A Hot Solar Market

A Hot Solar Market

June 01, 2008 | By Keith Martin in Washington, DC

There are two types of solar projects: photovoltaic, where sunlight is converted directly into electricity through solar panels or thin film, and concentrating solar power or solar thermal projects, where sunlight is directed by mirrors or lenses on to a heat exchange medium, the heat is used to boil water to make steam, and the steam turns a steam turbine. Both markets are growing rapidly.

Photovoltaic projects may involve solar panels mounted on roofs of big-box stores, schools, hospitals and commercial office buildings or on special structures erected over parking lots. Utility-scale photovoltaic projects can involve massive arrays in the desert mounted at angles to face the sun. The following is a transcript of a discussion among the heads of five companies that develop or install photovoltaic systems. The discussion took place at the Infocast “Solar Finance and Investment Summit” in San Diego in April.

The panelists are Karen Morgan, president of Envision Solar International, Jeff Wolfe, founder and CEO of groSolar, Arno Harris, CEO of Recurrent Energy, Andrew Beebe, president of EI Solutions, and Paul Detering, CEO of Tioga Energy. The moderator is Keith Martin from the Chadbourne Washington office.

MR. MARTIN: Working in an industry that is so heavily dependent on government subsidies can be like jumping and hoping someone will hand you a parachute. Solar projects qualify for large tax subsidies. Congress extends the tax subsidies, then they expire, and then it extends them again. Why does it make sense to commit so much effort to an industry that is so heavily reliant on the whims of the government?

MR. DETERING: We have to trust that the government will act sensibly and provide incentives that will remain on the books long enough for the industry to make it on its own. The government needs to do this to reduce our dependence on foreign oil and to encourage a shift from old forms of energy to renewable forms of energy. Photovoltaic solar is still too expensive to compete at parity with other forms of electricity. The hope is that the subsidies will accelerate the day when it can compete.

MR. MARTIN: Andrew Beebe, how much does solar electricity cost per kilowatt hour compared to electricity from fossil fuel?

MR. BEEBE: You mean pre-subsidy?

MR. MARTIN: Yes.

MR. BEEBE: The comparison is unfriendly. But as the interest in solar grows, more people get into the business and the cost curve declines while the cost curves for plants that use coal, gas or other forms of energy stay fairly fixed, giving this industry a really interesting end point. That is a fun thing to go after.

MR. MARTIN: Arno Harris, how much does solar electricity cost per kilowatt hour currently, pre-subsidy?

MR. HARRIS: Without an investment credit and direct state incentives, you are talking about something like 30¢ a kilowatt hour.

MR. MARTIN: Compared to what price for electricity generated from fossil fuels?

MR. HARRIS: At a wholesale level, probably 6¢ or 7¢ a kilowatt hour.

MR. BEEBE: It should be a retail comparison, right? So when you look at the avoided peak cost, the electricity price is more like 15¢ a kilowatt hour. Solar electricity is still double the retail rate.

Jeffrey Immedlt gave an interesting talk recently in which he answered your question. GE does not operate in a market setting. Much of what it does is in areas that are heavily regulated: for example, media, finance, health care. He was talking about why GE got into manufacturing jet engines and turbines, one of the more profitable areas. It was because of government incentives. We are still in the early days of a new industry and such industries are often seeded because the government wants to see them grow and provides inducements for people to get into them.

MR. MARTIN: Jeff Wolfe, you must obviously believe the gap between fossil and solar electricity will close fairly rapidly or you wouldn’t be putting so much time into this market. What do you think will be the time frame?

MR. WOLFE: It depends on a whole host of factors, including the investment credit, state subsidies and the price of fossil fuels. As fossil fuels go up in price, electricity prices also increase. We have done a better job as a country of raising electricity prices in the last few years than we have of lowering PV costs, but either direction is fine. [Laughter] It appears that if enough plates spin in the right direction that we will have grid parity somewhere around 2012 to 2015.

MR. MARTIN: Paul Detering, does that time frame sound right?

MR. DETERING: We are a little bit more optimistic. I think the decrease in cost of PV will come about more quickly than 2012 to 2015. Maybe it is because I come from the technology industry and from venture-backed start ups and know lots of people are tinkering with improvements in photovoltaic technology. To me, the bigger issue is how do we get these new technologies deployed more rapidly.

