trajectory. Scenario 2 is based on a market structure in which there are only minimal ICAP revenues, and generators must rely on the energy market for profitability to a much greater degree. The reference gas price trajectory is used here as well. Scenario 3 is based on the Scenario 1 market structure, but assumes a 20 percent higher gas price trajectory. Only one scenario of environmental regulation requirements is considered; alternatives are beyond the scope of the present study. Sensitivity cases were examined for variations in financing costs and for nuclear plant fixed operation and maintenance costs. The following observations emerge from these scenario results.

Clearly, both coal and nuclear benefit from an increase in gas prices. The 20-percent increase in Scenario 3 relative to Scenario 1 yields a parity capital cost increase of about $210 per kilowatt for both types of base load units.

Market structure is somewhat important in this study, with both coal and nuclear benefiting from a structure that incorporates an explicit ICAP market. The advantage in parity capital costs is about $100 per kilowatt for both. However, we have considered only a very limited subset of the many possible bidding strategies that market participants could adopt. Results may be different for alternative bidding strategies.

For coal units, the most favorable and least favorable markets are consistent in the three scenarios. This is substantially true for nuclear as well.

PJM appears to be among the best markets for both coal and nuclear.

MAIN appears to be among the least favorable markets for both coal and nuclear, probably because of the large existing population of base load generators. While it is important for every new project to minimize fuel costs, this may be particularly important for proposed coal plants in MAIN. This study assumed new projects would use a typical local coal with an escalated typical recent historical price. Very recent EIA data indicate that coal prices in Illinois in particular are being reduced, perhaps more than is reflected here. Also, it may be advantageous for a new MAIN project to utilize Powder River Basin coal.

Finding the Answers:
The Key Questions for Nuclear and Coal Development

How strong are the financial and economic drivers for expanded development of coal-fired generation and the reintroduction of nuclear? Under what circumstances are new nuclear and coal plants likely to be financially competitive with new gas-fired combined cycle plants? From the developer and financial community perspectives, taking relative risks into account, under what circumstances would there be a preference for nuclear or coal rather than gas? The key questions include:

1. What envelope of initial capital costs and ongoing fixed operation and maintenance costs must new nuclear plants and new coal plants stay within in order to be financially competitive with new combined cycle plants? Nuclear costs should include allowances for a decommissioning sinking fund and ongoing nuclear waste disposal costs. Coal plant costs should include meeting environmental requirements.

2. How does this cost envelope shift under alternative scenarios of natural gas price trajectories?

3. How does wholesale power market structure affect competitiveness?

4. Are some generation markets more attractive than others for new nuclear or coal plants in the 2006 through 2011 time frame?

5. How will financial risk premiums that investors are likely to demand for new nuclear and coal plants impact on competitiveness?

6. New coal-fired generation is experiencing a stronger rebirth at this time than nuclear plants. Since gas may not be the only competitor, how competitive are new nuclear plants in a scenario of widespread new coal plants?

7. What is the impact of alternative environmental regulation scenarios for coal plants? Under what scenarios would nuclear be clearly preferable economically and financially?

For nuclear, increasing the fixed O&M costs (including capital additions) from $65 per kilowatt year to $100 per kilowatt year reduces the parity capital cost by about $260 per kilowatt. This illustrates both the importance of controlling O&M costs and the financial risk of future O&M cost flyups due to new regulatory requirements and/or equipment replacement needs.

The base case financial parameters used in this study are summarized in Figure 1. The base cases assume that these parameters would be the same for gas-fired combined cycle, coal, and nuclear.

It seems reasonable to expect higher financing costs for coal because of financial risks associated with future environmental regulations. Higher financing costs in terms of a 10-percentage point increase in equity (vs. debt) requirement, and half a percentage point increase in debt cost reduces the parity capital costs by $80-100 per kilowatt, depending on the scenario. It seems reasonable to expect even higher financing costs for nuclear because of financial risks associated with new NRC requirements that may be triggered by events at the subject plant or even at another plant having a different owner, by equipment performance and replacement needs, and with public acceptance shifts. Higher financing costs in terms of a 20-percentage point increase in equity (vs. debt) requirements and a full percentage point increase in debt cost reduce parity capital costs by about $165-212 per kilowatt, depending on the scenario.

Making the Numbers Work: What are the Capital Costs?

