Offshore wind auctions should not be decided by a lottery

Governments can and should do better

The Danish Energy Agency has announced that it would draw lots to pick the winner of the Thor offshore wind tender, a 1 GW zone off the Western coast of Denmark. Surely we can agree that this is not how a well designed tender should be decided? Beyond the frustration for the losing bidders, who are not selected despite preparing fully compliant and competitive bids, it appears obvious that the competitive pressure was not applied in a way such that there would be only on clear winner, and that means that the result is sub-optimal from the perspective of the ratepayers and/or taxpayers.

The Danish system involves a positive price premium paid to the project in addition to the sale of power to the wholesale market, but imposes a minimum (effectively of zero) for such premium, which all bidders ended up proposing because they all knew that this was their only chance to win the project - because others would also bid that, and indeed they all did.

One simple way to avoid the problem would have been to allow negative premiums, which would have then made it possible for an undisputed winner to emerge. This was the logic behind the "secondary payments" discussed last year in the context of the German auctions, which would have allowed to distinguish between competitors that all went for zero bids. The advantage of allowing both negative and positive premiums is that bidders would have bid the number they were actually confortable with (including possibly positive numbers) rather than the zero number de facto imposed by the auction design. This would be better than the German system, which first imposed the zero bid, and then imposed a further payment by bidders, and better than the current Danish system, which got stuck at the zero bid.

The key conclusion here is that asymmetrical price formulas should be avoided. Any CfD (contract for difference) or tariff that acts only as a floor on the amount potentially paid to bidders will inevitably end up with zero bids and fully merchant projects. Governments may think that this is a great outcome, as they get projects built at no apparent cost to taxpayers (which, as an aside, is not true as development costs are not taken into account and even less true if the grid connection is paid by the grid rather than by projects), but (i) it is not so great from an energy policy perspective, as discussed below, and (ii) it prevents the most competitive bids from actually being selected.

The second item above has been resolved in some countries (like the Netherlands) by using qualitative criteria, but this does not appear to be a very fair solution and it always will end up giving the appearance that lobbying or other non project-related considerations have been given more importance than the value of the proposed project, so this also seems to be an inferior route.

But the most important argument is that governments actually give away the value for taxpayers and the influence on energy markets that they paid for by identifying the sites, conducting the early studies and preparing the permits that make the sites attractive in the first place. A random developer grabs the site, gets a de facto subsidy for all development costs - and in several countries, for the grid connection - and gets to build a project via a merchant route that will bring no protection whatsoever to the population against the volatility of international energy prices. 

Power prices are driven, most times, by the marginal cost producer, which is almost always a natural gas-fired power plant, and such price is therefore highly correlated to the price of gas. (see the graph below showing what it looks like in a market dominated by natural gas - source - a market with more renewables will have prices where spot prices fall below the price driven by gas during some periods, when wind and solar are generating enough volumes, as zero-marginal cost renewables replace more expensive gas-fired power - but the peaks will still be fully driven by natural gas prices)

[I could not easily find a more recent graph, but the point stands]

A fully merchant offshore wind project "rides" that price curve but does not influence it in the slightest. Wind projects are zero-marginal cost and effectively bid zero in the spot market at all times, and take whatever price actually clears the market at that time: they are "price-takers". Developers are thus making a bet that prices are high enough to pay over time for the financing of their initial upfront investment - and if they have raised debt, that they are high enough in each period to pay for debt installments.

Nobody likes to do upfront investment repaid via highly volatile spot prices, but if this is the only way to have projects, people will take the risk and make fancy assumptions about future prices that they can always hope they won't be around to see the consequences of, in order to build new assets (which all developers are desperate to do). It leads to projects allocated to the most bullish (or irresponsible) developers, build at a higher cost than necessary, while still handing out unnecessary subsidies.

Contrast this with a real CfD, where projects have to pay back to the grid or the government the difference between the spot market and the agreed bid price, if spot prices are higher: in that case, taxpayers and/or ratepayers get a benefit from the project in periods of high power (ie high gas prices) because they only paid the agreed strike price for power, not the much higher merchant price. Governments thus get a long term hedge against the international movements of gas prices that drive power prices at home. They also get cheap electricity, as a fixed CfD is very valuable to developers and their investors and helps attract cheaper capital (more debt and cheaper equity) which in turns makes the project's power cheaper.

And a fixed CfD auction can only have one winner - the lowest price, which is thus the market price for long term electricity. We don't call fixed mortgage rates "subsidies" when banks give us such a long term loan; we should similarly see a long term CfD as a market price for a valuable good; fixed price electricity over the long term.

The winner of the Danish lottery will not provide this, unfortunately.