David, As I understand this article, the annual NZ conference cost to 2030 is circa £26bn - how does that reconcile with the possible £50bn cost suggested by your post of 5th January? Guidance appreciated…
David, I found Figure 9 particularly interesting. You explain the % of the producers’ revenues accounted for by CfD subsidies (56%, offshore wind; 48% onshore wind; 46% solar). Do you know whether renewable energy schemes enjoy taxpayer-assistance on decommissioning costs as do, say, operators of oil & gas fields?
…but, as I understand it, N Sea oil attracts extra taxpayer support which is why, I gather, Miliband’s mad decision has/will accelerate circa £25bn of costs.
Hi David. Thank you for the analysis. Could you do a future article on curtailment and how it is funded/reported on?
Presumably the generator either gets the CfD price and a curtailment bonus or the curtailment payment must just be above the CfD price, presumably it is the latter to be a market solution.
As more wind and solar comes online presumably this cost will increase substantially.
If a wind farm is curtailed there is no CFD payment (or equally, no ROCs earned). CFDs and ROCs only pay out against actual generation. Curtailment payments are handled as a separate item by NESO, the grid operator, as part of the Balancing Mechanism. All generators submit prices at which they would be prepared to increase or decrease generation by a given amount. The control room looks at these bids and selects the cheapest way to adjust the generation portfolio to keep demand and supply matched while also meeting a variety of grid constraints (notably with wind generation, a shortage of transmission capacity between Scotland and England, which can result in other generation in England being fired up to cover the wind that can't be delivered because of the constraint). Constraint payments are paid as bid, not at a market clearing price, partly because some really only apply to handle a local problem rather than a national one which can lead to a large premium when NESO have no other real alternative - something that has become contentious with OFGEM investigations as to what rules they might impose.
Part of the problem with this is that these bids are driven by the subsidies that the wind farms would receive if they were to continue generating. Loss of subsidy is a big loss of income, so the most subsidised wind farms do not bid to curtail, which means that the consumer still gets to pay the full subsidy for all the costliest generation. Wind farms expect to be compensated for their loss of subsidy, plus any loss (or less a gain if the market price is negative) against market prices plus an amount for the insult (like a Bridge bonus).
The entitlements to subsidy vary significantly according to the terms under which the wind farm is operating. Those on ROCs may get anything from 0.7 to 4 ROC/MWh, with each ROC currently worth around £72, so ~£50-300/MWh. There are a number of wind farms that are operating on an unsubsidised basis (particularly those that bid aggressively low and unrealistic CFD strike prices which are waiting as long as possible before commencing the CFD and hoping that inflation will up the strike price to a more economic level in the mean time). All the AR3 wind farms were granted a judicial extension of the period before commencing their CFDs because of covid lockdowns delaying their progress.
There are a variety of terms under CFDs. The earliest CFDs pay out with a ceiling of the strike price when the day ahead market price is negative (IMRP is an hourly price). These wind farms only curtail when it is an operational imperative, and demand the full strike price to do so. The next tranche is on broadly similar terms, except with the wrinkle that if day ahead prices are negative for six or more contiguous hours (usually overnight) then no compensation is paid for the entire period of negative prices. This led to market games with prices being negative for 5 hours, then bid up to just positive for an hour, before being allowed to relapse below zero if necessary for the 7th hour onwards, thus ensuring that they maintained their subsidy throughout. When this gaming was discovered (I was actually the one who wrote to OFGEM with the evidence), the rules were changed for later CFD contracts so that any hour with a negative price pays no CFD compensation, so these wind farms then have the incentive to bid for curtailment payments.
That is actually not a sensible solution, because it means that all the wind farms with low strike prices will remain first in line to curtail, which means they will suffer the lion's share of curtailment on top of only getting a low uneconomic strike price when their CFDs come into operation. Rising curtailment volumes has seen increasing competition among wind farms who get no subsidy in surplus conditions, and so curtailment payments have eroded rapidly, as can be seen here:
The result is that they risk bankruptcy in due course unless other ways are found to subsidise them - which would leave us with a sudden shortage of capacity that DESNZ and OFGEM and the CCC simply haven't even thought about.
