Why it happened? Who lost in the bust? Who will survive to build another turbine?
The period from late 2001 to April 2002 witnessed a classic industry shakeout as a result of a merchant power development sector that became too ambitious in its power plant development plans.
A year ago, the merchant sector was moving full steam ahead on a timeline that would have added a whopping 380,000 MW of electric generating capacity to a 750,000 MW base.1
Certainly, these plans could only be deemed optimistic in an industry that has exhibited growth rates in the 2 to 3 percent range.
In fact, the 2000-2001 California energy crisis artificially extended the turbine boom because the state for almost a decade had discouraged new plant development.
But within a period of six months, starting with the Securities and Exchange Commission's (SEC's) Oct. 29, 2001 announcement that it was investigating Enron, plans for almost one-quarter of this planned new turbine capacity had been aborted, and the shakeout continues.
Many have tried to pinpoint the source of turbinemania that seemed to hit almost every energy company throughout the nation.
Spread of new gas turbine development could in retrospect be compared to the spread of tulip bulbs in Europe during the Tulipmania of the early 17th century. In the context of the merchant industry, however, GE was Holland, and we all expected a turbine in our neighborhood.
To those who aren't familiar with one of history's greatest speculative bubbles, during the seventeenth century the Dutch economy was severely disrupted by speculation in tulip bulbs. At its height, single bulbs of rare varieties were sold for as much as $25,000 in today's money. Then, in 1637, speculators took their profits and sold out, which made others nervous, so they sold too, triggering a panic and prices plummeted.
Of course, some have placed the blame for turbinemania and the following cycle of boom, bubble, and collapse of the merchant industry on electric deregulation itself and even on Enron, its biggest promoter. But with only a dozen or so states having moved to full deregulation and the fact that many turbines were located in regulated states, it's difficult to conclude that electric competition was turbinemania's smoking gun.
No, the principal cause of the turbinemania was the newly available "F Class" gas turbine technology advanced by GE (Frame 7FA), Westinghouse (501F), Siemens, and ABB. This new technology dramatically lowered the barriers to entry into the electric generating business.
The technology had an initial cost of one-third that of a fully loaded coal plant and was 100 times more siteable because it occupied a city block-size site, not a thousand or more acres.
In addition, the new turbines cut the Btus/kWh consumed from 11,000 of older displaced coal units to 7,200 and had a fraction of the emissions of a coal plant.
One Virginia developer, seizing the moment, claimed it was simple: "I got out a pipeline and power line map and where they crossed I sited my plant." Financing was expected because consultants "assumed" electricity was going to be high priced [based on the forward price curve] because the marginal producer would be inefficiently burning expensive oil or natural gas, in an old steam boiler. Tired coal units would fade away, chased off by the EPA.
The optimism was palatable. Calpine, AES, NRG, Mirant, and Dynegy stocks surged to near dot-com multiples on the idea that they would be building power plants for many years to meet America's insatiable demand for power as a result of what many believed would be an endless economic boom ().
Furthermore, financing for gas turbine construction was readily available. In fact, banks were lining up to finance projects.
And the merchant generation bubble grew until Enron's collapse essentially pricked it. But the underlying flaws in many of the assumptions that were made of the prospects for the U.S. merchant markets and their longevity were evident two years earlier.
The Early Birds Got The Worm
Joseph A. Schumpeter, the mid-20th century theorist of entrepreneurship and technological change, said that profits accrue to the early birds and are competed down by later entrants.
The pioneers in the merchant plant turbine business confirm Schumpeter. They cut their teeth in an entirely different market: the qualifying facility (QF) market of 1978 to 1992. In that market, they learned about project development, financing, and management of gas turbines (and coal plants), but with constrained risks.
Most QF projects were financed based on utility take or pay contracts that regulated utilities reluctantly awarded as required by federal law-the Public Utility Regulatory Policies Act (PURPA) of 1978 (). Their earlier projects, while they were developed based on utility purchase agreements, were not without risk, as AES demonstrated when its first two projects (at Deepwater in Texas and Beaver Valley in Pennsylvania) failed because they could not recover rising fuel costs at the agreed to power sales price.
This experience may explain AES's early avoidance of the market price risk in the new U.S. merchant environment. In the United States, AES became a toller,2 letting its partner-Williams Energy-take the spark spread risk.