MS. MORGAN: Let add a couple points here. It is important to see some standardization on the financing side so that we can do more of these projects. That is one of the major hurdles that we have: just getting order around the chaos. We have a slightly different perspective, because our company is very challenged by cost in that we focus on parking structures. Our projects are more expensive than just putting solar panels on rooftops. However, we see an opportunity to build esthetically beautiful environments in and around these solar parking structures. Our focus is also not just the United States. Consequently, we do not feel encumbered by the existence or lack of a US investment tax credit. Part of our mission is to drive wider deployment of solar worldwide because scale brings down cost.

Industry Economics

MR. MARTIN: Andrew Beebe, keeping the focus for now on the United States market, how important is a 30% investment tax credit for these projects? How important are utility rebates? Is there a way to quantify their importance?

MR. BEEBE: It’s pretty binary. Without both of those combined, at least in California where we do 80% of our business, the deals just don’t happen.

MR. MARTIN: Arno Harris, is there a way to quantify their importance to the economics of a deal?

MR. HARRIS: We cannot compete today for mainstream customers that are not necessarily ideologically motivated and are not interested in paying a premium for solar electricity without the investment credit.

MR. MARTIN: Jeff Wolfe, in how many states are there utility rebates?

MR. WOLFE: There are 20 with wildly different designs and patterns.

MR. MARTIN: Describe how they work in California.

MR. WOLFE: There is a commercial rebate, which is performance based and where for every kilowatt hour you generate you get both the retail electricity rate and a performance-based incentive that varies depending which utility service territory you are in. The incentive payment is either 22¢ or 26¢ for every kilowatt hour you generate for the first five years of the project.

MR. MARTIN: And that is a cash payment by the utility to whom?

MR. WOLFE: To the system owner.

MR. MARTIN: Can anyone describe how the program works in New Jersey?

MR. WOLFE: They are still figuring it out. I have been in the New Jersey market since 2004 and the program worked well in 2004 and 2005. They are trying to transition to a solar renewable energy credit market where utilities must purchase a certain number of megawatt hours of SREC credits. Those SREC credits have a capped value of 71¢ a kilowatt hour, or $710 a megawatt hour, but the utilities feel that they can probably purchase them for less. The idea of the New Jersey program, although it is not fact yet, is that a developer will sell his SRECs under a long-term contract to a utility or other purchaser at a price that creates enough value to finance a project.

MR. MARTIN: So the SRECs can be sold separately from the electricity?

MR. WOLFE: Correct.

MR. MARTIN: And for how long a term will the utility buy the stream of credits?

MR. WOLFE: The renewable portfolio standard is a permanent regulation, and they need to continue to buy SRECs for own generation to fill that RPS, and it’s an escalating requirement through 2020. It’s a long-term need.

MR. MARTIN: What other states have good utility rebates?

MR. BEEBE: Hawaii.

MR. DETERING: Oregon has a program that’s interesting, Colorado.

MR. WOLFE: Connecticut, Massachusetts.

MR. MARTIN: Paul Detering, suppose a solar PV project costs $100. The investment credit is $30. The ability to depreciate the project over five years is worth $26. So you have $56 so far covered. How much of the remaining cost is covered by a utility rebate?

MR. DETERING: These numbers are evolving and moving because the rebates, specifically in California, are coming down. That said, we look at it and say it’s roughly a third, a third and a third. A third of the project cost is covered through tax benefits. A third is usually covered by a state rebate. A third comes from the underlying economics of selling electricity.

Insolation

MR. MARTIN: Let’s move to another topic. Andrew Beebe, what is insolation?

MR. BEEBE: Insolation is the amount of sunlight that hits a given space.

MR. MARTIN: And how is it measured? How is it expressed?

MR. BEEBE: There is a theoretical maximum of 1,000 watts per meter squared. We look at it in a given area based on NREL data; it is surprisingly well measured. We assess the financial viability of projects in different regions based on three measures: policy, prevailing electricity rates and the sunlight. We look at them in that order. Insolation is actually third on the list. To echo what Jeff Wolfe said, we are seeing extraordinary increases in electricity rates in California, and we expect that to continue as the state moves away from coal-based power. So Pasadena, the Los Angeles Department of Water and Power and Southern California Edison, just in our little neck of the woods in southern California, have now all in the last three months announced 10+% rate hikes, with additional rate hikes of roughly the same amount expected next year. The trailing average in California for the last 30 years has been 5.8%. The point is that electricity rates are a much more important driver for solar projects than insolation.

