Developing a geothermal project can strike an outsider as a laid-back and low-intensity line of business. Steam, while a useful component in power generation, is not a valuable commodity in its own right. Steam properties do not attract the high-powered geological analysis that oil, gas and mining prospects attract.

"You tend to settle on a suitable site for a project simply by noticing steam coming out of the ground. It's only after that that you start work on drilling wells and estimating capacity," notes one adviser active in the sector. Geothermal power, a time-consuming niche within renewable power development, tends not to reward intensive geological searching.

Beautiful and baseload

But geothermal assets can be extremely lucrative. They produce power 24 hours a day, and so unlike almost all other renewable capacity are considered to be baseload producers. Their operating costs are minimal, although since projects are typically smaller their labour costs are high proportionate to conventional types of generation.

Suitable sites for geothermal projects are concentrated in areas of high geological volatility: California, Nevada, Iceland, Central America, the Andes and the various Asia-Pacific nations located on the so-called "Ring of Fire" of high volcanic activity. While this is a diverse and seemingly uncorrelated group of host countries, all, except for Iceland, demonstrate huge growth in power demand.

Moreover, Icelandic players are poised to play a disproportionate role in the acquisition and development of geothermal projects. Reykjavik Energy, the municipal utility of the country's capital, has one of the best stores of institutional experience with the technology. Glintir Bank counts the financing of geothermal projects as one of its three niche investment banking capabilities, alongside the seafood industry and offshore services vessels.

Two current acquisitions illustrate the depth of investor interest in the assets. ArcLight Capital's purchase of the majority of Caithness Energy's 720MW renewables portfolio included 324MW of geothermal capacity centred on the Coso projects. Citigroup, the buyer's financial adviser, closed a $674.3 million lease bond financing for the portfolio's 240MW Coso and 18MW Beowave projects.

Entering the Ring of Fire

Meanwhile, on 4 December, the Red Vulcan consortium won the bidding for a 60% controlling stake the Philippines' PNOC-Energy Development Corp, a state-controlled, but listed. geothermal developer and operator. Red Vulcan consists of First Gen, a Filipino independent power producer, and Spalmare, a joint venture of Reykjavik Energy and Geysir Green Energy. Geysir's main shareholders are Glitnir, FL Group, Atoka Group and VGK-Hönnun.

PNOC, originally a spin-off from the state oil company, operates 1,198MW of geothermal capacity domestically, equivalent to over 7% of the country's total, and over 11% of the Philippines' total 2006 power production. With the acquisition, Red Vulcan could use PNOC-EDC's roughly three decades of experience to pursue other projects in the region, especially in Indonesia, which is currently fourth place in geothermal production, behind, in order, the US, the Philippines and Mexico.

The buyers are looking to combine available funding from both the buyers' and sellers' financial advisers. ING Capital and the Development Bank of the Philippines were advising the seller, and may provide a bridge of $400 million, while Calyon, Standard Chartered and Bank of Tokyo-Mitsubishi UFJ have already made a $287.5 million facility, priced at roughly 110bp over Libor, available to the buyers.

The auction exceeded the government's expectations, since the Ps58.5 billion equivalent winning bid was well above the reserve of Ps45 billion, and a 55% premium to the price of PNOC-EDC's free float of shares. Losing bidders included groups led by International Power, Filinvest, San Miguel, Avenue Asia, and Aboitiz, while Ashmore prequalified but did not submit a final bid.

Attention now shifts to Indonesia, where levels of expertise are lower, but project potential is greater. In May 2007, Standard Chartered closed a $298 million financing for the Wayang Windu two-phase geothermal project, which included tranches for syndication to both local and international banks. The two phases, each of 110MW, have exclusive rights to develop capacity over a 13,000-hectare area and sell power to state electricity company PLN under contracts running till 2036.

Wayang Windu is notable for two reasons: its sponsor is a local energy group – Star Energy – and it is one of the first financings for an IPP that sells power to PLN since the late 90s Asia crisis. Geothermal projects should be more attractive to local offtakers because of their lower operating costs, but during Indonesia's late 90s financial crisis, PLN rejected several power purchase agreements, including two with CalEnergy-owned geothermal plants.

CalEnergy, which was later acquired by MidAmerican Energy, recovered $290 million in political risk insurance proceeds from Opic and Lloyds after an arbitration process that was a landmark in the project finance market. Opic, together with JBIC, is now considering coming back to the country and sector. According to Medco Energi, the developer, along with Ormat, Kyushu Power and Itochu, of the 300MW Sarulla project, JBIC and Opic are likely to contribute $560 million of the project's $800 million cost. Sarulla is set to come to market in the third quarter of 2008.

