This is the first of several articles based on studies of electricity markets in the UK, Europe and the USA which I will be publishing on this Substack over the next 2-3 months.
The world’s first offshore windfarm in 1991 used 450kw turbines. The latest offshore turbines are rated 18MW. So we’ve scaled 40x in 35 years and yet the technology still needs to be subsidised. Yet nobody in the offshore wind industry seems to be making money if you look at the share prices of Vestas, Orsted and the renewable energy funds. It seems to be the frictional costs that are siphoning off the subsidies, landowners, planners, lobbyists, financiers, price insurance, market traders.
To be fair the gap between the average market price and the subsidised price paid for offshore output is much lower than 20 years ago - the earliest commercial offshore wind farms in the UK were North Hoyle and Scroby Sands (both 2 MW turbines) which started operations in 2003-04. Two factors have offset the apparent economies of scale: (a) offshore wind farms have moved to sites with much deeper water and further offshore; (b) to generate 15-18 MW wind turbines have greater hub heights and require more complicated seabed installation as well as stronger vertical structures. The industry as a whole has vastly overestimated the cost benefits of opting for ever larger turbines. That is what has messed up project economics, affecting both turbine manufacturers & developers/operators.
There is another factor. Much of the industry assumed that the cost of capital - crucial for a high capital-intensive industry - would not increase even though interest rates and equity returns had reached historically low levels by the late 2010s. If you expect the cost of capital to be 3-4% real but it turns out to be 5-6% real, that has a huge impact on both finance and the offtake prices required to break even.
It would be interesting to consider what things would be like if we immediately halted all current renewables projects and jumped to a full nuclear solution. Would the inherent inflexibility of nuclear be better or worse than the variability of renewables?
Modern nuclear plants are designed to load follow - usually their output can vary from 50% to 100% of nameplate capacity with little variation in their marginal operating costs. That would allow systems served substantially by such plants to have much lower variation in market prices. Of course we would still be stuck with the inheritance of renewable generation capacity that will last until 2045 or after.
No, but nothing matches the flexibility of gas. It is the unescapable trade-off - either build a nuclear plant at $6,000-8,000 per MW of capacity but low running costs or a gas CCGT at $1,200 per MW but higher running costs + however you cost CO2 emissions. In the USA no technology gets close to the cost of building new CCGTs.
The French have long experience with up to 70% of annual supply being nuclear. There is very little flex in nuclear operation, although there is a seasonal pattern where maintenance and refuelling take place during lower demand months. Flex is provided by hydro (including pumping as with Wylfa/Dinorwig historically in the UK), gas and exports. Adding renewables to the mix has been unhelpful because it has required more flex to balance it resulting in less favourable trade positions. Nuclear plus flex has given them a much cheaper system than ours, albeit we lack the degree of hydro they enjoy.
Correct. As you note, France has much larger hydro resources than the UK to provide flex. However, the EPR and other modern nuclear designs are expected to be able to reduce output by 50% without significant extra costs. Even the older French nuclear designs are able to load-follow to a more limited extent (up to 25% reduction), though this has pushed up maintenance costs. But matching nuclear with intermittent renewables is a marriage made in hell.
For nuclear, being highly capital intensive with very low marginal cost, any reduction in utilisation is a cost because it is not earning a margin to pay interest and amortise loans. Being able to bypass Xenon poisoning (the effect of cutting output in conventional designs is a build-up of Xe139, which is a very efficient neutron absorber with a half life that renders the plant inoperable for ~3 days) reduces the pain - but regular cycling so that utilisation falls to the ratio between average and peak demand - around 60% - increases cost by almost the inverse 5/3rds ratio. Perhaps some back credit for the difference between peak and off peak prices.
The same economics apply to wind farms subject to extensive curtailment (at least if the curtailment tends to zero value, as appears to be happening now there is more competition).
The world’s first offshore windfarm in 1991 used 450kw turbines. The latest offshore turbines are rated 18MW. So we’ve scaled 40x in 35 years and yet the technology still needs to be subsidised. Yet nobody in the offshore wind industry seems to be making money if you look at the share prices of Vestas, Orsted and the renewable energy funds. It seems to be the frictional costs that are siphoning off the subsidies, landowners, planners, lobbyists, financiers, price insurance, market traders.
To be fair the gap between the average market price and the subsidised price paid for offshore output is much lower than 20 years ago - the earliest commercial offshore wind farms in the UK were North Hoyle and Scroby Sands (both 2 MW turbines) which started operations in 2003-04. Two factors have offset the apparent economies of scale: (a) offshore wind farms have moved to sites with much deeper water and further offshore; (b) to generate 15-18 MW wind turbines have greater hub heights and require more complicated seabed installation as well as stronger vertical structures. The industry as a whole has vastly overestimated the cost benefits of opting for ever larger turbines. That is what has messed up project economics, affecting both turbine manufacturers & developers/operators.
There is another factor. Much of the industry assumed that the cost of capital - crucial for a high capital-intensive industry - would not increase even though interest rates and equity returns had reached historically low levels by the late 2010s. If you expect the cost of capital to be 3-4% real but it turns out to be 5-6% real, that has a huge impact on both finance and the offtake prices required to break even.
It would be interesting to consider what things would be like if we immediately halted all current renewables projects and jumped to a full nuclear solution. Would the inherent inflexibility of nuclear be better or worse than the variability of renewables?
Modern nuclear plants are designed to load follow - usually their output can vary from 50% to 100% of nameplate capacity with little variation in their marginal operating costs. That would allow systems served substantially by such plants to have much lower variation in market prices. Of course we would still be stuck with the inheritance of renewable generation capacity that will last until 2045 or after.
Encouraging but still not the flexibility we are currently used to from gas
No, but nothing matches the flexibility of gas. It is the unescapable trade-off - either build a nuclear plant at $6,000-8,000 per MW of capacity but low running costs or a gas CCGT at $1,200 per MW but higher running costs + however you cost CO2 emissions. In the USA no technology gets close to the cost of building new CCGTs.
The French have long experience with up to 70% of annual supply being nuclear. There is very little flex in nuclear operation, although there is a seasonal pattern where maintenance and refuelling take place during lower demand months. Flex is provided by hydro (including pumping as with Wylfa/Dinorwig historically in the UK), gas and exports. Adding renewables to the mix has been unhelpful because it has required more flex to balance it resulting in less favourable trade positions. Nuclear plus flex has given them a much cheaper system than ours, albeit we lack the degree of hydro they enjoy.
https://gridwatch.templar.co.uk/france/
Correct. As you note, France has much larger hydro resources than the UK to provide flex. However, the EPR and other modern nuclear designs are expected to be able to reduce output by 50% without significant extra costs. Even the older French nuclear designs are able to load-follow to a more limited extent (up to 25% reduction), though this has pushed up maintenance costs. But matching nuclear with intermittent renewables is a marriage made in hell.
For nuclear, being highly capital intensive with very low marginal cost, any reduction in utilisation is a cost because it is not earning a margin to pay interest and amortise loans. Being able to bypass Xenon poisoning (the effect of cutting output in conventional designs is a build-up of Xe139, which is a very efficient neutron absorber with a half life that renders the plant inoperable for ~3 days) reduces the pain - but regular cycling so that utilisation falls to the ratio between average and peak demand - around 60% - increases cost by almost the inverse 5/3rds ratio. Perhaps some back credit for the difference between peak and off peak prices.
The same economics apply to wind farms subject to extensive curtailment (at least if the curtailment tends to zero value, as appears to be happening now there is more competition).