A review of which technologies and companies stand to win and lose as a result of the 2003 blackout.
Mishap, human error, and malice regularly crash the electric system. We have lurched from the Western economic power crisis of 1999-2000 to the Eastern reliability power crisis of 2003. Neither more studies nor more blackouts have changed what's been built-an excessive quantity of large generation plants dependent on relatively few major transmission lines. On its current course, the grid's inevitable destination is disaster.
Recent congressional testimony blames the root cause of the blackout on everything from deregulation to inadequate central authority over the grid. While there is an underlying truth to all these casual factors, this line of thinking misses the underlying problem.
The Rocky Mountain Institute's 1981 study explains why the power grid, our nation's most complex and critical infrastructure, remains profoundly vulnerable: the grid's centralized system architecture makes it inherently prone to precisely the sort of instability we've witnessed during the past few years. America relies on aerial arteries and precise electronic signals to keep huge machines rotating in exact synchrony across half a continent. And continued growth of the grid, combined with the liberalization of the country's wholesale power market, has placed even more strain on the system. When more and bigger power lines link more and bigger power plants, the grid becomes less stable in new ways and over wider areas. Restructured electric markets challenge this constrained system with transmission transactions of a frequency and complexity for which the grid was not designed.
The electric industry once again finds itself at a crossroads, confronting it with three basic choices: the supply-side path, the distributed path, or the status quo. This article does not debate the competing claims for solutions to the reliability crisis, but rather observes who wins and loses if we pursue any of these paths.
Who Wins and Loses If We Do Nothing?
In this scenario, we expect that the distributed generation companies with proven technologies will benefit significantly. Shipments of diesel generation sets and conventional gas turbines doubled to 5,000 MW/yr. after the California power crisis. Companies like Caterpillar, Cummins, and Briggs & Stratton would experience substantial growth. Similarly, providers of on-site power storage and power conditioning equipment such as American Power Conversion, Beacon Power and Active Power would also see expanding sales, as reliability-sensitive customers invest in a private solutions to meet their needs. Providers of innovative technologies-including companies like American Superconductor and Composite Technology Corp.-that enhance the capacity of existing transmission also will win, since new transmission would be difficult to site, making it imperative to upgrade existing lines.
The other obvious winners in this scenario are strategically located generators-especially those with peakers or rapidly dispatchable gas-fired units-that can continue to collect rents from underserved load pockets and from periodic price spikes occasioned by chronic transmission con-straints. On the other hand, we do not expect the innovative distributed generation companies to fare as well. Customers seeking business interruption insurance want proven solutions. They are not willing to fund technology development.
The big loser in this scenario, aside from consumers who are plagued by periodic outages and the associated rate impacts, may be distribution-focused utilities, including players such as Consolidated Edison, Pepco, and Northeast Utilities. A world in which the value of reliability continues to increase, while the cost of self-providing enhanced reliability decreases thanks to innovations in distributed resources, could lead to growing competition to wires companies' "natural monopoly."
Distributed generation poses four primary threats to the existing distribution utility business model. First, distributed generation results in the loss of revenue under traditional tariff structures; the customer simply is purchasing fewer kilowatt-hour or fewer distribution services. Second, more substantial market capture by distributed generation can create a new class of stranded asset within the distribution system-grid capacity no longer needed. Third, the ability of distributed generation to enter more rapidly than centralized generation or transmission upgrades can partially strand new capacity additions. Fourth, the combination of the first three threats can create a "death cycle" in which the higher prices to remaining customers induce more of them to leave this system, creating a self-reinforcing cycle of ever-increasing unit prices.
Even modest revenue losses have substantial impacts on profit. The problem for utilities is that their high unit gross margins (revenues less cost of goods sold) are volume needed to cover largely fixed operating and depreciation costs. Hence, they are highly vulnerable to volume losses, as evidenced by their earnings' strong sensitivity to weather (see Figure1).1
The Supply-Side Response Path
The under-investment in the U.S. transmission grid, especially relative to the increased demands being placed on it (see Figure 2), and the role of regulatory shortcomings in this neglect are the subject of endless discussion. The most commonly cited problems include: lack of adequate price information to signal the need for investment, controversy over the allocation of the costs of transmission investment, overlapping or poorly defined regulatory jurisdiction, a lack of mandatory reliability standards, excessive red-tape and complexity in transmission siting, and uncertainty over what regulators will demand next. This summer's blackout appears to create the potential for legislators to grapple with at least some of these topics. Combined with favorable tax treatments and lucrative guaranteed rates of return, the path could be clear for accelerated investment in the grid.
Just maintaining the historical capacity relationship norms of 201 MW-miles/MW demand throughout the next decade would requires the construction of 26,600 GW-miles, compared with planned construction of only 6,200 GW-miles.2 If new construction maintained current levels of transmission capacity, investments would total an estimated $56 billion. This is equal to the current book value of transmission assets and about half of the $105 billion investment forecast for new generation capacity over the 10-year period. Putting aside the question of who will pay for this, the impact on power costs will depend on how the cost of the increased capacity (at about 1,500-1,800/kW) will compare to savings through reduced transmission losses (1 to 2 percent, say some experts) as well as improved dispatch efficiency.