Curtailment volumes are now at the stage where they are increasing quadratically with increases in capacity: hours already in surplus get bigger surpluses, and there get to be more hours with surplus with each capacity addition. This effect eventually saturates as the extra capacity does very little to solve the problems of low wind hours, while simply piling up more surplus in hours already in surplus. The result is that a new wind farm could expect to see very little of its output earning a reasonable income: the effective price at which it must sell to cover its investment starts to multiply. If Miliband were to achieve his 2030 target a new wind farm would need perhaps 4-5 times the price to cover this. Something I explained already to the Energy Select Committee in written evidence several years ago now.
They seem to think that this will be solved by storage, but of course that adds enormously to costs. Storage needs volatile markets and frequent turnover to make money. There are already storage markets that are saturated: you only need so many batteries to act as online immediate backup if one of the interconnectors goes offline for instance. Wind surpluses are variable in size, and very intermittent. Whether you use batteries or make hydrogen in electrolysers you face the economics illustrated in this chart:
The economic capacity for storing surpluses will remain quite small: it will be cheaper simply to curtail instead. I think we can forecast that something will hit the fan!
Wow, incredible - there's the article I wanted right there! I really appreciate it and all the work you have been doing to get some sense into energy policy.
It's counterintuitive that although curtailment may increase the prices would potentially decrease because of the new CfD contracts, v.interesting. And also v.problematic for future wind as curtailment increases, so we will see future CfD prices increase hugely without storage.
I've thought of storage as potentially effective for solar; it's more predictable and it would potentially be used every day. Storage used for wind would be so irregularly used to render it infeasibily expensive and like you say the peak to trough difference so large that it just doesn't make sense.
I'd be interested to hear what your ideal system would be considering where we are and the direction of travel regarding net zero.
I’ve looked, albeit some amateurishly, at the economics of storage, with particular reference to PSH. Obviously it’s an arbitrage on power prices between periods of low and high demand. It’s sensitive to capital cost, but also to round trip energy losses, how frequently it’s cycled and how fast you can recharge. By my reckoning PSH can work for peaking — which is just as well, as it is being used for that which gives me a little more confidence in my approach — but it’s difficult to see how to make large schemes work for the long term storage needs to counter periods of dunkelflaute.
It would be interesting to consider battery backing specific solar generators. They don’t generate at night, the time of lowest demand and price, which is a distinct disadvantage. PSH is slow to recharge, for every hour of generation you need al least an hour of pumping water uphill, but it may be possible to “fast charge” batteries (I don’t know). Were that the case, it opens the possibility to quickly completely charge the storage during the shorter periods of day time minima. I suspect that, again, it’s capital costs that will determine.
At least with current battery technology fast charge is possible but is less efficient (more losses) and reduced battery life (so higher CAPEX).
PSH makes a lot more sense for backup as the storable power is much greater compared to the charge/discharge rate. The problem with it is the sites available in which to do it are severely limited. Also, what would the environmental cost be to have a reservoir operating like this?
IIRC that was one of the reasons hydro on the Severn didn't go ahead despite the huge amounts of energy which could be harnessed.
Thats a very comprehensive response thanks for taking the time. In my viewconstraints are an absolute scandal which is only going to get worse over the next five years until the Eastern Green Link DC cables are commissioned. The cost to wind farms is bad enough but those cost are insignificant in the scale of cost of alternative CCGT generation. I don't blame the CCGT owners as the market is rigged against them now so they need to exploit every opportunity to generate. However, this is surely a flawed mkt now as the CCGT operators can see when wind is going to be too high and just stick in high offer prices to generate when NESO comes looking for substitute generation. Quite frankly NO more windmills should have been authorised where there are known transmission constraints its only worsening the situation and not lowering bills.