MR. MARTIN: Sticking with you, Andrew Beebe, what are the top three states, given that checklist, for solar?

MR. BEEBE: It is an extraordinarily challenging question to answer because the market is such a moving target and, given that a development time frame of six, optimistically, but more like 12 to 18 months on a given project, we have to answer the question — what are the likely places to turn systems on in 12 to 18 months? — if we are talking about a fresh start from today. In the 12- to 18-month time frame, California absolutely would remain number one, but behind that I think it’s a little bit of a crap shoot. Hawaii and Oregon both offer good state tax benefits, but the tax benefits are challenging to use given that there are few companies with large enough tax bases in state to use them. New Jersey, if it works out the kinks in its new program, will be an obvious number two. We are spending a great deal of time in Florida, Texas and Arizona at a legislative level because we think that all three of those states could pop in the next 18 months.

MR. MARTIN: Arno Harris, same top three list?

MR. HARRIS: The southwest delivers the best sunlight in terms of insolation. But to prove Andrew Beebe’s point about the relative importance of sunlight, Germany, which gets half the insolation of most of the United States, is the largest solar market in the world because of the rich incentives.

MR. MARTIN: Jeff Wolfe, come back to insolation. Is insolation so unimportant that you could do a solar business anywhere in the United States? For example, could you do it in Seattle where it rains all the time? [Laughter]

MR. WOLFE: The United States is almost unfairly endowed with sunlight as we have been endowed with oil and coal and wind and almost any resource we want. You look at the insolation, which I prefer to call a radiance map so it that is not confused with insolation, and Vermont has about 30% more sunlight than Germany does. Southern California has about 30% more sunlight than Vermont has. The utility rates in Vermont are 30% to 40% higher than in Arizona. Other things being equal, it makes more sense to build a solar project in Vermont than Arizona. The only place in the mainland U.S. that is like Germany is Seattle, and I believe Seattle is better than or equal to most of Germany. So anywhere in the continental U.S., and even parts of Alaska, are wonderful solar sites.

Conversion Efficiency

MR. MARTIN: Karen Morgan, let me shift gears. What type of technology are you using, and how efficient is it at converting sunlight into electricity?

MS. MORGAN: We use both thin film and solar PV glass. The efficiency is far greater in the glass than the thin film. The biggest challenge with thin film is the efficiency, but we are customer centric. There are sites where thin film is more appropriate. The challenge is financing it. Thin film may be better suited for some microclimates — for example, where there is fog.

MR. MARTIN: How efficient are panels versus thin film in converting the energy in sunlight into electricity? 15% conversion efficiency? 12%? 16%?

MS. MORGAN: Fifteen to 18% for panels and about half of that on the thin film.

MR. MARTIN: Jeff Wolfe, how much has that efficiency percentage changed recently?

MR. WOLFE: I’m a fan of evolution rather than revolution, and it has been an evolutionary change. You know, when I got into the industry 10 years ago, your average crystalline pod was in the 12% range. It is now 15% on a panel basis. There are outliers, obviously. The cad-tel thin film was in the 6% range. I believe it is now approaching 10%. ASI transparent panels were in the 6% to 7% range. ASI panels are still in the 6% to 7% range.

MR. HARRIS: One of those interesting things from the perspective of a developer is we are at the edge of the commercialization of a lot of these technologies. We have seen a lot of upstream investment in manufacturing and comparisons of different technologies in the lab and the new technologies are now starting to move into manufacturing. Developers are at the tail end of the value chain in receiving that technology and deploying it on a commercial basis. The challenge is to get those technologies out of the lab and on to rooftops. We don’t know until they are put into use how viable they are commercially and in terms of the predictability of their performance over a long term, their stability in an exposed environment, and ultimately the creditworthiness of the warranty behind them has as big an impact on whether we will deploy them and how we use them as technical performance.

MR. MARTIN: Andrew Beebe, are the choices between crystalline and thin film? Are they at different ends of a spectrum? Describe the range of possibilities.

