The geothermal industry has had unusually bad luck in its choices of financing partners. It uses proven technology and does not suffer from fuel or weather risk, but struggles to attract commercial bank lenders. Most developers are listed on the Toronto Stock Exchange, not so much because their projects are located in Canada, but because the city's investor base is more comfortable with hard-to-quantify resource plays.

Even as once-speculative technologies like solar and wind build up bank followings, geothermal developers still scramble to find construction debt. This stage in the financing chain is the last one for which developers lack dependable options. Equity is available, and assets can be refinanced in a variety of markets, but this poor following in project debt markets creates a huge bottleneck in the development pipeline.

The struggle for capital

Developers have turned, variously, to investment banks for construction debt and tax equity, as well as the private placement market. Goldman Sachs, Morgan Stanley, Merrill Lynch, Citigroup and Lehman Brothers all offered debt and tax equity products, while TCW, Prudential and John Hancock provided long-term debt financing either to refinance, or in place of, investment bank debt.

But European lenders, even those with renewables specialisations, tended to steer clear of the market. German and Spanish lenders, for instance, could bring little home market experience to bear, saw little relationship advantage, and could find little in ancillary business in serving Canadian-listed clients. The only European utility with a presence in US geothermal, Enel, eschews project debt. "I troop up to Andover to see them every six months or so," says one New York-based banker, "they explain how much more cheaply the parent can borrow at, and I troop back home again."

The only commercial banks interested in the sector were those of Iceland. These banks had home-country experience, existing relationships, and impressive amounts of risk appetitie. One of them Glitnir Bank, opened up an office in New York in 2007, and quickly snagged several advisory mandates, including one for Nevada Geothermal, for which it closed a $20 million resource verification loan.

During that period to the end of 2007 the headline deals were Goldman Sachs' $34 million construction debt and tax equity financing for US Geothermal's Raft River in 2006, and the tax equity refinancings of several of Ormat's producing assets with Lehman Brothers and Morgan Stanley between mid-2007 and early 2008, which brought in $118.4 million.

Among the most prominent victims of the crunch, when it came, were the Icelandic banks and US investment banks' taxable profits. Short of becoming reliant on bond insurers for long-dated financing, the geothermal industry could hardly have hitched its star to a less reliable set of institutions.

The chain

The benefits of geothermal make it attractive to utilities. It is baseload, entirely renewable, and can be profitable without the use of tax subsidies. The assets use proven technology, and are predominantly located in the power-hungry states of California, Nevada and Arizona, as well as the slightly more frugal Utah.

However, the industry remains dependent on a small number of equipment suppliers and contractors, one of which, Ormat, is a successful developer and operator in its own right. The rigs that are used for geothermal well exploration are the same rigs used in oil and gas drilling. When oil prices rise, and demand for rigs increases, geothermal developers have to contend with sharply higher day rates.

Most significantly for developers and lenders is the cost and uncertainty involved with discovering a geothermal resource. Drilling accounts for roughly 50% of the cost of a geothermal asset, and banks cannot be certain about the size of a given resource until a large amount of time and capital has been expended on it.

Lenders are now more comfortable with describing the resources of geothermal fields in probabalistic terms, much as they would for wind, oil and gas, or mining. Few institutions, however, have the in-house expertise in geothermal that they do in mining or oil and gas. Wind production can vary from hour to hour and day to day, but geothermal resources' total capacity is only verified by drilling, and can only be expanded with more expensive drilling.

Since geothermal wells decline over time, geothermal assets require ongoing drilling to maintain resource levels. This drilling can be expensive, while there is uncertainty about the results that this drilling will produce, notes Richard Donner, vice-president and senior credit officer at Moody's.

These costs can be a drag on sponsors' resources and their ability to refinance their assets. When ArcLight Capital was looking to refinance the holding company debt for its acquisition of a portfolio of geothermal and wind assets from Caithness Energy in mid-2009, the costs of maintaining the output of the Coso geothermal assets came into stark relief.
According to Moody's B1 rating of the $240 million in debt, which BNP and Citigroup closed for the portfolio in October 2009, Terra-Gen's Coso geothermal assets require $102 million in capital expenditure up to 2011, of which ArcLight has funded $38.5 million, and promised to ccover another $30 million. These sums are on top of its $700 million equity funding for the acquisition of the assets. (For more search "Terra-Gen")

Conventional equity access

During the period of constrained credit, geothermal developers still enjoyed access to equity markets. Magma Energy, for instance, founded in 2008, has raised C$184 million ($174 million) from a series of private placements and an initial public offering on the Toronto Stock Exchange. It now owns 23MW Soda Lake, in Nevada, which is operating at 50% capacity, and 43% of Icelandic producer HA Orka, which owns 175MW. Vulcan Power, which controls a substantial acreage in the Western US, raised $145 million in preferred equity from Denham Capital in July 2008.

