No Solar Way Around It

Nobody who has paid attention to what's happened to solar panels over the last several decades can help but be impressed. Prices declined an astonishing 75 percent from 2008 to 2012. In the United States, solar capacity has quintupled since 2008, and grown by more than 50 times since 2000, according to US Energy Information Administration data. In 1977, solar panels cost $77 per watt. Today, they are less than a dollar per watt.

So it came as a shock to many and an offense to some to learn that new nuclear plants still cost substantially less than solar. Solar advocates have challenged our recent analysis finding that the electricity from Finland's beleaguered Olkiluoto plant is still four times cheaper than electricity from Germany's solar program, claiming that we cherry-picked cases to make nuclear look good and solar look bad.

It is an odd objection, given that we selected perhaps the most expensive nuclear power plant ever built for our comparison. The complaint is odder still because many of the same critics who accused us of cherry-picking then turned around and, without any apparent irony, cherry-picked small, one-off solar projects as evidence that our analysis is slanted toward nuclear.

The reason we compared the Finnish plant to the German solar program is not just because renewables advocates have long claimed that the two examples prove that solar is cheap and nuclear is expensive. We also compared the two because both projects exist in the real world at significant scale, which helps avoid the cherry-picking problem of overgeneralizing from particular cases. Thanks to generous subsidies, Germany generated 5 percent of its electricity from solar last year — a huge amount compared to other nations. By contrast, last year the United States produced just 0.18 percent of its electricity from solar, according to the EIA.

Some have reasonably asked if there aren't broader surveys of the costs of new solar and new nuclear. There are. Both the International Energy Agency and the EIA have done them, and both find that solar costs substantially more than new nuclear construction.

While those figures represent the cost of the average solar installation today, they don't tell us what it costs for a major industrial economy to scale up solar rapidly, such that it gets a significant percentage of its electricity from solar. To date, Germany is the only major economy in the world that has done so. The costs of Germany's solar feed-in tariff represent the only real world figure we have.

As solar has scaled up in Germany, the costs have declined. But the dynamics are not dissimilar with nuclear. France saw significant cost declines as it scaled up standardized plant designs in the 70s and 80s. The new plant in Finland is a first-of-kind design. Subsequent builds are already showing significantly lower costs. The EPR under construction in France, initiated around the same time as the one in Finland, is expected to cost slightly less. The third and fourth versions of the EPR, currently under construction in China, will be a third the cost of the Finnish plant.

Had we chosen to use the two new Chinese plants, solar would have cost twelve times more than nuclear, rather than just four times more. Of course this comparison would almost certainly have raised further objections that we had compared German apples to Chinese oranges. Yet it turns out that the German solar program has benefited enormously from the scaling up of Chinese solar manufacturing — or in the eyes of the US Solar Energy Association, the US Trade Commission, and the European Union, the outright dumping of solar panels by Chinese firms. Indeed the flood of Chinese solar panels, which take up as much as 80 percent of market share in Europe, has depressed the cost of solar panels by as much as 88 percent according to EU officials.

Surely, if it is appropriate to tout solar cost reductions that have been driven by Chinese mercantilism and industrial policy it is also appropriate to consider the cost benefits that Chinese manufacturing and construction costs are bringing to nuclear ­— even more so given that the vast majority of future carbon emissions will come from places like China, not Finland or Germany.

Our analysis was further biased toward solar over nuclear by not accounting for the high costs of backing up and integrating intermittent solar electricity. Leading anti-nuclear greens, including Bill McKibben and Robert F. Kennedy Jr., note that for a few hours during a sunny weekend day, solar provided 50 percent of Germany's electricity; at the same time, as we pointed out, only five percent of the country's total electricity came from solar in 2012. What that means is that if Germany doubled the amount of solar, as it intends to do, there might be a few hours or even days every year where the country gets 100 percent of its electricity from solar, even though solar only provides 10 percent of its annual electricity needs.

What happens beyond that is anyone's guess. Some say Germany could sell its power to other countries, but this would mean other countries couldn't move to solar since Germany would provide electricity at the same hours it would seek to unload it on their neighbors. Solar advocates say cheap utility-scale storage is just around the corner; in fact, choices are extremely limited and expensive. As a result, analysis by the Clean Air Task Force suggest that integration costs for solar and wind are likely to surge dramatically should renewables rise much above 20 or 30 percent of total electrical generation (see graph below).