DESNZ tacitly concede defeat
Regular readers will be aware that I have been exchanging emails with (a rather reluctant-to-engage) DESNZ about the absurd assumptions they use for estimating the cost of renewables, and in particular offshore wind. My last post on the subject can be found here. This article reports on the latest reply, from Helen Huish of the Correspondence Unit (the Permanent Secretary, Jeremy Pocklington seems not to like to reply in person).
The full text of her letter, and the enclosure, together with my further reply, can be found at the bottom of this post.
Capital costs
I had pointed out to DESNZ that their capital cost assumption, of £1.5m per megawatt of capacity had already been falsified, because Moray East, one of the windfarms that will commission in 2025 had already spent more than that putting in their foundations. The suggestion that their figure is nonsense has been confirmed by the latest accounts for Dogger Bank A, which show a very similar picture. The outturn is likely to be around £2.8m/MW for both developments.
DESNZ’s response to this is to ignore the point entirely. Their latest letter makes no mention of capital costs at all. This presumably means that they accept that their input assumption has been falsified but are unwilling to say so. They have tacitly conceded defeat. The effect of correcting this figure is that their central LCOE figure rises from £44/MWh to £62/MWh (both in 2012 prices), the latter figure being equivalent to £83/MWh in 2023 prices.
Operating costs
I had pointed out that the financial accounts of recent windfarms reveal opex costs of around £150,000/MW/year, and, because those costs increase as the turbines age, are likely to have an average of £200,000 or more by the end of their working lives. DESNZ say they are working with a figure of around £90,000, and I seem to have been labouring under a misapprehension that they had been using a rather lower figure. But even if that is the case, their figures are still grossly understated, and they attempt no further justification.
If we correct add this to the list of required corrections, we are up to over £70/MWh.
Capacity factors
As noted in my last post on this correspondence, DESNZ’s capacity factor, of 61%, was extremely high. Existing windfarms might expect to deliver 40% of capacity or less over their lifetimes. Their previous explanation was that “it is standard for LCOE estimates to be calculated assuming that they operate at their technical maximum”. I had read this as being an assumption that they were using a momentary peak output (at times of high wind), instead of a lifetime average output. This would at least explain where 61% came from. However, in their new reply they explain that their capacity factor figure includes all the low wind days but then also includes an availability factor, to account for maintenance downtime and so on. This can be seen in their underlying model, which shows that 61% is in fact the net of a 64% capacity factor assumption and a 95% availability factor. Which is fine, but puts us back to square one as far as understanding why they think output of future windfarms is going to be so much higher than existing ones. As my previous email noted, larger and taller turbines are unlikely to deliver the kind of dramatic increase the department is claiming.
If we also correct the capacity factor to a still-very-optimistic 50% (net of availability), a more reasonable levelised cost is over £80/MWh.
Strike prices
There is some interesting consideration of strike prices in the latest response. (In this section all prices are in 2012 terms – I have adjusted published figures to make them comparable.)
We know that if the strike price is too low, then no bids will be forthcoming. In the AR4 auction, in 2022, there had been a number of successful bids at £37/MWh. A few months later, in August 2023, DESNZ published its Electricity Generation Costs report, which gave the central levelised cost of offshore wind as £36/MWh. (The lowest cost was £32/MWh.) At around the same time, the Administrative Strike Price for AR5 was set at £44, around the highest cost in the report.
It's worth noting, also, that a CfD at any given price will cover a levelised cost that is rather higher, because the price is fully indexed, while the majority of the costs are sunk, and therefore not subject to inflation. In other words, the AR5 strike price was good to cover levelised costs of £50 or more.
But despite this, nobody was interested! There were no bids for AR5 at all.
So for AR6, in 2023, the ASP was raised again, this time to £73. Now the offer was double the LCOE (central) figure. It turns out that at the time, DESNZ had published a partial explanation of how they came up with such I high strike price, but I had never seen it. Capital costs, in particular, were up 26%, and elements of opex seemed to be up a similar amount. This was put down to “supply chain issues” resulting from Covid. This was a remarkable increase given that the CfDs are already index-linked.
However, a 26% increase is still not enough to explain the ASP. Either way, the market cleared at £59, a mere 63% premium on the levelised cost figure.
Conclusion
The LCOE report and the Administrative Strike Prices very different. When going out to developers – when the truth matters – DESNZ are talking about costs that are much higher than those they use in public-facing documents. It is now surely beyond doubt that the LCOE report is grossly misleading, and the failure of the department to even try to defend their capital or operating cost figures shows that they know it.
I have asked the Permanent Secretary to issue a correction. We shall wait and see.