Solar supporters agree that solar electricity—particularly photovoltaic (PV) technology—has the potential to make a tremendous impact on our nation’s energy mix. The disagreement comes when you ask how best to create policies to maximize that impact.
The optimum policy will achieve a number of objectives, including: drive demand for fast development and construction of projects, use public and rate payer funds as efficiently as possible, set appropriate incentive levels, and result in solar being located where it makes the most sense.
So far, no subsidy program or policy worldwide has gotten better than mixed marks in all of these areas. However, regulators in California may be close to cracking the code. The California Public Utilities Commission (CPUC) recently approved what’s known as the RAM (reverse auction mechanism) program, which mandates that utilities procure 1,000 MW of distributed solar generation. While this program has yet to prove a success, its design makes it the first large-scale solar program that has the potential to effectively accomplish virtually all of the major policy objectives.
California borrows elements from feed-in-tariff (FiT) programs that have been used in Europe and Ontario. Such programs give developers a financeable 20-year cash flow, which is critical to attracting the investment dollars required to rapidly deploy solar energy systems. In these programs, a developer will sign a 20-year power purchase agreement (PPA) with the utility company. This gives a project investor a credit-worthy counter party and eliminates market risk, such as the price of energy or tradable credits (renewable energy credit, carbon credits, etc.). Furthermore, these programs encourage developers to locate systems in the most cost-effective locations, by paying based on performance and disaggregating the cash flow from the customer’s retail energy price—a problem with the net-metering system most commonly employed in the United States. Essentially, California’s program implements elements from Germany’s program while modifying the allocation method to improve on its shortcomings.
One of the key innovations of California’s proposed program is the use of the reverse auction as a procurement mechanism. This appears to be the first large-scale use of this procurement mechanism for distributed generation. Virtually all other solar FiT type programs (both in North America and Europe) make use of a fixed subsidy-level allocation with a date-based trigger. In a date-based trigger, developers are guaranteed to receive a subsidy at a given level if they are able to complete their projects by a specified date. The mechanism gives developers a clear deadline to meet project milestones and qualify for a guaranteed subsidy level, and also has the advantage of not allowing developers to act like squatters, which reserve allocations only to sit on them and wait until conditions allow a project to move forward—a prevalent problem in previous evolutions of California solar programs.
However, the big problem with the date-based trigger is that it requires that policy makers correctly predict a workable subsidy level. If the level isn’t set correctly, bad things can happen. If subsidy levels are set too low, then the program will fail and few projects will get developed. If subsidy levels are set too high, then there will be a flood of demand as developers and equipment providers look to take advantage of an over-subsidized market. The date-based mechanism doesn’t easily allow for the government or utilities to cap their exposure. As a result, an over-subsidized market can result in aggregate payouts of subsidy dollars far higher than anticipated or budgeted.
The highest-profile case of these risks happened in Spain during 2008. The Spanish government put a FiT into place similar to the program that had been successful in Germany. They used a date-based trigger, allowing developers who completed their projects before year-end to sign a 20-year fixed-price PPA with the government. Unfortunately the program designers set the subsidy level far too high. The Spanish market grew from negligible size to more than 2 GW in a single year. This resulted in far more solar being developed in Spain than the government had hoped for—or could really even afford. As a result, the Spanish government was forced to completely overhaul the program and essentially default on its obligations by reducing FiT payments to projects already up and running. All these changes brought turmoil in the Spanish and global solar markets, resulting in the Spanish market shrinking back to almost zero before climbing back to a more sustainable level of approximately 500 MW per year.
The elegance of the reverse auction mechanism is that it relieves the pressure on the program designers to set the correct incentive level. Instead the mechanism allows the market to set the incentive levels by forcing developers to compete to participate in the program. The mechanism also easily allows program designers to cap their risk and achieve their goal of developing a specific amount of generation capacity. The only apparent danger here is that it could attract bidders developing highly speculative projects. If participation hurdles—cash deposits, development experience, and demonstration of project viability—aren’t set sufficiently high, then the program will be flooded with projects that will never be built.
This is a problem that California is already facing, with previous efforts to meet the state’s renewable energy portfolio (REP) standard through the development of large utility-scale power plants. Unfortunately, to date, the vast majority of the PPAs signed by the utilities haven’t resulted in renewable energy being produced. If the program designers can avoid this trap, then the program should successfully result in the rapid development of a massive amount of solar energy at the best market prices.
Lastly this program seeks to encourage growth in distributed generation by allowing only projects between 1 MW and 20 MW in size to participate.
Distributed generation can offer several advantages over central plant or utility-scale renewable projects, most notably that distributed projects have a higher success rate, in terms of actually getting built. Also they don’t have the same transmission, interconnection or environmental approval hurdles that large solar farm projects usually face. Utilities can expect a distributed generation developer to start delivering power within 18 months of receiving a signed PPA, compared to the several years that large central station projects take to complete the steps required for permitting and interconnection. As a result, 1 GW of distributed solar generation probably can be interconnected much faster in 1- to 20-MW chunks than 1 GW of utility scale solar in 50- to 500-MW chunks—counterintuitive as that might seem.
One obvious way that the California program could be improved is to directly encourage solar projects on rooftops or parking lots. Such installations offer environmental and community advantages, because they make good use of real estate from which no other additional energy generation or public benefit can be derived.
Ultimately solar’s greatest advantage as a technology might be that you can put it almost anywhere, and therefore don’t need to compete with agriculture, wildlife or wind farms for real estate. Solar has a large footprint requirement, compared to other technologies, but the fact that you can put it in places where other technologies can’t go improves the value proposition of distributed solar.
CPUC’s new program recognizes this value proposition and seeks to make it as economical as possible by combining the German type FiT with a new allocation mechanism. The 1-GW RAM fills a void in existing energy policy by incentivizing the development of distributed generation projects that have a high probability of entering service. If implemented properly, the RAM could serve as an international model for how to successfully and efficiently drive the development of renewable energy.