As Giles Hundleby* explains, the cost of energy from the next tranche of offshore windfarms built in the UK will be lower than previous projects but won’t be as low as some in Europe because the auction model used there is different to those used in the Netherlands and Denmark
The upcoming results of the contract for difference (CFD) auction in the UK will be scrutinised closely, following the record low prices contracted in last year’s auctions in the Netherlands and Denmark and this year’s bids by Dong Energy and EnBW at zero subsidy in Germany.
I can foresee people jumping to conclusions, many of them partly or wholly wrong, and lots of high horses being mounted in supposed defence of the poor, ripped-off UK consumer. So before the madness starts, it’s helpful to take a measured look at where we are, what we might expect to see and what that could mean.
The next UK round is a competition between different developers (or consortiums of developers), each with a different windfarm, bidding for a CFD power purchase agreement. The projects must be operational by 2023 to 2025, and we expect around 2 gigawatts (GW) to be awarded to a maximum of four projects, with the biggest project being between 700 megawatts (MW) and 1,200 MW in size. The CFD includes the cost of the offshore and onshore substations and transmission cables between the two.
BVG Associates (BVGA) analysis suggests that the levelised cost of energy (LCOE) (including transmission system) for the potential UK projects varies between £70 per megawatt hour (MWh) and £80/MWh (€80–92/MWh), based on 6.5% weighted average cost of capital (WACC) and a 25-year project life. Due to the CFD being for 15 years, this would imply CFD bids around £80/MWh. However, if developers pull all of the LCOE levers at their disposal in the way they did in the other recent auctions, volumes could be higher and prices lower.
In contrast, three of the four winning bids in the recent German offshore wind auction for projects coming online in 2025 rejected the comfort of a minimum reference price. By contracting to deliver energy without even a ‘subsidy free’ minimum reference price, the windfarm developers must be expecting to make a better return based on the wholesale market price than they could under a subsidy regime.
BVGA has calculated the LCOE of these windfarms taking into account their characteristics, expectations about new technology and costs from 2025, published information and statements from the winning bidders Dong Energy and EnBW. The use of much larger turbines (13 MW or more) was implied by both bidders. This is in line with expectations of the technology that will just be coming available by the time these projects are being built. The sites are also windier on average than many other sites bidding for offshore wind support.
Treating the two neighbouring Dong Energy sites (OWP West and Borkum Rifgrund 2 West) as a single site, BVGA calculates an average LCOE (in today’s terms and including the cost of the grid connection) of around £48/MWh (€55/MWh) at 6.5% WACC.
The refusal of a government-backed minimum reference price does appear to make the commercial viability of these windfarms as stand-alone projects riskier. The time horizon for them to come online further complicates that risk. Those risks and timescales present an interesting dilemma for all stakeholders for this project. If they turn out to be optimistic on LCOE and/or overly pessimistic on the wholesale price side, will the developers still want to proceed with a potentially loss-making project? In the converse situation, will consumers and governments eventually find that they are paying more than they needed to?
In July 2016, the announcement of the Borssele I and II auction results was greeted by the offshore wind industry with a mixture of astonishment, delight and (from competitors) a little concern. Dong Energy had beaten the €100/MWh LCOE barrier well ahead of the 2020 target that industry previously committed to.
Dong Energy’s bid secured 15 years of SDE+ (the Dutch equivalent of CFD) funding for the project followed by the ability to sell energy at market price for the rest of the project’s life.
Dong Energy’s bid was actually at €72.7/MWh, which implies an LCOE of about €68/MWh (or £59/MWh) excluding transmission. The Dutch Government has already paid for key elements of project development activity and is taking the risk on the transmission connection for which it will add a (fairly modest) €14/MWh to the costs, giving a total project LCOE of about €82/MWh (£71/MWh). That was a dramatic decrease from other levels seen at the time.
The Shell consortium that won last December’s Netherlands auction for an offshore windfarm at Borssele III and IV achieved an even lower bid price of €54.5/MWh, implying an LCOE including transmission of around €70/MWh (£61/MWh).
When Vattenfall’s bid price of €50/MWh for Kriegers Flak was announced, it seemed at the time a crazily low price that surprised much of the industry.
The fact is that, with the same technology and timing, the windfarm costs should be pretty much the same at Kriegers Flak as it is at Borssele I and II (see above). That’s because Borssele’s advantages of being a little bit bigger and a little bit windier are mostly offset by Kriegers Flak being a little bit closer to shore and having a little better wave conditions. Kriegers Flak has to be operational by the end of 2021, and while Borssele could be as late as mid-2020s, we expect it to be delivered earlier than that. Both are in similar water depth, and both bid prices exclude transmission and some development costs. But the LCOE for Kriegers Flak implied by the €50/MWh bid price is around 20% lower than the equivalent LCOE for Borssele I and II – around €68/MWh (£59/MWh).
Some commentators have indicated that Vattenfall will use a next-generation 10+ MW turbine at Kriegers Flak (though having these ready for a project completing in five years’ time is a bit of a stretch) and is banking on reducing operational costs through the project life, but our calculations suggest this explains at most a quarter of the benefit. The explanation appears to lie either in increased revenues, reduced project returns or a combination of the two.
It’s clear that, in all these recent auctions, the developers are using all six levers of LCOE: energy production, annual operating cost, total capital cost, project lifespan, WACC and timing of capital expenditure.
The Netherlands system of doing development and having developers compete for sites delivers benefit in reducing risks (and thus WACC) as well as removing costs from developers.
The German system is an interim measure before it moves to a new system similar to the Netherlands. However, though superficially attractive, zero bids add risk to the process and are not necessarily helpful to ensuring capacity is built and delivers to consumers at lowest cost.
The CFD system used in the UK provides opportunity for consumers to ‘claw back’ subsidies if strike prices are below the market price for energy supply but does not deliver as much competition as the Netherlands system due to developers bidding on different projects.
The winning strike prices in the upcoming UK auctions will likely imply slightly higher LCOE than the other markets, but the UK LCOE will not be grossly out of step after adjusting for timing (and assumed technology), project sizes, wind speeds, depths, distances to port and synergies with other projects. The UK LCOE will be marginally higher due to the less directly competitive situation and the carrying of development risk by each different developer.
A combination of the best elements of the UK and Dutch systems could be the optimum way forward. The sooner the UK moves to the Netherlands-style system while retaining use of CFDs, the more level the LCOE playing field will be across the whole of the European offshore wind market and the more transparent the results. To reap the further benefits will require a clearer definition of the future pipeline of projects so developers can plan accordingly, not project by project as now.
*Giles Hundleby is a director at BVG Associates.