At the Renewable Energy Finance Forum (REFF) – Wall Street conference held in New York City this June, Starwood Energy CEO Himanshu Saxena shocked the room with a statement about five to seven year contracts for solar projects. He further went on to explain that this was translating to longer and longer “merchant tails”, or periods where uncontracted energy will be sold on the wholesale market.
Only one month later Duke Energy revealed that it was buying a 200 MW solar project under construction in Texas which holds a 12-year hedge with a subsidiary of Goldman Sachs. The Holstein plant is one of many projects in Texas that are proceeding without any power purchase agreement (PPA).
These moves towards shorter contract lengths, longer “tails” and alternative contract structures all suggest a move in the direction of merchant power. However, these developments also bring up many questions. Chiefly, is the use of hedges in place of PPAs part of a larger trend, or merely a quirk of the market driven by the expiring Investment Tax Credit (ITC)? And ultimately, can merchant solar emerge as a viable business model?
Shorter contract lengths
Sexana’s comment at REFF-Wall Street spoke to a documented trend: over the past two years PPA terms have plummeted. A list of project PPAs supplied by Bloomberg New Energy Finance (BNEF) shows typical contract lengths falling from 20-25 years in 2017 to 12-15 years in 2019, with a 10- or 12-year term being the lower bound for PPAs reported by the analysts who pv magazine spoke with.
These shorter PPAs are driven by a number of factors, but chief among these is that corporate off-takers are demanding them. Colin Smith, a solar analyst at Wood Mackenzie, describes a “tremendous amount of pressure” from Google, Facebook and Microsoft for shorter terms.
“They want to secure low-cost contracts, but don’t want to be stuck with something 10 years long,” Smith told pv magazine. “Facebook hasn’t been around for 20 years – it’s hard for them to sign a contract that is longer than they’ve been around.”
There are also novel deal structures emerging, and BNEF reports one PPA where Adobe buys the first five years of output from a solar project, and Facebook takes over on year six.
The balance of power
Utilities are also calling for shorter PPAs, and are in a better position to do so than in previous years. Many utilities have already fulfilled their responsibilities under state-level renewable energy mandates, and now are voluntarily procuring renewables. This means they are under less pressure to sign deals.
“There is the possibility that tomorrow they are going to secure a longer-term contract at a lower price,” notes Smith. “The solution to that is securing shorter contracts.”
Adding to this, the utility-scale solar market is starting to grow in states like Florida where utilities are vertically integrated. This means that utilities can build their own solar, and state regulators are increasingly allowing them to pass the up-front costs on to their customers through a process called rate-basing. So when utilities do choose to contract with outside developers, the availability of other options further tilts the balance of power in contract negotiations.
Kyle Harrison, senior associate of corporate sustainability at BNEF, says that these dynamics end up pushing more risk onto developers. “[Offtakers] want developers to take the risk,” explains Harrison. “They don’t want weather risk and they want shorter contract lengths.”
And to build any solar plant you first have to fund it, so the dynamics on the investor side are also shifting the balance of power. As investors get more and more comfortable with solar as an asset class, the “wall of money” looking for solar projects to invest in grows. And with more investors looking to put money into fewer projects, investors can’t afford to be too particular about contract lengths, so they are also taking more risks.
Hedges and other models
Beyond shorter contracts, there are changes to the kinds of contracts that solar projects are securing. Specifically, hedges are increasingly becoming a substitute for PPAs in the Texas solar market, with not only Duke’s Holstein project but also the 240 MWac Misae solar project in Childress County holding hedges instead of PPAs.
These hedges provide some level of security for investors, by providing a floor for prices. But while this is not the same as a PPA, in that sales into the wholesale market form the basis for revenues for asset owners, it is also not the same as pure merchant risk.
Starwood’s Saxena estimates that there are a few gigawatts of projects with hedges in the Texas market, but also says that there are limitations. “There are challenges as to whether these deals are going to get done or not,” Saxena told pv magazine. “It’s not like there is an unlimited market for financial hedges. Each hedge needs to be pricing a little bit lower than the one before.”
Saxena says that while there are a lot of developers in the advanced stages of planning projects backed by hedges, many of them are having a hard time securing capital. “The challenge comes when you are looking at hedge prices in the low 20-dollar range,” explains Saxena. “A lot of investors like us who are evaluating these projects are having a hard-time making the numbers work.”
He also notes that for all the projects under development in Texas with hedges, few have actually begun construction.
But while hedges have become a popular substitute for PPAs in Texas, they are not the only potential form of alternative contract structure. BNEF’s Harrison notes the emergence of proxy revenue swaps in the wind market, where an insurance company takes the risk on the volume of electricity generation and thus the weather, and the settlement unit is the revenue.
“Insurers are more willing to take a gamble on weather risk for wind projects, as they have more insight into historic generation,” says Harrison. “The main differentiator is that you are locking in some portion of your revenue, and it is a lot easier to secure funding.”
While proxy revenue swaps have not yet caught on in solar, there are many ways to guarantee a certain amount of revenue, and to allocate risk between the various parties.
South of the border
There have been merchant solar projects both in the United States and Latin America, but the first wave of these projects largely ended in disaster. Solar projects which took advantage of high hourly prices at certain nodes in northern Chile were later unable to repay their lenders when more solar came online and power prices fell. This led to a scandal in the United States as they were backed by loans from the Export-Import Bank and Overseas Private Investment Corporation (OPIC).
Additionally, First Solar had to write down losses on its Barilla project in Texas, which it built to test merchant solar; since then the project has secured a PPA. And merchant solar projects have also re-emerged in Australia and Brazil, Chile and Mexico. This has been going on for several years in Chile and Mexico, but the projects in Australia and Brazil appear to be more recent.
Saxena says that the turn to merchant solar in Mexico is the result of policy changes, and the difficulty in securing even 10-15 year PPAs in that nation. “It seems like people will build it now, and try to sell the power over time,” he observes.
Beyond the compelling dynamics of mid-day power pricing, there is another reason that so many projects are moving forward in Texas, PPA or no: the ITC. This federal incentive falls from 30% to 26% at the end of this year, to 22% at the end of 2020 and 10% at the end of 2021. This means there is a big financial incentive to get steel in the ground in order to claim as much of the ITC as possible.
Both developers and financiers are able to take more risk if they can secure the tax equity portion of the financing. “Some developers are saying – I’ll take a $20 hedge in order to get tax equity,” says Saxena.
He goes on to note that these are risky bets, but also expects the use of hedges in the Texas solar market to be a temporary phenomenon. “The market is acting because of the transient nature of the tax credits,” he states. “If the tax credits go away these short hedges should go away with it.”
Future of merchant?
The analysts who pv magazine spoke with agree that it would be very difficult to move to a market where projects are built on a merchant basis in the United States, and don’t expect solar projects to be built without PPAs in the near term in states other than Texas. However, the shorter length of contracts means that unless they can secure additional contracts or extensions, asset owners are going to have merchant projects on their hands after PPAs expire.
This will often be after much or all of the financing has been paid off, but still the use of hedges means that both project developers and financiers are getting a taste of using models other than long-term PPAs to get solar projects off the ground.
None of this is happening in a vacuum; as the price of solar continues its historic fall and as financiers get more and more comfortable with solar as an asset class, hedges and other models may end up spreading. But while we aren’t sure what exactly will replace long-term PPAs, the era when a 20-year power contract was needed to build a solar plant is over, and it isn’t coming back.