Last week, I had the interesting experience of reading a piece which casts doubt on the future growth of global solar markets, “How Fast Are The Costs Of Solar Really Coming Down?”, by the Breakthrough Institute (BTI) and specifically Alex Trembath.
This was novel for me as I don’t read this perspective often. Most of the organizations that are looking at how solar fits into the future global energy picture, from the International Energy Agency to solar-specific market analysts, are predicting continued rapid growth over the next few decades, though the rate of this predicted growth varies widely.
And so the particular collection of straw men, faulty logic and ignorance of the actual global solar industry and markets that this article represents provides an interesting opportunity to talk about what is actually happening in the solar industry and what we can expect for the future of this technology.
I will deal with this piece in two parts. In the first, I’ll look at the fundamental errors in BTI’s prediction, and in the second, I’ll explore why the BTI would be motivated to lead readers astray on this topic.
Manufacturing costs declines: 36 years, not three
Trembath et al. argue that costs will not continue to fall at a rapid rate in the global solar industry, and that as a result the technology will not see the wide deployment that many are predicting. This argument has several components, so let’s take them one by one.
BTI argues that the current, rapid global fall in component costs is due to Chinese dumping. We in the solar industry are all very well aware that overproduction, centered in China, has led to a dramatic decrease in the cost of PV products starting in 2011. We are very familiar with it, because it has destroyed many companies and made whole sections of the value chain unprofitable.
However, I would like to point out that even before the Chinese took over the bulk of PV manufacturing, roughly in 2010, costs had been falling rapidly for roughly three and a half decades, as illustrated in the following graphic, based on Bloomberg New Energy Finance data. You will note that over-production in China was a factor in only the last three years.
It would be interesting to know what analysis BTI has for the fall in prices over the previous 33 years. I will offer a few reasons: incremental solar technology improvements, improvements in manufacturing processes and, in the last decade, growing economies of scale. Furthermore, there is every reason to believe that the industry will continue to make incremental progress in cost reduction due to these very factors. It won’t be as dramatic as the fall in the last two years. It doesn’t have to be.
Why the rest of the world doesn’t need German soft costs (except maybe the US)
Second, BTI also argues that the recent fall in soft costs in Germany “has not been transferable across national borders”, noting that this cost reduction is largely a result of the nation’s feed-in tariff.
First, a technical point. Germany does not “have the cheapest solar in the world”. The cost of PV varies widely according to scale, and BTI may be talking only about rooftop PV. Later in the article the organization noted the recent PPA for the Macho Springs PV plant at $0.059/kWh, suggesting that this was a one-off. However, a recent PPA to buy power from three PV plants with the City of Palo Alto also came in at $0.069/kWh, and in January the LADWP put the price of large solar plants in the California desert at USD 0.095/kWh. I have not seen any other region with solar PPAs between 5.8 and 9.5 cents per kWh.
(BTI is correct that without the New Mexico state credit the cost would be around USD 0.085/kWh. This is still well below other regions. I am unsure how the BTI came up with its other assumptions about the value of subsidies.)
To return to the main point, this argument about the transfer of Germany’s soft costs is notable for its extreme geographical vagueness. The global solar industry involves dozens of national markets, each of which is different, including very different soft costs at different scales. In many of those markets, notably China and India, the cost of PV at the scale that it is commonly deployed at is as low or lower than in Germany.
In trying to make sense of this statement, I’ll note that in the United States, rooftop PV soft costs have been compared to Germany’s and found wanting. And while, as noted by the article, we have only installed a small percentage of the rooftop PV that Germany has, the US residential market is growing without feed-in tariffs, instead through third-party solar financing arrangements.
I would also like to note that the United States represented roughly 10% of the global solar market at its height in 2012, so its future is not terribly critical for the global industry.
PV already competitive in a number of markets
After noting the errors in BTI’s first premise, the argument about cost declines, one must then look at the logic of the article – that increased deployment is predicated on costs continuing to fall as rapidly as they have in the past few years.
The multiple articles referenced by BTI may or may not have made that argument, and I’ll let the authors cited argue on behalf of their claims. I don’t make that argument, because the cost of PV does not need to continue to collapse at the current rate for the industry to grow rapidly. In fact, as far as component costs, it would be a good thing for all of us in the industry if they didn’t, because they are currently below the cost of production.
PV is already a competitive source of power in an increasing number of nations around the world. Again, the details are geographically specific. In the Middle East and North Africa, the measure is usually not parity with retail electricity prices. Instead, it is the opportunity cost of subsidizing fossil fuel for electricity generation, when these fuels could be exported to the world market. This is likely why Saudi Arabia is planning 41 GW of solar PV and CSP over the next two decades.
Also, the prices of Spanish and Italian feed-in tariffs are not terribly meaningful. The cost of installing PV, even at small scales, is around the price of retail electricity in these nations. The industry’s view of the future of both nations does not involve the feed-in tariff, but rather self-consumption and PV plants that compete at market rates without subsidies. Already Tentusol has announced plans to build a 250 MW PV plant in Spain, without subsidies.
BTI misses Asian markets
However, looking at the future of existing markets in Europe or the United States as the hope for the near-term future of solar is a straw man. China and Japan already are looking like they will be the largest PV markets this year. (Also here).
These nations show no sign of slacking their commitment to PV. Following the first annual review of Japan’s feed-in tariff, payments were reduced only 10%, and is still the most generous of any large nation in the world.
The Chinese market is particularly missing from BTI’s analysis, which looks only at its manufacturing policy, and ignores its significance as a market. The Chinese government has made it clear that it will continue to stimulate demand to soak up the extra output from its factories, in light of US and EU tariffs. There is also the factor that China needs power, and is adding generation of all kinds as fast as it can.
But an even more significant trend is that well beyond China and Japan, the global PV market is becoming more geographically diffuse all the time.
South Africa, India, Thailand, Saudi Arabia, Romania, Chile and dozens of other emerging markets don’t need to emulate the fall in German soft costs to grow. And while some of these markets will fail and others succeed, the overall trend has strongly been that solar is becoming a truly global phenomenon.
Reductionist arguments miss complexity of market adoption
BTI utterly fails to look at how PV fits into bigger picture of power production in different regions and nations. But it doesn’t stop there. The crux of their argument – that there is a direct relationship between continued rapidly falling PV prices and deployment, is far too reductionist to come close to addressing reality.
The cost of PV had to fall to a certain point to become competitive. After that, there are multiple factors and market barriers which effect the speed of uptake of PV in different markets. There is access to capital and finance models – which have been a huge factor in US residential PV markets. There is the need for additional generation which can be rapidly deployed at low up-front cost – particularly in developing nations. There are the “peak shaving” benefits of solar. And most significantly, there are whether the policies of a particular nation or region block the adoption of solar or assist it, and whether particular administrations have the political will to create a supportive policy environment.
In many markets, this does not mean subsidies – it is means correctly accounting for the cost of solar, removing permitting barriers, and enabling a financing environment that works for solar.
So why would BTI take the time to spin such a fantastically misleading assessment? In my next post, I will look at the BTI’s proposal for a technology solution, what we know about who the BTI is, who pays for their work, and why they are doing what they are doing.