AES's 2001/2002 stock crash followed Enron's, but was driven, in part, by its international exposure, particularly in Latin America. In February, AES decided to exit the competitive electricity supply business. Two other U.S. pioneers, PG&E and Edison International, were also hit hard by poorly performing international projects. Needless to say, in 2001, PG&E, largely due to the California price squeeze it was caught in, declared bankruptcy.
Furthermore, early birds such as Cogentrix, Sithe, LS Power, and Tenaska did well by sticking to U.S. projects and not going public.
In 2001, LS Power sold its plants (5,633 MW) to another ambitious but late arriving merchant-NRG.
One of the few exceptions to the "early bird" trend of eventually selling off its plants or halting its merchant development program was Calpine,3 which became the most ambitious publicly traded merchant company.
Notwithstanding, Calpine, too, was forced to modify its ambitions in response to the supply glut, and due to the high cost of development capital as a result of debt ratings downgrades by ratings agencies concerned that the company was over-leveraged.
Calpine, in January 2002, halted development of 34 projects in "advanced development," which represent 15,100 MW. Nevertheless, Calpine, with 11,100 MW operating in January 2001, continues to construct 15,000 MW.
The Convergence Investors
Convergence was a 1990's fad. The idea was for the fuel owner to capture the "spark spread" by entering the generating business, usually through turbine ownership. However, in the case of Williams Energy and El Paso, the so-called convergence play was achieved by signing tolling agreements with risk adverse owner-operators.
Tolling is an arrangement whereby a party moves fuel to a power generator and receives kilowatt hours (kWh) in return for a pre-established fee ().
Most convergence players are characterized by "deep pockets," or strong balance sheets, by access to natural gas, by ownership of gas pipeline or regulated utilities, and by stable regulated cash flows. But this group has its winners and losers, as well as early and late entrants. The financially strongest among this group will likely be among the vulture acquirers of the weak merchants (), or at least many of their plants.
Duke is the largest convergence merchant developer. Dominion is a later entrant, which is likely to experience lower margins pursuing its Maine to MAIN strategy. PSEG has plants under construction from ECAR to PJM, with a focus in deregulated New Jersey. Dynegy has retrenched, and NiSource and CMS were too late to get a foothold. As noted, El Paso and Williams prefer to trade gas and electricity and toll electricity, rather than build plants themselves.
Meanwhile, among the largest players in the turbine market are the still-regulated investor-owned utilities (IOUs) and member or government-owned monopoly generators who are quietly increasing market share by building a large share of the new turbine crop to serve their captive customers. These players are the public and private power companies, partially listed in Exhibit 3.
Of the combined cycle (CCs) and combustion turbines (CTs) that came online between 2000 and 2002, these IOUs have built 27,902 MW or 32 percent. And these companies, along with the deep pocket vulture investors, will pick up the best turbine projects abandoned by the weaker companies.
The final group (7) is the most vulnerable to the turbine downturn: the pure merchant generators. Some are free standing but staggering: Mirant, NRG (being re-acquired by 74 percent owner Xcel), and Reliant Resources.
These three companies and others have nixed () thousands of megawatts of new turbine capacity and written off hundreds of millions of dollars. Even pioneer Calpine has moved to this group.
Calpine has ceased development of all its projects not under construction, cutting its plans to build 70,000 MW to 25,000 MW, canceling 35 turbines and taking a $168 million write-off in the process.
Other affiliate merchants have re-trenched, including PPL Global and Entergy. In December 2001, PPL Global canceled 2,100 MW, or $1.36 billion in projects, in Pennsylvania and Washington and wrote off $150 million in cancellation charges.4 PPL Global still has 2,150 MW of turbines under development.
On April 11, 2001, Entergy took at least an estimated $1.15 per share write-off on 225 million shares, with details to be announced later. A year earlier, Entergy had sold its rights to 22 turbines to a "special purpose entity ... formed through equity contributions from an unrelated third party."5 Sound familiar? Entergy retained a guarantee "of up to $309 million" due to this turbine "transfer." Apparently this guarantee was triggered in April 2002.
But Entergy is unfazed, announcing: "... we expect to be more a buyer than a builder of generating assets. ... Entergy is in a strong financial position to seize opportunities in the 'distressed' asset market we see developing."6
In addition, TECO is an ambitious merchant with a rich cash flow fueled by utility depreciation and earnings and Section 29 tax credits (of 56 million in 2001). But TECO hired an Enron subsidiary to build its plants. As a result, TECO is facing $63 million in cover payments and $200 million in accelerated payments due to Enron's collapse.