We envisage four potential sets of winners in this scenario. First and most obviously, there are the transmission equipment manufacturers and engineering firms who build and maintain the grid, such as ABB, the Shaw Group, and Quanta Services. Note that even though about 58 percent of expenditures in transmission and distribution networks is for capital rather than operating expenses, approximately 60 percent of total expenditures are for ongoing testing and maintenance rather than for the major equipment (towers, transformers, substations) and construction projects often associated with grid upgrades. As a result, a plethora of smaller engineering and construction firms and highly specialized testing and diagnostics firms would benefit from a grid overhaul as well as the better known names. Second, the emerging set of pure-play transmission system owners or operators-as well as private investors such as KKR and Berkshire Hathaway that are acquiring positions in power transmission-are likely to benefit from increased investment and focus on the grid, including names such as National Grid, TransElect, and American Transmission LLC.
Owners of more efficient but remotely located generation (i.e., baseload coal and nuclear units), are likely winners in this scenario, displacing higher-cost, strategically located units. We also would expect that large-scale wind, which can now be produced at 2.5 cents/kWh (inclusive of 1.7 cents/kWh federal tax credits), will become highly competitive with gas plants, since wind power prices are far less volatile, and wind-hydro combination create an equivalent firm power. The Dakotas have a technical potential of more than 600 GW in wind energy that is trapped because of lack of transmission capacity. Finally, improved interstate transmission capacity and pricing could result in a shift of rents to Powder River Basin (PRB) coal players such as Peabody and Arch, which may be able to use wire rather than rail to transport PRB coal to their customers, at the expense of BNSF and Union Pacific. Intuitively, the net impact on natural gas-the preferred fuel of the smaller peaking and load-following units that currently profit from transmission constraints-in this scenario likely would be negative.
Implications of the Distributed Resources Path
Proponents of the distributed resources path argue that it costs less and is inherently more reliable. Without arguing the merits of this claim, if it were true, are there enough potential distributed resources to achieve the same result as the proposed supply side response? What might such a distributed system cost?
Demand response signals customers when power is scarce, so they can choose convenient ways to trim or defer power use. Smart meters coupled with two-way control technology, along with time-variant price signals, are needed to enable demand response. EPRI has estimated that demand-response programs could reduce peak demand in the United States by an additional 45,000 MW, or about 6 percent of peak baseline usage.3 Such a demand response capability could cost $200/kW to $600/kW to install.
The most recent comprehensive study of national energy efficiency, Clean Energy Futures, performed by the five national energy laboratories, found energy conservation could deliver a 6 to 10 percent national electricity demand and would cost an average of 2-3 cents/kWh.4
Proponents of the distributed resource path emphasize the value of diversity in generation resources, so we assume a portfolio of these resources would be employed. Cogeneration represents the lowest cost distributed resource at $600/kW to $800/kW, while fuel cells are the most expensive at $4,000/kW. Sixty gigawatts of distributed generation already exist, and several studies suggest the economic potential is between another 60 and100 GW.
Investing in energy efficiency would avoid investments in transmission, distribution, and generation. The magnitude of the savings is augmented by several factors, including line losses and lower planning risk (since distributed resources are inherently less lumpy than large transmission lines) and reduced exposure to energy market volatility. Despite all the existing obstacles, distributed generation is already growing rapidly- partly because its hidden economic benefits often boost value by about tenfold.5
Demand response acts as a hedge by dampening price spikes when power is scarce, and by providing cheap insurance against artificial scarcity. Had California installed more load management, equivalent to 1 percent of its peak load, shrewd investors could simply have shorted the power market (bet on lower prices) in 2000-2001 when suppliers were withholding supply to raise prices-then activated their load management, dropped prices, averted shortages, and taken more than 1 billion from the miscreants.6 McKinsey & Co. estimate that if national load were reduced by 5 percent, the resulting savings from avoided peak power prices would be $15 billion a year.7
The second-order effect on gas pricing is equally if not more intriguing. Consider that almost all peak power is made by inefficient gas-fired combustion turbines, so shaving just 5 percent of U.S. peak electric load would save 25 percent of the gas used for power, or 1.5-2.1 Tcf (9.5 percent of total U.S. natural gas use)-enough to return gas prices to their previous normal range for years.8 The value of the gas price reduction has been estimated at between $15 billion to $40 billion a year.9
There would be many winners from the distributed resource path. Society at large would prosper because electric service could be provided at lower cost with higher reliability. Regulators would achieve their objective of fair and competitive electricity markets at the wholesale and retail level, since distributed generation would add more competition and liquidity. Further, grid reliability and energy system resilience (hence security) would be enhanced. Business customers would benefit from a wider spectrum of options to manage their energy needs, greater grid reliability and the ability to reap commercial profits from advantaged locations. Progressive utilities would benefit by sharing in the savings from the lower revenue requirement-in effect, earning a higher return on assets. Clearly, distributed generation manufacturers and energy service companies would become high-growth industries, attracting capital and creating jobs. The environment will benefit from lower air pollution more than it would with centralized generation.