The Energy Information Administration's "advanced nuclear case" shows a capital cost of about $1735 per kilowatt in 2001 dollars for plants in the 2005 time frame, and about $1560 per kilowatt for plants in the 2010 time frame3 These figures are quite a bit higher than our parity capital costs. In a recent interview with Entergy executives on nuclear issues, they said it would take a nuclear design whose cost is less than $1000 per kilowatt to stimulate high interest on the part of Entergy. In addition they indicated that there are several designs on the drawing boards that meet this criterion.4 We hope so, since our parity capital costs suggest that this is the cost range needed for financial competitiveness.

Cost estimates for coal-fired plants of $1082 per kilowatt to $1190 per kilowatt would meet cost requirements if they were constructed with a modular approach.5 This represents a 10 percent to 15 percent cost reduction relative to stick built plants (IBID). Accordingly, a 2001 capital cost of about $1150 per kilowatt for conventional pulverized coal plants with in-service dates in the 2005 time frame (Footnote 3, Table 43, IBID). These figures appear to be almost competitive with our parity capital costs in the most favorable markets, assuming our reference gas price projection. More comfort is obtained with the 20 percent higher gas prices.

This study assumed a relatively favorable scenario of environmental regulations. Alternative scenarios may significantly reduce the parity capital costs for new coal plants, indicating reduced competitiveness. These scenarios may increase parity capital costs for nuclear. Not surprisingly, new coal and nuclear projects clearly involve bets on gas prices, market selection, and environmental regulations.

Power Markets: Projections For Nuclear and Coal

Competitive generation market context is important to understanding the results of any study of this type. This study looks at the entire Eastern Interconnection plus ERCOT. It assumes that a multi-regional competitive power generation marketplace exists throughout this very large geographic area.

For modeling purposes, this area is divided into about 165 individual markets that are linked together in accordance with today's transmission system.

Think of these markets as bubbles on a bubble chart that are joined by lines representing transmission links. Many of these markets are individual merchant plants with no native load obligations. About 50 of these markets are load regions, most having generation within them as well. Major transmission interfaces may limit power flows and transaction opportunities at various locations and times within this broad geographic area. The model includes 30 such interfaces.

It is assumed that regional transmission organizations (RTOs) form everywhere, but in a relatively Balkanized pattern rather than the very few super-regional entities that the FERC is attempting to promote in its orders of July 2001. These RTOs reduce the pancaking of wheeling charges and greatly increase power transaction opportunities relative to the historic marketplace, but probably not to the level envisioned by the FERC plan.

We expect that most new generation, including the gas-fired, coal-fired, and nuclear plants that are the direct subjects of this study, will be merchant plants. Each merchant plant can obtain revenues from the wholesale power marketplace that can be viewed as having three segments: an energy market, an installed capacity or ICAP market, and an ancillary services market. Revenues from the three market segments must be sufficient to cover not only variable operating costs but also fixed operating costs, paying off debt financing, and providing additional revenues so that the equity holders realize target returns on investment or internal rates of return (IRR). It is neither necessary nor realistic that all generators meet financial targets, but most existing generators and new projects needed to maintain generation reserve margins must be able to do so. Revenues from these market segments that are net of the variable costs of securing these revenues are deemed contributions to fixed charges (CTFC). In recent history, the tendency has been to construct gas fired generation because it is the least capital intensive, fastest to permit and construct, and supported by favorable expectations of natural gas availability and cost. Merchant plant developers will be induced to develop more capital intensive technologies, such as coal-fired and nuclear plants, only when the additional contributions to fixed charges from the marketplace are at least sufficient to finance the capital premium while meeting financial targets.

Our market model assumes that reliability will be maintained by enforcing a generation reserve margin obligation in each reserve-sharing pool. In this study, reserve-sharing pools are typically assumed to coincide with a NERC region or sub-region. Reserve margin targets, based on today's targets, range from about 13 to about 20 percent of seasonal peak load, depending on the region. (See sidebar, )

Raber Consulting maintains an ongoing list of announced new generation projects and tracks their development status. As we sense that a project is likely to come into service, we add it to the database of our energy market model. In the early years of the study, this list of new projects determined whether each reserve-sharing pool has more capacity or less capacity than is called for to meet its reserve obligation. The model adds expansion unit projects to each reserve-sharing pool when the existing level of generation, net of economic retirements, falls to the minimum level required by the reserve margin obligation. In this study, it is assumed that all expansion units will be natural gas fired. A developer perspective on which of these technologies will provide the better IRR drives the mix of simple cycle combustion turbines (CTs) and combined cycle units in each pool. An "equilibrium" mix, which would produce the same IRR for both technologies, is sought. In most market areas, the expansion unit mix is heavily tilted toward combined cycle. ECAR is a notable exception, wanting perhaps 30 percent to 50 percent simple cycle CTs in the mix in order to balance out existing base load generation that will continue operating over many years.

