The Resounding Success of German Solar Auctions

The German government has decided to dump the previous simple feed-in tariff system and change it to a model based on auctions. As a first step, the last reform of the Law on Renewable Energy in 2014 introduced this model for large scale ground based solar projects.

In January, a report on the first three auctions under this new model was published. Sigmar Gabriel, the head of the Federal Ministry for Economic Affairs and Energy, called the new model a “resounding success” (voller Erfolg).

Since I was rather skeptical of this transition to an auction model, which was introduced by legislation in clear contradiction to basic values of democracy, I would like to discuss this report. Maybe I was wrong and auctions work as well.

There are several ways to measure the success of the new model.

One is how much new solar capacity it delivers. I recall that there is somewhat of a problem with emitting CO2 and it might be advisable to transition quickly to renewable energy. I also recall that prices of solar are already way down as a result of the wildly successful German feed-in tariff system enabling mass production, so it doesn’t make any sense to step on the brakes because of cost concerns.

The auctions in question here are a miserable failure under this most important aspect. They put a cap on new capacity under this model, set at 400 MW a year average for 2015 to 2017. This cap is included in Article 3 of the Federal Government Ordinance of 6 February 2015 that regulates the details of the auction procedure.

There can’t be more than 500 MW of projects in the first three auctions held in 2015, and no more than 400 MW in 2016, and no more than 300 MW in 2017. The Federal Government decided in its wisdom that it is a good idea to reduce the amount auctioned by 100 MW a year. Considering the urgency of climate action, this alone is already enough to qualify this whole mess as an utter failure.

It is also clear that the auction model is further away from a market model than a feed-in tariff. Under a feed-in tariff, the question of how much new capacity is installed is left to decision by market forces.

I am so old that I can remember these market forces resulting in 3000 MW of solar installed in one month. That was a resounding success, as far as climate policy is concerned. Aiming at 300 MW in one year (2017) is less by a factor of 100. Again, the wrong direction considering the urgency of the climate crisis.

It is anybody’s guess what percentage of the projects that were successful in the auctions actually get realized. Successful bidders have two years to start delivering electricity.

A large scale solar project in Germany is typically built in about two months. There is no need for a two years period.

So the successful bidders will wait for at least 18 months before they buy the first panels for their project, at further reduced prices at that time in the future.

And then some of them might reconsider, go out of business for some reason or other, or fail to deliver the project on time. In that case, they will be required to pay a penalty of 50 Euro a kW of the project, or 50,000 Euro for each MW. That’s in the general vicinity of 5% of the cost of building the project.

But anyway, under this auction system, there is no way to know exactly how much is actually built as a result of an auction until those two years have passed. So the 1.5 GW over three years the Government is aiming for may be reduced further.

The first three auctions were experimental. The second and third auctions had the rule of “uniform pricing” for determining the price, while the first and all other auctions run under “pay as bid”.

“Pay as bid” means exactly what it says. Every successful bid gets paid whatever they bid. So if A bids 7 cents and B bids 8 cents and both are successful, A gets paid 7 cents and B gets paid 8 cents.

In contrast “uniform pricing” means that everyone gets paid the same amount, in the example above 8 cents, the highest bid still successful.

“Uniform pricing” is also adopted as the model for setting wholesale electricity prices. Everyone gets whatever the highest successful bid was. This is also called “merit order“, and it is a pricing model that doesn’t work very well.

The Government report explains on page 8 that some bidders entered extremely low bids of less than 1 cent in the two “uniform pricing” auctions. Such a bid would obviously result in a large loss under “pay as bid” rules. But under “uniform pricing” rules, it makes sense to make sure of a winning bid and hope that someone else bids a more realistic price.

The Bundesnetzagentur recommends using “pay as bid” as a rule. I agree. And while “uniform pricing” doesn’t make sense for these auctions, it also doesn’t make sense for setting prices on the wholesale market (merit order).

The first auction resulted in an average of 9.17 cents. The second and third auctions were concluded at 8.49 and 8.00 cents (page 6 of the report).

It remains to be seen what percentage of projects will be actually realized at these prices. The same is true for the question if those numbers are higher or lower than the feed-in tariffs in place when the projects start delivering electricity.

Things could have been even worse. There might have been no bidders even for these low caps. Or the successful bids might have been at much higher costs.

To sum up, I am not convinced that these auctions have been a resounding success.

But since the Government thinks otherwise, I am looking forward to having them extend this model to the 10 kW home roof market next year. That would be even more fun.