Prognos AG has published a study commissioned by coal interests that concludes somewhat unsurprisingly that coal capacity will be necessary for decades ahead to deal with peak demand. Thanks to this tweet by Karsten Wiedemann for the link.
I recall that the mid term German grid development plan also assumes that coal capacity will stay largely unchanged for the next decade or two.
Anyway, there must be some way to generate sufficient electricity on a windless November evening. I would prefer moving to gas generators, with most of them installed in citizens’ basements (Zuhausekraftwerk) with 90% efficiency, and powering them with hydrogen made in time slots where demand is insufficient to meet the supply from renewable power.
But with this post, I won’t discuss these questions. I would rather like to focus on the very thorough explanation this study gives in its chapter four (pages 7 to 15) for the price mechanisms in the German wholesale electricity market. They do that to explain the well known fact that fossil fuel power stations are having trouble to stay profitable with more solar and wind energy in the mix, which cut into their capacity factors as well as average wholesale prices.
Before doing that, let’s just note that already in 2011, the installed capacity of all renewable energy was at 70 GW, which is close to the sum of coal, lignite, and gas combined (71 GW). It is sure to surpass them this year (page 3).
Now for some basic information about the way the wholesale electricity market works. Everybody bids for delivering electricity in a time slot at a certain price. The bids are successful in the order of their price (merit order). And the highest bid that was successful sets the price for all bidders.
That is an interesting difference to normal contract law. Let’s say solar park A bids at 0.1 cents per kWh, since they have close to no marginal costs. If someone accepts that bid under normal contract law, they pay those 0.1 cents, not some other price based on what gas plant B has bid (5.3 cents) based on their higher marginal cost, including the cost of fuel.
As a consequence, A would get the difference between the price of 5.3 cents and his marginal cost of 0.1 cents to pay for the fixed costs of the capital needed to build the solar park. The study calls this difference “Deckungsbetrag”, which might be translated as “coverage amount”.
That in turn means that getting paid for fixed costs depends on the random factors of how much some other party (in the above example B) is asking, and if demand in the time slot in question is sufficient to make B’s bid successful.
This is a rather weird way of running things. Why doesn’t everybody just bid with the sum of their fixed and marginal cost, as averaged over a year?
This way of running the market also gives a strong incentive for manipulations. If you own sufficient capacities, you might want to hold back some of the cheaper coal production capacity and make sure that gas plants get into the mix.
There have been allegations of exactly that happening in 2009, ripping off consumers to the tune of over 13 billion euro.
This weird way of setting a price might have been adequate for a market where most of the generating capacity is owned by a few large scale utilities. As Prognos remarks (page 9), in such a market it doesn’t matter so much that the highest bid actually gets no “coverage amount” and therefore has no way of paying their fixed costs. It all washes out in the average of the utility owning multiple power plants.
The study then also notes that there is a need to discuss if the present system is suited to a changed market structure. They also note the possibility to have demand play a role for setting the price.
I think the present system is a complete failure and needs to be redesigned from scratch. I think the simplest solution would be to have everybody bid their prices including the fixed costs, and pay everybody exactly what they have bid, not what some other unrelated party has bid. Electricity in the feed-in tariff system which has priority would be assigned some fixed value and be successful even if it happens to be not the cheapest at the time, and the feed-in tariff costs would be calculated from the difference between that value and the tariff, removing the perverse effect that surcharges rise with sinking wholesale market prices.
I am not aware of any other auction where participants get paid on the basis of what some other party has bid, so even if this failed system stays on, it is rather interesting to discuss as a weird deviation from standard contract law.