German Solar: Too Much of a Good Thing?

Until the recent reduction of its feed-in tariff, Germany provided some of the most generous solar incentives in the world. However, based on a statement last week by the head of the German energy agency, DENA, the rapid solar buildup threatens to overwhelm the country's power grid. Stephan Kohler proposed capping the amount of new solar that could be added each year at 1,000 MW, or around 10% of the capacity in place as of the end of 2009, in contrast to the 3,800 MW added last year, and as much as 6,000 MW expected to be added this year. Germany's solar incentives are often held up as a model for others to follow, but that rarely takes into account a growing list of unintended consequences that now appears to include grid congestion at high solar penetration.

The problem that Herr Kohler identified is rooted in the large disparity between the average and peak output of solar panels installed in high latitudes and under Germany's notoriously cloudy skies. The principal consequence of this disparity has been economic: it takes a lot more megawatts (MW) of solar capacity to produce the same output in Germany as in a sunnier location such as Spain, North Africa, or the US Southwest. The German government has overcome this impediment by throwing money at the problem. Until recently Germany had some of the most generous solar incentives in the world--generous enough that Germany accounted for more than half of all new solar installations last year. Even after several rounds of cuts this year, the owner of a new building-mounted solar array can still collect up to €0.33/kWh, equating to $0.46/kWh at the current exchange rate. Under the feed-in tariff system, utilities pass on the extra cost of buying renewable power to ratepayers, and as reported by the German Energy Blog recently, that will add €0.035/kWh ($0.049/kWh) to the average consumer's bill next year. Nearly half of that premium is attributable to solar power, even though it apparently accounted for only about 7% of all renewable power generated in Germany in 2009, because the country is such a poor location for solar power.

On average, every MW of solar capacity installed in Germany generates only about 100 kW over the course of the year. If that were a constant, it would be a lot easier for grid managers to accommodate. But of course that capacity generates nothing at night, while still putting 1 MW into the grid at noon on a bright summer day. That's more than twice the peak-to-average output ratio for solar in a good location in Southern California, Arizona or Nevada. The difference affects how much backup capacity must be available to the grid and likewise how much other capacity must be taken offline as solar output ramps up daily and seasonally. It also determines the nature of that swing capacity. While in a sunny location it might suffice to keep a few "peaking" gas turbines on standby--a role that might even be filled by electricity storage in the future--in a place as un-sunny as Germany it requires substantial capacity capable of running economically for many hours a day, week after week. That doesn't sound like a recipe for replacing German coal-fired power plants (or nukes) with photovoltaics.

Everyone knows solar power is cyclical. However, while I've tended in the past to ignore peak output and focus on the average output of solar in a given location, because that's what determines how much energy is actually delivered over time, the implication of Herr Kohler's comments is that the low capacity utilization inherent in solar installations in northern, cloudy regions amplifies the impact of solar's cyclicality. It's starting to look like the German feed-in tariffs, which were certainly effective as a solar policy in maximizing installations, despite Germany's disadvantages of climate and geography, weren't a very smart energy policy. They've placed too much emphasis on a technology that under German conditions only yields a third as much energy, on average, as the same amount of wind capacity, while still being capable of swamping the grid when the sun does shine. I hope that policy makers and grid planners in such similarly sub-optimal locations for solar as New Jersey and Ontario, Canada are paying very close attention.