The IASS has produced what is probably the most comprehensive overview on the results of recent renewable energy auctions yet. The study, which is soon to be published in English (it’s available here in German), thus fills a crucial gap. Policymakers will want to know the findings so they can improve policy design – and the study sheds light on some surprising gaps. Another question is, if the auctions outcomes have been analyzed so little, why are so many policymakers convinced of their greatness?
Anyone – well, any policy geek – who has followed renewable energy news over the past few years has probably noticed the onslaught of record-low prices. In August, Chile broke Dubai’s record for solar, with both coming in below 3 US cents per kilowatt-hour for the first time. But just a few weeks later, an auction in Abu Dhabi came in at 2.4 cents – a 20% reduction below the previous record. The same thing is happening with other renewable energy tech as well; this summer, auctions in the Netherlands produced a price of around 8 US cents for offshore wind power, some 30% lower than the previous record.
Analysts are rushing to claim that the policy of auctions is the main reason for these price drops. BNEF’s Michael Liebreich is one prominent example (start at 5:30 in this lecture, for instance). But while Liebreich shows auctions taking over the planet as a policy tool (in some 60 countries, according to Ren21), the support schemes used are still pretty diverse in the countries with a significant installation rate. For example, 22 countries installed 200 MW or more of wind capacity in 2013/2014, but only 6 of them used auctions as a national support scheme in those years. The case for photovoltaics is similar.
The authors (Benjamin Bayer, Dominik Schäuble, and Michele Ferrari ) don’t intend to provide a comprehensive analysis and comparison of all the countries with auctions for renewables. Rather, they selected four countries with a significant deployment in wind energy or photovoltaics: Brazil, France, Italy and South Africa for those two years.
The biggest two countries for new renewable energy installations, however, are China and the United States. China still uses feed-in tariffs (FITs) as the main policy, whereas the US has a mixture of policies at the state level – auctions for PPAs (power purchase agreements), but also tax incentives for wind in particular, which are closer in spirit to FITs than to auctions, as the IEA’s Simon Müller likes to point out in his presentations. (The IASS’s study did not include auctions in the US because they are not part of any federal policy but are organized at the state or even utility level.)
The authors looked at four indicators:
- realization rates (were projects completed [on schedule]?),
- market concentration (how many winners were there and what is their market share?),
- market participation among small players, and
- auction prices.
The third item may seem unusual to non-Germans. Germany is concerned about the impact of the switch from FITs to auctions on small bidders. SMEs, newcomers, and community projects have been the main drivers in the Energiewende, but Germany is the exception here. The study found that there were few small players (< 50 MW) before and after auctions in Brazil and South Africa, for instance. What’s more, the data from Italy did not even allow for such an analysis.
Project realization rates
The authors encountered data availability obstacles in all categories. For realization rates, the reporting was only good in Brazil, where completion by the deadline was also the lowest at 14%. On the other hand, 89% of the projects are expected to be built (just not all on time). The reasons were also not always part of any flaw in the auction’s design. Grid delays, the country’s domestic content requirement, and the handling of environmental impact assessments slowed things down in Brazil. Policymakers can remedy such matters with good auction design and accompanying rules; however, issues pertaining to individual circumstances (such as bankruptcies and poor project management) also caused delays, and policy design may not help as much there.
In contrast, it was hard to tell the reasons for delays/cancellations in France, where only 44% of the volume had been finished by the time the study was completed, because only the volume is known, not the firms behind numbers. The researchers were thus not really able to investigate conclusively. Note that all of the projects in South Africa were completed on time.
One concern is that auctions will lead to “predatory bidding”: prices so low that (small) competitors leave the market, and the surviving oligopoly will then raise prices. The very low bids in Brazil that drew attention in 2013 have not, however, led to a concentration; the researchers found that the share of the big five fell in Brazil and Italy from 59% to 37% and 33%, respectively. South Africa began with a greater concentration of 80%, but it fell to 65%.
One issue here is how to count what. German wind energy association BWE counted eight winners in the first round of auctions in South Africa and only four in the second, which looks like a 50% reduction. But our researchers point out that all four in the second round were new winners, not repeats from the first round, so the two rounds produced 12 winners in total – and an increase of 50% total winners in the second round. Furthermore, the authors argue that competition remains high; round four was still oversubscribed fivefold. And again, no data on specific firms were available for France. At any rate, the authors say they currently find no sign of auctions leading to market concentration.
Did prices at least come down?
Yes – way down. When adjusted for inflation, they fell in France by 35%, in Italy by 27%, and in South Africa by 55% (wind) and 87% (PV). But they went up in Brazil, nearly returning to the price level of 2009. On the other hand more, whether prices went up or down, the researchers found the main reasons to lie outside of the policy mechanism itself.
The reasons for higher prices in Brazil include currency devaluation of the real vis-à-vis the euro/dollar and the attribution of grid costs to the operators. Likewise, costs came down for both solar and wind worldwide during those years regardless of what policy was in place.
Other factors that don’t fall under the auction’s design also played a great role. For instance, market maturity in a given country brought prices down, such as in South Africa. So did interest rates; the WACC (weighted average cost of capital) plays a tremendous role for wind and solar in particular, which have no fuel costs, so almost the entire expense is faced during construction. Simply put, countries with low interest rates have cheaper wind and solar power – with or without auctions.
Of course, the biggest factor in pricing is also unrelated to policy: resource quality. A site with excellent solar or wind conditions should lead to lower power generation prices. Countries like the UEA, Chile, Brazil, etc. have some of the best conditions in the world, and they are mostly still undeveloped. Countries with more modest conditions – or whose best sites are already used – cannot hope to replicate the low levels reached in outlier cases simply by switching to auctions.
Those who, like Liebreich, celebrate auctions as the main reason why wind and solar prices are dropping should realize the disservice they do. A project’s LCOE alone simply does not tell the whole story – or, as the study itself puts it: “Low prices are not a good way of judging whether auctions are a good policy to promote renewables.”
What does the study conclude, then?
So how awesome are auctions? The outcomes were all over the map: very low on-time realization rates in Brazil but nearly 100% in South Africa, falling prices in South Africa and Italy but increasing prices in Brazil, etc. The outcomes were specific by country and context. The authors thus find a messy situation that is hard to draw conclusions from. “The German Economics Ministry wants to know how to design auctions,” coauthor Dominik Schäuble told me for this report. “Our study basically tells them not to look only at auction design when you want to have good results. A lot of ‘external’ factors are crucial.”
The German government switched to auctions after years of policy makers having trouble chasing after the plummeting price of solar by adjusting FITs. For several years, FITs for solar produced high margins in Germany. “But auctions can only ensure tighter margins if there are enough bidders,” coauthor Benjamin Bayer says. In light of all the external factors, ”the success of auctions is best reflected in the margins of developers, not final prices.”
And here’s the tricky part: no data on profit margins is available. Until such comparisons are possible, the authors say auction prices alone should not be used as a sign of policy success.