MR. BEEBE: For five years, we have been doing large-scale commercial installations in many different locations. For the first four years, these were design-build-transfer contracts where we installed and sold solar systems. In such transactions, there was usually a willingness on the part of the customer to take bigger risks because the customer saw warrantees from manufacturers and wanted to try the latest thing. We wanted to push the envelope.

We are now making the transition — and I would call it revolutionary, not evolutionary — to financing of systems around a power contract with the customer rather than a sale of the project. In fact, 80% of our deals over the next 12 months will be power contract deals, or PPAs.

In a PPA market, developers have a very different view. They want systems that are bankable and financeable over long periods of time. So we are seeing — I don’t know if it is a regression or progression — to a more predictable model of technology. Therefore, to answer your question, we will continue to focus on baseline polycrystalline technologies because they are bankable and they work. And that’s probably a good approach for the industry because we won’t see projects having massive technical or systemic failures across multiple sites.

MR. MARTIN: Paul Detering, have you used thin film and if so, what problems, if any, have you had with it?

MR. DETERING: We have not deployed any thin film. We have proposed projects with thin film. I actually am encouraged and look forward to deploying newer technologies. What I like to say is that we want to be on the leading edge, but not on the bleeding edge of the new technology. We need to work very closely with the sources of capital that back us for these projects and make sure they are comfortable with the economics as well as the reliability of the technology, not just in the first year, but also over the life of the projects to make sure that we can pay back the debt and earn a decent return.

Leases v. PPAs

MR. MARTIN: Karen Morgan, do you use leases with customers or power purchase agreements and whichever one you use, why?

MS. MORGAN: The answer is both, and it is driven by the customer’s situation. PPAs are the choice du jour, but leases work as well.

MR. MARTIN: Why are PPAs the choice du jour?

MR. HARRIS: The difference is in who carries the risk of performance. The reason why PPA financings are becoming more common is today’s mainstream customers don’t want to take the risk of operation or ownership. They really just want to buy electricity by the kilowatt hour.

MR. DETERING: Each customer has to decide what business he or she is in. Is the customer in the business of owning, operating and maintaining equipment on a roof or parking structure or the back side of a hill? Or is the customer in the business of doing something else and letting a power company own, operate and maintain the systems?

MR. MARTIN: I can see you guys are good salesmen. How much difference is there really for a solar panel between leasing it and buying the electricity it produces? There is not much maintenance required.

MR. BEEBE: For the last 18 months, our value proposition is to offer customers three options. The customer can buy the system outright, lease it or enter into a PPA to buy the electricity. If the customer wants a PPA, then we will go out and find the customer the best deal. It is an honest broker approach. Customers really appreciate it. We don’t take anything off the top in any of the three scenarios. So I feel like we are in a position to provide data on the reactions from customers. Every single customer will look at those three options, and everyone will walk from the lease for exactly the reasons these guys are articulating.

MR. MARTIN: Jeff Wolfe, which model do you use and why?

MR. WOLFE: We use what the customer wants, and that’s usually a PPA because customers are interested in buying electricity and not in owning the generating assets. Most manufacturers, hospitals, schools — you name it — have a certain amount of capital dollars, and they want to use those capital dollars in their core businesses and not in what is really a side business.

MR. MARTIN: Arno Harris, you scour the market for potential customers, people who want to put panels on their roofs or on the ground. What’s your business proposition to them?

MR. HARRIS: Our focus is on what we see as a vast underserved market: rooftops that are sitting vacant on leased properties. If you look broadly across the types of roofs that we think are attractive, 40,000 square feet or bigger in solar-friendly states and in utility territories where the electricity rates are relatively high, we see 60% or more of rooftops being unaddressed by a market that has traditionally focused on owner-occupied buildings. Leased properties face a number of challenges that we call the lease barrier and that have to do with the fact that triple net leases create some disincentives between owners and tenants around ownership. They also represent a bit of a challenge from a financing perspective, because unlike, say, a Wal-Mart or a Kohl’s, you have a building in which you may not have a 15-year tenant with great credit that wants to sign a 15-year power purchase agreement. We have focused on developing a set of structural and financial solutions that allow us to make those types of properties financeable by our partner, Morgan Stanley, and we think it opens up a massive opportunity for us.

