How to design an offshore wind auction

As the offshore wind sector struggles with cost increases and the often-hostile scrutiny that this brings, it is worth taking a step back to think about how governments should design auctions to get projects built at the lowest possible cost.

Several of the recent highly publicised project cancellations or delays are directly linked to poor auction design, so it makes sense to avoid or correct certain mistakes.

The goals of auctions are many, and some governments may prioritise some over others, but it is worth listing them explicitly.

• Getting low-carbon power generation capacity built
• Getting it done as rapidly as possible, or according to a long-term strategy
• Getting power at the cheapest cost possible, both at the project level and from a system-wide perspective
• Getting local benefits, whether in the form of jobs or industrial development

Getting capacity built means setting goals and providing a framework that is attractive enough for developers and investors. This necessitates reasonable economics, predictable timelines, and goals that do not change (too much) along the way.

For a capital-intensive industry that spends most of its money upfront and then takes a long time to get its payback, that ideally means rapid development cycles and predictable revenue streams.

In the context of volatile but unavoidable spot power markets, the latter means having access to some form of price stabilisation mechanisms, whether CfDs (Contracts for Difference) or PPAs (power purchase agreements). CfDs are essentially long-term swap contracts that trade volatile short-term prices for a stable long-term one.

In the development phase, this includes transparent regulatory processes to get permits and, ideally, limits to the legal contestability of project permits. At the very least, developers require clear timetables as to how such challenges are resolved to be able to allocate capital with a degree of confidence.

Local benefits will largely flow on their own if there is a pipeline of projects rather than a series of one-offs. The majority of project man-hours are structurally local (operations and maintenance, port activities) so there will always be local activity once projects are built. 

The offshore wind sector deals with very large physical objects, and it is often simpler to manufacture or at least assemble them close to where the projects will be. But the supply chain needs predictable volumes before investing for a specific market, so plans for multiple auctions, under a well mapped schedule, work best.

Joined-up thinking

System costs are largely linked to grid access and a project’s proximity to the local power network. It makes sense for regulators to guide where projects should be built, and to direct the grid operator to undertake the grid upgrades or modifications required to inject the additional power.

It also makes sense to attribute costs arising from projects – the grid connection itself, as well as any local reinforcements of the network to absorb such capacity – to the projects directly. 

The cost of electricity itself will be driven by three things: whether projects bear price risk, how competitive the auction is, and how quickly projects can get built after the price is fixed. The first item was identified above: long term fixed-prices, through CfDs or otherwise, allow developers to attract cheaper capital and therefore significantly lower the cost per megawatt-hour produced.

The second item – auction competitiveness – means providing a predictable path for permitting, grid connection and understanding of the project economics. This requires real competition between parties or projects (more on this below).

Timing is everything

The third item – timing mismatches – is at the heart of most of today’s project cancellations in the UK and the US. Realistic bid prices require developers to know when they can build their projects, and in particular to know that they will have all permits to go ahead with construction when they get their tariff (or within a few quarters, under a predictable schedule).

The early French tenders saw massive delays in permits (due to legal recourse). Similarly, in the US, bidders did not know when (or even whether) they would not get their permits for several years. When this happens, there is no link between the actual cost of building the project and the bid made a long time before: the cost of capital, materials and labour will have changed and the technical and physical parameters of turbines and other equipment will also have moved on.

That means that price bids in such circumstances are at best educated guesses, and at worst cynical bets on the future capacity of the developer to lobby for better terms if circumstances are changed against them (and keep the upside if they have changed favourably).

(Un)co-ordinated approach

The problem is compounded in countries that do two-step auctions – one for site control (seabed leases) and one for tariffs – such as the UK or in the US, where this is compounded by having separate regulators – federal and state – in charge of the two separate processes.

The timing of price auctions matters. It is essential that prices are set only when projects can actually be built.