US Geothermal, which is listed on both the TSX and the American Stock Exchange, closed a C$10.9 million private placement through Dundee Securities and Clarus Securities in August 2009. Raser Technologies, headquartered in Utah, which develops geothermal plants as well as automotive technologies, raised $23.8 million.from a stock and warrant sale led by Calyon, RBC and JMP Securities.

The most eye-catching debut, however, has been Ram Power. Ram's founders and principals largely consist of former Ormat Technologies executives. It is named after Hezy Ram, Ormat's former North American head of business development. Ram was founded in 2008, and in late 2009, merged with two Canadian-listed developers, Polaris Geothermal and Western GeoPower.

The combination included a 100MW prospect in British Columbia, South Meager, two Californian late-stage projects, Orita and Geysers I, and a Nicaraguan asset, San Jacinto, which was operating on a small scale and in the throes of closing an expansion financing. Ram says it has raised C$180 million in equity capital, and hopes that this will be sufficient to leave it independent of the vagaries of the project finance market.

San Jacinto, however, managed to close its debt financing just before the end of 2009. The senior debt was a $70 million 11-year loan led by the Central American Bank for Economic Integration (administrative agent), FMO (environmental agent) and EDC, each of which provided $20 million, as well as a $10 million participation from Cordiant Capital, with $10 million. FMO also provided a 12-year $7 million subordinated loan.

The closing of San Jacinto capped a protracted process that started long before the Ram acquisition, and involved CABEI, the earliest of the banks to commit, working hard to attract participations from other lenders. Of the leads, FMO was participating off the back of a commitment to Ormat's Olkaria geothermal financing in Kenya. It was also the first large-scale project financing of any size in Nicaragua, involving a total investment of $149 million.

Ram is already looking to close a follow-up financing, to take San Jacinto from 46MW to 72MW, for which it is looking for at least $60 million in A and C loans, as well as B loan mobilisation, from the International Finance Corporation. The $216 million total expansion financing requirement would include the refinancing of the earlier debt.

Desperate debt manoeuvres

If San Jacinto, in virgin project finance territory, was a tough process, US developers have endured worse during the last two years. Raser Technologies closed a $45 million construction loan and $25 million tax equity commitment with Merrill Lynch for its Thermo No 1 plant, five days before Merrill Lynch was sold to Bank of America, and six days before the Lehman Brothers bankruptcy. Nevertheless, Prudential refinanced the construction debt in November 2008 with a $31 million long-term tranche.

Despite the refinancing, Raser mandated Calyon to look at attracting outside partners and in January 2009 took out a $15 million corporate line of credit priced at 10% per year with a group of lenders, of which one was connected to is chairman. In April 2009 it restructured an agreement with United Technologies under which it had committed to buy a fixed number of units, managing to secure the return of a $7 million deposit.

Then, in September 2009, it signed an extension agreement with the Thermo lenders, reflecting delays to the completion of the plant, and the sponsor's decision to apply for cash grant financing for the project rather than production tax credits. This announcement in turn came two weeks after the US Department of Energy rejected an application for a loan guarantee for its East Thermo plant, saying that while it thought the project had fundamental strength, it thought it needed more development.

Raser eventually restructured the financing for the Thermo 1 project, cancelling in the process an ambitious multiple financings agreement with Merrill Lynch. Under that agreement, Merrill had the right to finance 155MW of capacity. With the proceeds of a $33 million cash grant, the tax equity commitment, as well as some of the debt, are largely repaid.

Raser has now formed a partnership with a fund called Evergreen Clean Energy, under which Evergreen will fund drilling at Raser sites, roughly $30 million at each 20MW prospect, in return for project-level equity. Raser has also indicated that it will try and close a prepayment for 110MW of capacity at the Thermo site with the Southern California Public Power Authority. The authority has considerable experience with prepaying for wind capacity, but the arrangement has never been used in the geothermal industry, and few industry participants could see how the arrangement would work. Moreover, any prepayment would need to be used to take out a construction financing. Raser, for all its experimentation the last three years, faces the same challenge of bridging early-stage to permanent financing of its peers.

