The goal of market competition

A recent IZES study discusses specific energy policy models Germany could adopt if it discontinued feed-in tariffs as proposed by 2017. To see what policy design is best, we first have to define the goals. Craig Morris investigates.

Centralised PV

Reverse auctions could lead to more centralized PV and fewer rooftop installations. (Photo by, CC BY 2.5)

Germany is to switch from feed-in tariffs to reverse auctions (see my previous post). The IZES study lists three ways in which auctions are held to be better than feed-in tariffs:

  • they keep costs down because the lowest bidder gets the contract;
  • targets are limits – the market stops when the target is met; and
  • lobby groups for renewables do not have as much influence.

Interestingly, two of those three “benefits” are actually designed to keep the transition to renewables from moving too quickly! Furthermore, those three benefits are not exhaustive; an equally important issue is competition between large corporations and SMEs (not to mention citizens and energy cooperatives); the study calls this goal “diversity among providers.” Corporations have the expertise and resources (both human and financial) to place bids, while smaller entities do not, so auctions could lead to an oligopoly by shutting out small competitors from the outset.

The study puts it this way:

“The Energiewende is more than a mere accounting balance of income and expenditures. It is a societal project for the future requiring a large number of people to take part. People therefore have to be able to and want to take part in its success.”

In placing a bid for a 100-250 kilowatt PV array in France, for instance, the study found that a firm needs “several weeks,” so each bid entails sunk costs – auctions thus potentially produce more losers than winners, and even the firms that win have to include such costs and risks in their bids. The smaller the firm, the less likely it is to even bother to bid.

In Germany, normal citizens and energy co-ops accounted for nearly half of the installed capacity in renewables and a third of the capital invested as of 2012. A switch to reverse auctions would therefore gradually revert ownership back to conventional utilities.


The example of wind power is illustrative. As I recently pointed out (PDF), the American Wind Energy Association provides a list (to paying members) of wind farms including an indication of the owner, which is almost always a single firm – but no such figures are even available in Germany, where ownership is splintered across numerous small investors. (German wind farms are listed by postal code.) Based on data for projects, IZES estimates from 2011-2012 that the average wind farm in Germany has fewer than five turbines with a total of less than 10 megawatts.

At a target of 2.5 GW for onshore wind and an average project size in Germany of 10 MW, Germany would therefore need to have 250 rounds of bidding, roughly one per business day – or, more likely, multiple projects in different locations would be included in collective rounds of bidding.

Of course, small projects below 3 MW or three wind turbines are to be exempt from having to take part in auctions. But wind turbines in particular continue to get bigger. As this comparison shows, the bidding process is likely to fundamentally change the German onshore wind power market; clearly, feed-in tariffs bring about a market that looks much different than the one resulting from auctions. Despite the small size of the average German wind farm, even these projects would be awarded in a bidding process.

Craig Morris (@PPchef) is the lead author of German Energy Transition. He directs Petite Planète and writes every workday for Renewables International.


Craig Morris

Craig Morris (@PPchef) is the lead author of Global Energy Transition. He is co-author of Energy Democracy, the first history of Germany’s Energiewende, and is currently Senior Fellow at the IASS.

1 Comment

  1. photomofo says

    Why not have distributed power aggregators bid into one of these auctions? Or skip the auction altogether and just match the lowest bid after the fact.

    A distributed power aggregator could easily beat the bids of the best solar farm builder. Excess production from end-users is a secondary by-product because end-users extract most of the value from on-site production.

    In the future when you go to install a PV system you sign up on-line with an aggregator to sell your power for whatever the going bid is. Once the allowance for that bid has been used up you hold a new auction and post the new rates.

    You can imagine the bid in these auctions as a new FiT. Can installing PV still be profitable at this new FiT? If you can reach IRRs of 6% then everything should be fine right? Truth is, the bid in an auction isn’t of critical relevance. You have to look at the relationship between self-consumption rates and the FiT to see this.

    The extra net present value of a jump from a 5 cent/kWh bid to a 10 cent/kWh at a 30% self-consumption rate is about 250 Euros per KW of installed capacity.

    If you had a choice between a 5 or 10 cent/kWh FiT on one hand or a 30 or 40% self-consumption rate on the other which would you choose ? Turns out a 40% self-consumption rate gives you an extra 350 Euro in net present value over a 30% rate. This is better than the choice between a 5 and 10 cent/kWh FiT.

    Here’s the pinch. The extra net present value of a jump from a 5 cent/kWh bid to a 10 cent/kWh at a 40% self-consumption rate actually goes down to 200 Euros vs the higher NPV when you were at a 30% self-consumption rate.

    The value of higher FiT rates (or auction bids) gets pinched out of the equation the more you self-consume your power. Obviously there’s some upper practical limit for self-consumption and there’s a cost involved in improving your self-consumption rate via smart appliances and such. You should have these aspects of self-consumption in mind when you debate the transition from FiTs to Auctions.

    We tend to think that self-consumption only helps PV but what if there was a way to rig self-consumption to work with wind? Well there’s actually some potential there. All the smart loads that have to regulate their operation to follow PV output to maximize self-consumption can also be made to regulate their operation to follow Wind output. i.e. The same loads that follow the sun can be taught to follow the wind. A wind aggregator could inform controllable appliances in households/businesses when cheap wind power was available and/or forecasted to be available. The loads would then reserve schedules with the aggregator until the wind availability was used up. May sound overly complicated but it may work. One way to encourage this type of a set up would be to temporarily allow the energy sold by these hypothetical wind aggregators to avoid the surcharge and/or part of the surcharge. If this system of regulating how wind is bought worked it would eventually help bring down the surcharge. This is because these smart appliances wouldn’t just be trained to follow the wind output from aggregators – they could also be trained to follow the wind from legacy plants getting the FiT. If you had loads following wind output you’d lift prices on the wholesale market precisely when wind was available. Lifting wholesale would depress the surcharge.

Leave a Reply