MR. MARTIN: Paul Detering, what is your business proposition to customers?

MR. DETERING: At the end of the day, it’s pretty simple. Most of the time, but not always, we can show them a lower cost of electricity compared to what they are paying today. The second part is, as they look into the future, they fear an escalating rate of increase in electricity prices. The second thing we offer is the ability to hedge against those future increases. Third, and certainly not least, is the fact that they want to move to green renewable energy and, in some cases, they are willing to pay a premium for that as well.

MR. MARTIN: Karen Morgan, you have a very different proposition. What is it?

MS. MORGAN: I would flip what Paul Detering is saying in terms of how we position our value proposition. It is very much about the esthetics and the beautification of ugly parking lots. As our chief operating officer likes to say: “The entryway to your facility is no longer at the lobby. It’s at the curb cut.” The customer may be able to cover part of the cost of the electricity by charging drivers extra to park under the solar shade.

MR. MARTIN: Jeff Wolfe, you’ve been in this business perhaps longer than anybody else on the panel. How have you seen the value proposition to customers evolve, if at all?

MR. WOLFE: It has gone from being a pure values proposition to both a values proposition and economic proposition. I think what is often misunderstood is that while the financial models need to work for your business, but financial models alone don’t sell a project. We have seen many projects with great finances fall apart because the customer just didn’t want to do it at the end of the day. There had been a movement away from selling on the values, which is how pretty much everything else is sold in this country: value, want, desire. Now things are moving back. It is easier to sell something that has value if it is cheaper.

MR. HARRIS: In our market, our customers really aren’t so much interested in fixing their long-term costs of power. They are much more interested in getting green power and just making sure that they are never exposed to an above-market price. We tend to enter into contracts that offer an indexed rate that rises along with utility rates, but with a discount to the utility rate. The result is very different. Customers who want hedges are trying, in a sense, to capture the option value of the solar power system up front. Our model makes us more of an energy company.

Contract Terms

MR. MARTIN: Andrew Beebe, how long would you typically enter into a lease or power purchase agreement with a customer?

MR. BEEBE: We have never done a lease. We have offered them, but never done them. We actually don’t enter into the PPAs. We do it through partners. They are for 12 years on the low end and 25 years on the high end.

MR. MARTIN: Paul Detering?

MR. DETERING: We look typically at 15 to 20 years, but in some cases as short as 12 years. A short contract makes it harder for the economics to pencil out. The economics look better with longer contracts. Adding to what Arno Harris said, what we are seeing today in the marketplace is a lot of customers who see solar as a hedge against rising electricity prices. We have structured proposals where we essentially provide a floating rate off of some agreed benchmark, but we have not had many takers.

MR. MARTIN: Karen Morgan, how long are your contracts typically?

MS. MORGAN: They run 20 to 25 years for PPAs, shorter for leases.

MR. MARTIN: Jeff Wolfe, what do you do about vacancy risk? If you need the economics of a contract that runs 12, 15 or 20 years, what do you do about the possibility the customer will go out of business or move?

MR. WOLFE: I don’t offer the PPAs myself; we work with other parties to do that. It really comes down to doing credit analysis and creating a large enough pool of projects that any one project doesn’t put the whole program at risk.

MR. MARTIN: Arno Harris, you have vacancy risk. What do you do about it?

MR. HARRIS: Our approach is basically similar to what Jeff described. We look to mitigate that risk through the portfolio of projects that we have operating. We are very careful in the site selection.

MR. DETERING: Let me add to that. The other thing that we do is credit analysis very early in the sales cycle. We look at what the underlying credit of the offtaker is, and we have developed some proprietary tools that we use so that we don’t waste sales time up front.

MR. MARTIN: Does anybody on the panel have a sense of the growth in the PV market? Is there any way to measure it?

MR. WOLFE: A lot of money has been put into the manufacturing side and the technology side of the solar business. Manufacturing will catch up with demand. Somebody will figure out how to stamp the wickets. On the financial side, it is extremely complex financing. A lot of very good people are working on it. We are getting it. Once the financial side gets worked out, we will know what our financial markets can do. The stumbling block is that at some point these things need to go on a roof and we need a roofer, electrician, mechanic, somebody to build these projects. The challenge will be how to scale up to the ability to move from putting megawatts to gigawatts out into the field in a very short order.