Building Power Plants in Competitive Markets:
An Explanation of Assumptions

The competitive generation market is viewed as having three segments: an energy market, an installed capacity or ICAP market, and an ancillary services market. The energy market is modeled using the Inter-Regional Electric Market Model.6 This model combines economic dispatch with broad-based inter-market energy transactions to arrive at energy market prices as a function of time, and energy market contributions to fixed charges for each individual generating unit. The ICAP market is modeled with Raber Consulting's spreadsheet model, which is basically an income statement that accepts IREMM results along with accounting, financing, and tax parameters to develop IRR values as a function of time for individual projects. This model is also used to compare new merchant plants of the three technologies. Ancillary services market revenues are expected to be relatively small for the types of generators of interest here, and will depend on market development, generator location, and generator operating and electrical characteristics. A simplified approach was adopted in this study, assuming that the ancillary services market provided a 5-percent bonus to the combined contributions to fixed charges from the energy and ICAP market segments.

Two long-term natural gas price trajectories were considered. The reference projection is based on the reference scenario in EIA's Annual Energy Outlook 2001.3 Near term prices were updated to EIA's March 2001 Short-Term Outlook. Escalation rates for fossil fuels other than gas were also considered.3 The high gas price trajectory has gas prices 20 percent above the reference prices starting in 2003. The prices for other fuels are unchanged from the reference in the high gas price scenario. Gas prices are region-dependent. As seen in the results, the financial performance of coal and nuclear plants depends strongly on the spread between their fuel costs and gas costs.

This study assumes that several large companies who already own and operate multiple nuclear plants would be the developers of new plants. These include Dominion Resources, Excelon, Southern Energy, Constellation, Entergy, and Duke. Several of these companies have expressed interest in doing so. New nuclear plants will almost certainly be developed at existing nuclear plant sites.

Economies of volume in manufacturing new nuclear plant components and erecting new units will almost certainly be necessary if competitive cost levels are to be achieved. This study assumed that commitments for eight new plants would be made among the companies listed above, with costs shared in some manner. Four of these plants are assumed to enter service in 2006, which is probably earlier than practicable. The other four are assumed to enter service in 2011. These units are assumed to be conventional pressurized water reactors, for example the standard design AP600, with the characteristics shown in Figure 1, page 44. In this study, fixed O&M costs are assumed to include capital additions. Nuclear fuel costs are assumed to be independent of plant location, and constant in time at about $0.70 per million BTU. This figure is viewed as covering the disposal costs for high-level radioactive wastes whose production is proportional to plant operation; for example, spent nuclear fuel.

The new coal plants to come into service in 2006 are assumed to be pulverized coal units having the characteristics shown in Figure 1, with high efficiency scrubbers for SO2 removal.

With regard to environmental requirements, this study assumes that extensive NOx emissions controls will be required everywhere within the EPA "SIP Call" regions. Most existing generators and all new generators are assumed to have NOx emission controls, most often selective catalytic reduction or SCR for major units.

Outside the "SIP Call" regions, it was assumed that NOx emissions controls are not required for either existing or new units. This may well be unrealistic, but tends to maintain competitiveness of new coal-fired merchant plants relative to existing coal-fired units. Unlike nuclear, fuel costs for coal plants vary significantly from region to region.

Each new merchant coal plant was assumed to utilize coals, and see coal prices, that are typical of existing units in that region. In our modeling, each coal-fired generator is assigned its specific coal price and sulfur content based on recent historic data. Obviously, a new unit should seek the best coal prices available because this will directly affect competitiveness. Our study does not assume that these new plants can achieve coal prices below what has been typical for their market areas. This may be important for a new coal unit in MAIN in particular. We assumed it would utilize local coal rather than perhaps cheaper coal from the Powder River Basin, which might enhance its competitiveness.