The only geothermal developer active in the US with the financial flexibility to ignore the construction debt conundrum is Ormat, which brings in sufficient revenues from its existing assets, its technology, and its contracting business to fund construction on balance sheet. When bankrupt Lehman Brothers held an auction of its tax equity commitments, Ormat exercised the right to beat a low-ball offer from Goldman Sachs and buy back some of the tax equity in its own project. Ormat cannot use the production tax credits its project creates, but can book an accounting gain from buying back the tax equity units, and can resell them again when their value recovers. Its competitors do not have that luxury.

It was Ormat's contracting arm, dangling a notice to proceed deadline in front of Nevada Geothermal, that forced the latter developer into closing a richly-priced $180 million loan for its 49.5MW Blue Mountain plant with TCW in September 2008. Nevada Geothermal, which was unable to close a conventional loan through Glitnir and Morgan Stanley in time to make the deadline, paid 14% for its debt, fees of 200bp, and must keep $70 million of the TCW debt outstanding after any refinancing.

The timing was better than Nevada Geothermal could have imagined, because in the months after Lehman's collapse even a 14% loan might have been difficult to find. The developer has since received just over $59.5 million in US treasury grants (Enel, with $60 million, and Raser are the other two beneficiaries so far), which has gone towards prepaying the TCW debt. Nevada Geothermal now says it is working with John Hancock on a $95 million financing, which would fund additional drilling at Blue Mountain and pay down some more of the TCW debt.

Federal largesse

The John Hancock financing would be closed under the Department of Energy's Financial Institutions Partnership Program (FIPP), which is designed to shift more of the burden of project due diligence from the overstretched department to commercial lenders. Since the department has indicated that it will frown on mini-perm structures, institutions like Hancock, Prudential and TCW, which lend beyond the tenors of the commercial bank market, could be major beneficiaries.

The increase in interest in cash grants and loan guarantees indicates the extent to which the geothermal industry has latched on to the US incentive regime. It was only in 2004 that geothermal projects became eligible for production tax credits, and developers have since become adept at using them. Tax equity providers, which were attracted to geothermal's predictable rate of tax credit generation, proved more welcoming to geothermal than commercial lenders, although they never assumed construction or resource risk.

The 2009 stimulus bill allowed geothermal projects to claim the investment tax credit, a one-off credit equal to roughly a third of the value of a project, or a cash grant in lieu of the ITC. Sponsors will need to spend amounts equivalent to 5% of the project's value by the end of 2010 to qualify. Cash grants have established themselves in the market for a number of reasons. Unlike wind, geothermal assets, once complete, often have values in excess of their construction cost, even with financing fees thrown in, and the ITC, which can be claimed against project value, looks more attractive than it might for newer technologies.

The ideal structure, according to Rick Rodgers, managing partner at Montgomery Street Financial, is to use a partnership structure to monetise the ITC. "The developer will need to sell the completed project to a third party to realise this value, and the partnership's deficit reduction account must be structured carefully, but the benefits can be substantial." Still, taking the PTC, and using a tax equity structure, is likely to complicate obtaining a loan guarantee from the US Department of Energy.

Most of the developers at the Geothermal Energy Association's recent finance forum in New York looked to the loan guarantee programme to reduce the impact of construction finance bottlenecks on the market. The department, or even the institutions that act as its agents on FIPP, will have to spend time exploring the geothermal niche before getting comfortable with the risks.

The industry commands impressive political support. Senators from Western states, viewed as a new battleground between Republicans and Democrats, enjoy considerable clout. Senator Harry Reid, leader of the Democrats in the Senate, took time out from the US healthcare reform process, to address the forum, though he mentioned at the start of his speech that he was expecting a call from President Obama. He promised, however to do something about the arcane process by which the Department of Energy calculates credit subsidy costs for loan guarantees.

Nevertheless, both Paul Leggett, a vice-president at Morgan Stanley, and Árni Magnússon, executive director for sustainable energy at Islandsbanki, the name to which Glitnir has reverted, think that project financing will become more available during 2010. The terms and pricing however, have yet to be settled.