MR. MARTIN: Arno Harris, how long do crystalline panels last? What’s their useful life?

MR. HARRIS: Twenty five years is the warranty life. The actual life may be a lot longer.

MR. MARTIN: How rapidly do panels lose value or how rapidly do they degrade in terms of energy conversion?

MR. WOLFE: The theoretical and warranted rate of degradation tends to be about .8% per year in minimum power output. The panels are affected by sunlight and heat. The less sunlight and heat you put on them, oddly enough, the longer they last.

MR. BEEBE: So you guys cover your panels?

MR. WOLFE: Right. It makes them last a lot longer. [Laughter] Even in the southwest where it is very hot and very sunny, you will find maybe a .5% percent degradation or less, and that varies with panel manufacture and technology. Some are finding about a .25% degradation per year.

MR. BEEBE: We model .75% for PPA transactions.

MR. MARTIN: Does the value decline commensurately with the degradation in efficiency of conversion?

MR. BEEBE: One area of continuing discussion is the residual value of these systems. I think that’s a TBD — a to be determined — because, unfortunately, you can’t just look at the degradation curve to model the efficacy of the system seven years from now. It is a more complicated equation that must take into account the rate of technological change and the future interest of the market in solar.

MR. WOLFE: There are also very different residual values depending on whether the system is left in place or whether it is moved and reinstalled elsewhere.

MR. DETERING: The other thing that drives residual value is, what’s the cost of alternatives? If electricity rates are going up, that’s going to drive the residual value of the system, because the kilowatt hours it can produce are now worth more.

MR. MARTIN: Karen Morgan, is it reasonable to expect that you can tear panels off a parking structure and put them someplace else in 15 years?

MS. MORGAN: We have actually come up with removable solar trees so that can move them around. However, the practical answer is I would think the current technologies would stay in place and you would install new technologies alongside them instead of ripping them out.

MR. MARTIN: Arno Harris, any actual experience taking used panels and putting them up someplace else?

MR. HARRIS: A previous business had a customer that ended up selling a building and deciding to move the system from one roof to another and, through that process, we got a pretty good sense of what it takes. With today’s modular, non-penetrating mounting systems, there is nothing physically embedded in the roof structure and the process goes fairly quickly.

MR. BEEBE: We have an interesting residual value proof point, which is that last week thieves went up on a roof of one of our installations and duct taped about a dozen panels together and were repelling off the roof with expensive repelling gear with these panels as the police waited for them on the ground. That suggests there is a rather strong after market in used panels.

Technology Risk

MR. MARTIN: Paul Detering, there are lots of people in Silicon Valley tinkering in garages with new technologies. That suggests there is a fair amount of technology risk in this business. How does that affect residual value?

MR. DETERING: The technology risk associated with the facilities we are deploying today is fairly low, because we are using traditional polycrystalline panels and well understood inverter technology. The interesting thing is, as some of that stuff comes out of garages, how we work with them to be on the leading edge and not on the bleeding edge of that new technology.

MR. MARTIN: Jeff Wolfe, is there a technology risk in this business or do the customers, once they have the panels on the roof, remain happy with them for 15 years?

MR. WOLFE: They tend to be pretty happy. Once we get stuff deployed, it is there, we have a revenue stream, and everything is under contract, so there’s not really technology risk. The only risk is potentially to the residual value. If all of a sudden people are giving away panels, then that decreases the value of the panels on the roof. It does not bring it to zero, because the panels on the roof are already installed and free panels are not. That said, nobody is predicting free panels. People are predicting the value will decline over time and scale and improvements bring cost savings.

MR. BEEBE: I entered the market six years ago, starting by running a tracking concentration company called Energy Innovations, and the product still has not changed. During that period, I watched hundreds of millions of dollars of venture money come into the space with the expressed goal of lowering the cost of PV. In fact, during that period, PV panel pricing actually went up. So I’m thinking there’s an interesting venture-capital-invested-cost-change model that isn’t that pleasant and that the biggest innovations, in terms of engineering, come from the financing side because, over that same period of time, when we started selling PV systems, the thing that most shortened our time to close on a given sale is the power purchase model for financing systems, not the technology.