As the shakeout continues, the buyers include AEP, FPL Energy, CLECO, GE Capital, Aquila, and TXU. Leading sellers are AES, Cogentrix, Calpine, NRG, and Mirant. Meanwhile, as shown in Figure 1, collapsing equity P/Es have dried up financing for merchant plants. The bonds of many independent merchants have fallen to junk status.
New Turbine Capacity Factor & Efficiency Data:
Looking To The Crystal Ball
Had the turbine boom not encountered Enron's difficulties, it would have ended as new units failed to achieve the capacity factors needed to achieve the pro-forma financial results lenders had been promised. This metric applies to CTs and combined-cycle combustion turbines (CCGT), but the expected capacity factor is lower for the former peakers than for the combined-cycle units. About 40 percent of the turbines added to the U.S. generation base in 2001 were CTs and 60 percent CCGTs. The capacity factors for the 1999-2001 classes of CCGTs are shown ().
Certainly, capacity factor data can be influenced by start up problems, which have plagued the class of 1999 (six units) for three years.
The classes of 2000 and 2001 achieved respectively 45 percent and 42 percent capacity factors in 2001. These capacity factors are examined on a regional basis ().
The data suggest that merchant CCGTs located outside the markets of New England, Texas, and Arizona-Nevada may have difficulty achieving "bankable" capacity factors. The New England, Texas, and Arizona-Nevada (California) markets generally have steam gas or oil units operating as marginal (the highest cost) units most of the year.
Therefore, in these regions new CCGTs with superior efficiency will be economically competitive most of the time. In contrast, in 2001 some 30 plants located outside these regions achieved an 11 percent weighted average capacity factor.
If this level is sustained, new CCGTs in the following power pools could face difficulty achieving acceptable merchant revenues: PJM, SERC, VACAR, MAPP, MAIN, SPP, and ECAR.
Moreover, another test of CCGT viability is whether new units are achieving the expected high efficiency (low heat rate) levels expected of them (). As a rule of thumb, many analysts take the delivered gas price in $/MMBtu and convert it to a $/MWh power price by a ratio of 0.7 times 10. This assumes a 7,000 Btu/kWh heat rate. Accordingly, a $3.50/MMBtu delivered natural gas price is equivalent to a $24.50/MWh power price. This gets the unit to competitive price levels in all non-coal regions and against environmentally challenged northeast coal units and older, smaller coal units in a few other eastern regions.
The data confirm the potential efficiencies of the new CCGT technology; a 7,500 Btu average heat rate for units operating in the 45 percent capacity factor range appears attainable. But the data also shows that the heat rates of the low capacity factor CCGTs outside the oil-gas steam regions are not impressive and certainly not bankable.
Efficiency is laudable, but is the power needed? Figure 7 shows the range of reserve margins for the NERC regions of the U.S. and the entire U.S. The figure shows that due to turbinemania some regions have projected reserve margins approaching 90 percent, while other regions are below 10 percent. Historical margins have been around 20 percent.
Turbinemania has struck and the shake out is well underway. Like real estate, the early data shows the key to CCGT success is location, location, location. CCGTs need high cost gas and oil units to back out in order to achieve acceptable capacity factors and efficiencies.
Even in Texas and Arizona-Nevada, EVA's data show CCGT saturation is still likely. But outside these regions, the outlook is not good for new CCGTs. Siting new CCGTs in reliance on electricity growth of 2.5 percent per year will not support very many CCGTs. Many players are leaving the field (canceling projects). Whether or not there will be enough demand to bring the turbine market toward equilibrium must be addressed on a regional basis.
- See EVA's "Tracking The Boom of New Power Plants in the U.S.", March 2002, p. 16.
- In the U.S. in 2002 AES has 1,584 MW of new turbine tolling capacity operating, 2,566 MW of merchant capacity operating, and 1,266 MW of merchant capacity under construction. See AES 2001 SEC Form 10-K report, p. 13-14. In February 2002 AES's Board of Directors decided to exit the competitive power supply business. See , February 19, 2002, p. A4.
- Calpine press release dated January 16, 2002.
- PP&L SEC Form 10-K Report, p. 56.
- 2001 Entergy Annual Report, p. 38.
- Ibid, p. 6.
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.