Developers with deep pockets have shown willingness to pay large amounts upfront for leases in the expectation that only leaseholders can then bid for tariffs. Thus the cost of leases, if similar for all winners in the first (leasing) round of auctions, can be passed on to consumers without a loss of the relative competitiveness of the tariff bids in the second auction. Only those that have paid lease fees can bid for a tariff, and they are in the same position on that cost item.

This means that upfront lease payments to governments are effectively a discreet way to fund government by the developers, which pass costs through to ratepayers rather than taxpayers. And it means that the disconnect between the date of permitting and the date of price setting creates a real risk that projects will never get built.

So, the timing of price auctions matters. It is essential that prices are set only when projects can actually be built. This applies both to power prices and to any lease payment. This can be done either the Dutch way, where site development is undertaken by the regulator beforehand and the auction winner gets a fully permitted site with grid connection. Or it can happen the UK way, where only projects with full consent – obtained by the developers themselves – can bid in the CfD auctions.

Sharing costs equitably

The cost of building the grid connection should be borne by the project, with the ability to transfer the asset to other owners, like under the efficient OFTO mechanism in the UK. And once projects are allocated a tariff, penalties for abandonment should be extremely high – equivalent to, say, the price of two years of power generation – roughly the value of the loss to the system of the inevitable delay such abandonment creates.

If there is a preference for developer-led permitting, allocation of sites should be done exclusively on non-price criteria. This should not favour large players only; experienced small developers have a good track record of building projects and they send much less time lobbying to renegotiate terms. It should also include stiff penalties for abandonment other than for objective permitting failures or obstacles.

Governments should not try to get payments for leases because this will lead to higher power prices. It also creates perverse incentives to bid too aggressively and increases the risk that projects are not built. The Scotwind ‘disposal premium’ system, where the regulator can capture any excess value generated in a transaction done in the early years of development, is a good way to avoid windfall revenues accruing to those who win the early licensing rounds but do not take their projects all the way through themselves.

Long-term commitment

Tariffs themselves should be strictly two-way CfDs with the longest possible duration: at least 20 years, and preferably 25 or 30 years in line with the longer operational lifetime of new offshore wind developments. CfDs should ideally include partial indexation to consumer price index (CPI) as this makes macroeconomic sense and helps attract cheaper long-term capital. Pension funds value the inflation hedge and will offer cheaper equity.

CfD payments should be halted when power prices turn negative, as the UK CfD regime now stipulates and is now done in most other European countries. The reference price – against which CfD top-up payments or reimbursements are calculated – should be based on average prices and not instantaneous day-ahead prices, so that projects bear the capture price risk. 

Permitting – and any corresponding legal recourse – should be streamlined and at the very least kept to strict timelines. I have a preference for the Dutch system, where sites are assessed and pre-permitted by the public authority. But there is a legitimate case for a developer-led consenting process on sites pre–identified by the regulator as this can be faster, cheaper and foster original ideas to mitigate local risks.

Aiming for win-win

My personal favourite for a tariff design would be, at the time a site is fully permitted, for the government to offer a CfD at a predetermined price, say €50 or €60 per MWh indexed over 30 years. Governments could then ask developers to bid an amount (positive or negative) to obtain the site, to be paid at completion.

Such a mechanism would allow governments to obtain upfront payments (which makes for good headlines and politics) and secure stable prices for a long period of time at a level that is transparently competitive. This approach would also allow developers to raise ultra-cheap capital on the basis of stable prices and rapid construction. This way, nobody would bid silly amounts because the revenue side is known and capped, and bidders could easily be held accountable for their price offers.

Overall, a lot of these features are already visible in current regulations for offshore wind auctions, but a single gap can make the whole process inefficient. For instance, the fact that neither the Netherlands nor Germany offered two–way CfDs made otherwise smart systems very unfavourable to the development of the industry, and unnecessarily expensive for ratepayers. Similarly, the early price auctions in the UK and the US have led to unrealistic price bids that will need to be corrected or re–set, as we see today. This gives the impression that the offshore wind sector does not understand its costs and requires massive subsidies, when this need not be the case.

Jérôme Guillet is managing director of SNOW – Sustainable Energy Now and a contributor to E-FWD