Established US assets stretch the market

The US offers a much more stable backdrop, but can still suffer from unsettled power and financial markets. The most attractive areas for development – broadly California and Nevada – are served by utilities with a chequered credit history. But both states offer attractive incentives for renewable developers and both have instituted renewable portfolio standards (RPS), although wind and solar have grabbed much of the market's attention.

The larger geothermal portfolios have excited strong interest in the debt capital markets. Israeli-founded Ormat, for instance has closed project bond issues for two portfolios: the $190 million Ormat Funding transaction, from February 2004, and the OrCal Geothermal bonds, a $165 million issue from December 2005.

Dwarfing these two, however, is the $674.3 million in lease bonds due 2026 that Citigroup closed for the Coso Geothermal portfolio. It priced the issue at 315bp over the equivalent treasuries at the end of 2007. Lease bonds, which usually finance the acquisition and upgrade of thermal power plants, rely on strong, long-dated cashflows, and baseload geothermal plants are well-suited to providing these.

Coso Geothermal Holdings leases the assets from a special purpose vehicle, and pays rent on them equivalent to debt service. The issuer is a pass-through trust, but bondholders are exposed to the credit of the lessor. Fitch rated the bonds BBB-, noting that the deal is highly dependent on payments from Southern California Edison (SCE) for power, since revenues from these agreements make up 95% of the borrower's cashflow.

The assets offer some advantages over gas-fired capacity, since they will not incur any future carbon compliance costs, and are vital to allowing California's utilities, primarily San Diego Gas & Electric, Pacific Gas & Electric and SCE, to meet their looming RPS targets. Coso's 240MW of capacity is vital to achieving this target. The smaller part of the portfolio – Nevada's Beowave – sells power to Sierra Pacific Resources under a long-term PPA.

But the steam field resources are not infinite, and developers have to maintain and upgrade wells over time, a capital-intensive process. Coso's portfolio, for instance, is relying on its Hay Ranch upgrade programme to offset declines elsewhere, and it has run into some difficulty gaining the necessary permits to go forward. The proceeds of the bonds will not only prepay Coso's existing bond debt, and fund ArcLight's purchase of the assets from Caithness, but also provide funding for a long-overdue maintenance programme at the site.

Calpine's Geysers assets, which make it the largest producer of geothermal power in the US, have been at the centre of much of the power producer's bankruptcy court wrangling. Its debtor-in-possession lenders have security over the 725MW of capacity, and many of the proposals to bring Calpine out of bankruptcy have centred on the desire of its creditors to auction off the prized baseload assets. Calpine's latest reorganisation plan, which received a 91% acceptance rate on 12 September, involves it resuming trading as a listed entity. The geothermal assets will be a vital hedge of its primarily gas-fired portfolio.

Tax and blend

Beyond the limited number of established assets financing structures get more creative. The production tax credit (PTC), a staple of the wind industry, was extended to geothermal assets in late 2004. While geothermal power is normally economic without this kind of incentive, the credit does encourage greater equity interest in the field.
Ormat, for one, was quick to take up the offer. While tax equity investors in wind projects have to deal with the variability of wind levels, and thus their revenues, geothermal assets attract a more conservative investor base, and achieve even lower yields than wind.

Geothermal projects have such strong economics that some developers suggest that they may survive the recent revenue procedure from the Internal Revenue Service, which capped the amount of deferred tax equity that outside investors may contribute to a project earning PTCs. Deferred equity contributions, which are contributed over the life of the project as and when tax credits are paid, under a pay-as-you-go (PAYG) mechanism, allow developers to put debt on projects ahead of tax equity, and if this debt is substantially cheaper than tax equity, to realize better returns than an unleveraged PAP (partnership flip) deal.

The recent IRS ruling allows the tax equity providers to defer no more than 25% of their contributions, which has the effect of exposing them to much greater levels of operational risk. For wind projects, the stated targets of the ruling, the demands for increased yield from tax investors, which are already higher for PAYG than PAP, make such a structure even less attractive. Should the ruling apply to geothermal projects, and there is some debate as to the intentions of the IRS, then developers still might find the structure useful, since tax equity providers might accept lower yields on reliable baseload generation. The technology might become the last refuge of PAYG structures.