Traditional and next generation distributed resource providers both would realize increases in revenues far beyond their current valuation.
The major efficiency companies (e.g., Invensys, Integrated Electrical Services, Sempra Energy Solutions, Duke Energy Solutions) and the efficiency equipment suppliers all stand to benefit. So too the copper industry, as more efficient motors use more copper. Since reliability services are at a premium, uninterrupted power supply and storage suppliers would see rapid growth.
The losers are those parties that resist technological change and do not adapt to it. Incumbent distribution utilities would lose significant revenues and have a new class of stranded costs (stranded wires) if they clung to traditional regulation. Regulatory reform-decoupling revenues from kilowatt-hours sold- as practiced in California and Oregon today, would remove the disincentive for distribution utilities.
Generation companies also would suffer major losses, since the penetration of distributed resources acting as virtual peakers will significantly reduce peak power prices. Production-cost modeling of several U.S. power markets shows that just a 4 percent or greater penetration of distributed generation would effectively clip the peak, eliminating price spikes. It's obvious that these virtual peakers directly compete with gas turbine peak power plants. What's surprising is that this shift in the power markets would drop the average revenues earned from a new combined-cycle plant by 10 to 15 percent. The number of hours run decreases by 15 percent as well, and the combined effect lowers the total net free cash flow by 30 to 40 percent. Thus, the distributed generation penetration reduces the profits of new combined-cycle plants so much that they would simply not earn an adequate return on investment. Further, the profitability of utility generation assets, which typically have a mix of coal, nuclear, and gas thermal plants, would be reduced by a stunning 15 to 25 percent.10
It is the fear of these losses that creates resistance from the incumbent players to widespread adoption of distributed power.
In sum, the financial winners and losers from the August blackout depend on which path we take as a nation. If the political consensus galvanizes around the supply-side solutions, we open the door to fundamental realignment of the production of electricity. The distributed resources solutions have the potential to create a more resilient and secure electric power system at lower costs. Political support from this path is absent because it presents disruptive risks to the incumbent players. Yet, to the extent distributed generation moves the energy system toward a more efficient frontier, society wins. Since the distributed path creates a new playing field, utilities must shift their regulatory strategy to adapt.
- A typical vertically integrated utility with 2 million customers may have sales of 40,000 GWh/year, revenues of $5.5 billion/year, but net income of only about 10 percent or $500 million/year (which translates to about $0.013/kWh). If total sales revenue is on the order of $0.10-$0.11/kWh, and fuel and purchased power costs are around 0.04/kWh, then gross margins are high-about $0.06-$0.07/kWh. Direct costs are around $0.03/kWh, indirect costs (overheads) about $0.015/kWh, depreciation and income taxes are in the vicinity of 0.01/kWh, so total costs are around $0.05-$0.06/kWh, leaving only $0.01/kWh in profit.
- Peak Load Management Alliance, "Demand Response: Principles for Regulatory Guidance," February 2002, and Eric Hirst and Brendan Kirby, "Transmission Planning for a Restructuring U.S. Electric Industry," June 2001, pp. 8-9.
- EPRI, "The Western States Power Crisis: Imperatives and Opportunities," June 25, 2001, p. 24.
- Interlaboratory Working Group. 2000. Scenarios for a Clean Energy Future (Oak Ridge, Tenn.; Oak Ridge National Laboratory and Berkeley, Calif.; Lawrence Berkeley National Laboratory), ORNL/CON-476 and LBNL-44029, http://enduse.lbl.gov/Projects/CEF.html.
- Lovins, A.B. , Datta E.K. , Feiler T., Rabago K.R., Swisher J.N., Lehmann A., Wicker K. 2002 Small Is Profitable, The Hidden Economic Benefits of Making Electrical Resources the Right Size. www.rmi.org
- Id. at 362-365.
- McKinsey and Co., "The Benefits of Demand-Side Management and Dynamic Pricing Programs," May 1, 2001, p. 2.
- Jewell, R.W. "Natural Gas: What's Going On: The Double Benefits of Energy Efficiency and Conservation," March 18, 2003, estimates a 5 percent reduction in electricity creates a 25 percent decrease in gas demand, or 1.5 TCF. Current research at RMI estimates the reduction to be as high as 2.1 TCF.
- Elliot, N, Shipley, A.M., Nadel, S., Brown E., Petak K., Blustein, J. 2003. "Impacts of Efficiency and Renewable Energy of the Natural Gas Markets," ACEEE. Estimates reduction of natural gas prices by 20 percent, saving $15 billion/year. Higher estimates from unpublished re-search from the Rocky Mountain Institute.
- Lovins, et al., 2002 .
Articles found on this page are available to subscribers only. For more information about obtaining a username and password, please call our Customer Service Department at 1-800-368-5001.