The fixed operation and maintenance costs for coal units shown in Figure 1 are taken from Table 43,3 escalated to 2001 dollars. The nuclear plant fixed O&M costs are shown in Figure 1.7 This $65 per kilowatt-year is reasonably representative of 1997 through 1999 existing plant experience for the best quartile of the existing plant population. The alternative figure used in sensitivity cases, $100 per kilowatt-year, appears reasonably representative of the third quartile experience. Table 433 indicates a fixed O&M cost just under $60 per kilowatt-year in year 2001 dollars. In this study, it is assumed that fixed O&M costs include capital additions. Price projections for SO2 and NOx allowances were taken from the EIA.8

This study assumes no requirements for carbon emissions, again a potentially unrealistic assumption. Since coal units would be impacted more heavily by CO2 emission regulations and allowance costs than natural gas-fired units, this assumption tends to enhance the competitiveness of new coal-fired generation. This assumption also tends to reduce the competitiveness of new nuclear plants relative to either gas-fired or coal-fired generators.

The following sensitivity cases were examined in all three scenarios. Sensitivity case results are summarized in Table 1, page 43.

1. New nuclear plant fixed O&M costs increase from $65 per kilowatt-year to $100 per kilowatt-year, in year 2001 dollars.

2. Financing cost premiums are required. For new coal units, this premium is assumed to be a shift in capitalization from 40 to 50 percent equity, and an increase in debt cost from 8.25 percent to 8.75 percent. For new nuclear units, the shift in capitalization is assumed to be from 40 to 60 percent equity, and the increase in debt cost from 8.25 to 9.25 percent.

In essence, this study assumes that new generation will continue to be overwhelmingly gas-fired, with progressive improvements in CT technology. We are examining the financial competitiveness of a small population of new nuclear and coal-fired generators introduced into this "world of gas." In most respects, this should maximize their competitiveness.

Two alternative power market structures were examined for impact on the competitiveness of new coal and nuclear plants. The market model used in Scenarios 1 and 3 assumes that there is an explicit ICAP market in addition to the energy market. In essence, this ICAP market becomes the vehicle for enforcing reliability requirements. It is assumed that energy market competitiveness maintains energy market prices at or near prevailing variable costs. In this study, a minimum profit of $3 per megawatt-hour is added to variable cost to obtain generator bid prices during on-peak hours. This minimum profit before sale was scaled back to $1 per megawatt-hour during mid-peak hours, and to zero during off-peak hours. The ICAP market is then relied upon to provide the revenues necessary for new gas-fired combined cycle and/or simple cycle CTs to meet the specified market average target IRR values when new capacity is needed for reliability. This market model anticipates very low ICAP values in early years when reserve-sharing pools have excess capacity. This value was taken to be $10 per kilowatt-year in today's dollars.

In the market model used in Scenario 2, it is assumed that there is a minimal ICAP market, limiting its revenues to $10 per kilowatt-year in today's dollars. The energy market is now called upon to provide the necessary revenues to allow new gas-fired projects to meet target IRRs when they are needed to meet reliability. This was done by increasing only the on-peak minimum profit before sale on a region-wise basis, but applying it to all generators within each region. In this market model, this energy market bidding strategy was allowed to persist during the early years of excess capacity. This produces much higher IRR values than the market model used in Scenario 1 for merchant plants introduced prior to the need for new capacity to maintain reliability, typically in the 2004 to 2008 time frame.

The study horizon was 2020. For new merchant plants installed in 2006, a target IRR of 16 percent in 2020 was assumed. For new merchant plants installed in 2011, the target IRR was assumed to be 12.5 percent in 2020. Capacity prices and energy market minimum profits were adjusted to have gas-fired merchant plants close to these targets. Capital costs for nuclear and coal units were then adjusted to have their IRR values match those of contemporary gas-fired combined cycle units.

1 , May 30, 2001. Article entitled "Nuclear Power Enthusiasts Grapple With Wall Street Skepticism", page 17.

2 Betsy S. Vaninetti, "The Race for New Coal-fired Generation". Article in , July/August 2001.

3 Energy Information Administration Report: Annual Energy Outlook 2001, DOE/EIA-0383 (2001); December 2000. Also, companion documents presenting assumptions and bases.

4 , July 2001 interview article entitled: "Keuter: On Entergy Nuclear's Acquisitions and Growth."

5 Jerry Gotlieb et al., "Power Plant Design: Taking Full Advantage of Modularization." Article in , June 2001.

6 The Inter-Regional Electric Market Model is developed, maintained, and licensed by IREMM, Inc. See their Web site at or call Mr. Wayne H. Coste at 860-651-1600 for more information.

7 Nuclear Energy Institute status report entitled: "The Outlook for Nuclear Energy in a , January 2001.

8 Energy Information Administration report: "Analysis of Strategies for Reducing Multiple Emissions from Power Plants: Sulfur Dioxide, Nitrogen Oxides, and Carbon Dioxide. SR/OIAF/2000-05. December 2000.

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.