In June 2006, Ormat completed a $71.8 million tax equity financing of its Steamboat Hills (12MW), Galena 2 (10MW) and Desert Peak (10MW) projects, by selling them to a special purpose vehicle, OPC LLC, in which Lehman Brothers and Morgan Stanley took stakes. The two, which have to date had a smaller presence in US renewables than players such as GE and JP Morgan, are then understood to have raised debt against these cashflows. Ormat is likely to close a $46.6 million follow-up financing with a similar structure in April 2008 for its Galena 3 project, located, like the other assets, in Nevada.

The results of the US Bureau of Land Management's auction of leases, for areas with geothermal activity, surprised casual observers. The BLM, which controls vast swathes of the remote west, brought in $20 million in proceeds, at prices per acre many times what it normally brings in. Ormat accounted for $8.2 million in spending, but several smaller developers, as well as some speculators, have also submitted bids.

But the small number of listed developers is unlikely to expand substantially in the short term, because of the time it takes to develop new capacity. The New York markets host Ormat and Raser Technologies, while Toronto's venture exchange, already home to several junior miners, hosts another five, including US Geothermal and Nevada Geothermal.

US Geothermal struck an agreement in August 2006 with another US investment bank, Goldman Sachs, to provide it with $34 million in tax equity financing for its 13MW Raft River project. Its CEO, Daniel Kunz, says that the financing encompassed both early-stage development and construction funds. Raft River is now online, and Kunz says that the developer is now working to advance its Neal Hot Spring project. Nevada Geothermal signed a $20 million bridge loan with Glitnir to bring its 30MW to 47MW Blue Mountain plant to a stage where it can get long-term financing. Morgan Stanley, with its experience with Ormat's tax deals, is likely to join Glitnir on the take-out.

Raser and margins

Raser recently signed a commitment letter with Merrill Lynch for the debt financing of its 10.5MW Truckee geothermal asset in Nevada. The agreement, which covers $44 million in 15-year debt financing, priced at 500bpm over Libor, is the first in a series that commits Merrill Lynch to 100MW of additional capacity, and the right of first refusal on another 55MW. The Truckee financing would incorporate a leveraged pay-as-you-go tax equity element, with a yet-to-be determined provider providing the class A equity units. These entitle the tax equity provider to 95% of income and tax benefits for the first ten years of the project's life.

Raser Technologies, best known as a developer of new motor technology, began to expand into geothermal power around 18 months ago. It has three projects, including Truckee, all of around the same size at late stages of development. It uses United Technologies' binary generation equipment, which allows generators to produce power from a geothermal resource with much lower temperatures than the existing generation of technology.

The technology is of a relatively recent vintage, and lenders hope that it will be more reliable than the generation of geothermal projects that came online in the 1970s and 1980s and saddled lenders and tax investors with heavy losses. If the new generation of machinery operates as planned, and manufacturer warranties are robust enough to satisfy lenders, then geothermal could take off at a much faster rate than has been possible until now.

Raser promises such a rapid rate of growth in capacity because it is using lower-temperature sites that have been adequately measured but rejected by developers using the older technology. It will not for long be the only developer interested in the properties, as the BLM auctions indicate. But it has a solid head start, and has achieved qualifying facility (QF) status for its projects.

The QF status would oblige local utilities to buy the output from Raser's projects at the utilities avoided cost of power, and also oblige them to allow the project to connect to the grid. But since utilities in the western US are struggling to meet their targets for renewable electricity use, the developer should be able to command a premium over this and a contract length of much more than the 15-year term of the Merrill Lynch debt.

Despite these advantages, project finance banks have been largely indifferent to the sector. At least part of the explanation is a reluctance to repeat the errors of the 70s and 80s, and project banks went through just such an acceptance period with the newer generation of wind technology.

Still another explanation is that the necessary engineering and geophysical expertise is difficult for banks to justify keeping on staff. The nearest analogy is the independent oil and gas sector, where a larger subsector of project finance banks keeps petroleum engineers on staff to evaluate reserves-based deals. Most investment banks and tax equity providers will come onboard when steam resources are sufficiently developed for production risk to be minimal.

But finding the early-stage capital to develop these resources is still a challenge. The two most conspicuous US-listed developers – Ormat and Raser – both have sidelines as technology providers. They can afford to finance some of the early development, and have pre-existing relationships with investment banks. Below this level, in North America, Jacobs & Associates, founded by a former banker at Dundee Securities, and Glitnir, which recently launched a New York office, are set to dominate the sector.