I don't think that CCGT gets a particularly good price when replacing constrained wind: there is far too much potential competition to allow that. They will be paid a small premium to marginal cost, and gas prices will tend to be dampened by the low CCGT demand. Look at this chart:
When it gets windy the volume of CCGT that gets to run is quite small - probably only the most efficient stations in the best locations. Interconnector imports get backed out (except that if Ireland too has a surplus we may be asked to dispose of it), and there may be exports, depending on the extent of Continental surpluses.
In reality, most of the CCGT that is run on the excuse of alleviating transmission constraints would be run anyway even absent the transmission constraints to provide inertia, but NESO can hide behind the transmission constraint to pretend that wouldn't be a factor - they get more grid that way.
For clarity any generator with a CfD still has to sell their output on the open market to which they could have received a better or worse price than the IMRP and given most of them do so via long term PPAs, which are commercial sensitive contracts, it then becomes a matter of whether they show their hand in the annual accounts.
On a straight average basis though the CfD strike price will drop as the AR3 sites and beyond actualise their CfDs or will they? Then we have ones like Viking which is operational now but has nearly two years before it needs to activate. Conversely Neart na Gaoithe should have been completed four years ago but is only just completing and is allowed to keep its CfD!
Also with the change to the rules on negative pricing from AR4 onwards there's going to be an increasing a number of occasions when the CfDs wont pay out so going to be interesting to see how the generators behave then. Its a foregone conclusion that with the daft amounts of solar now with a CfD let alone smaller commercial schemes that we will see negative pricing on more occasions in high summer. Those with PPAs will have to honour them but maybe able to source energy more cheaply on the wholesale mkt.
All in all a buggers muddle which is then magnified by all the other add ons that are needed to actually make even the expensive CfDs viable like capacity mkt (definitely not going to get cheaper as the decade marches on), BM costs (definitely not going down as wind generation will outrun grid capacity for the next five years) and then the increase in TNUoS charges to pay to remove those grid constraints.
In reality most CFD PPAs price off a basis similar to IMRP because the CFD provides the wind farm with a guaranteed revenue price if they sell on that basis. There are exceptions: Seagreen reports that its contracts with its shareholders protect against both volume and price variation, though of course it isn't operating its CFD. This is hard to interpret, and may mean no more than it gets curtailment payments on an agreed basis.
It is also clear that most operators of CFDs based on baseload prices do not hedge/sell on the basis of baseload prices, but rather decide on the basis of the subsidy or tax (CFD strike price minus BMRP) which is a constant for each summer or winter, and the day ahead market price compared with their costs whether to generate. This resulted in no output during the energy crisis except when prices went sky high during extreme peak periods.
You are quite right to draw attention to the implications of rapidly rising curtailment and negative IMRP levels. The economic logic is that those with no active subsidy will be first in line to curtail, followed by those on the lower CFD strike prices. Here's the consequence at Seagreen
The problem becomes that as there is more competition for curtailment payments as surpluses become more frequent and extensive with rising capacity the prices paid will erode rapidly. This is already happening as he data at REF reveal
That combination is going to hit the economics of more recent wind farms very hard: some risk bankruptcy unless markets are rigged in their favour again. Those not on CFDs do have at least the compensation of much higher prices if they are able to generate at times when we need dispatchable generation and imports. The corollary is that dar from reducing market volatility, it will increase rapidly. This is something being factored into the plans of those investing in grid batteries for whom returns are greatly enhanced by frequent and high market volatility.
Thank you David for another data-driven analysis, which deserves several readings. It has provoked some questions and observations. Can a modern energy system be built without subsidies? In particular, nuclear power, of which I am convinced on account of energy density, small land footprint, reliability and safety. Subsidies, which transfer resources (currency) from the general population to particular interests, are evil, but may be a necessary evil in order to get some things done, where there is a payoff to society, such as keeping the lights on. The UK is in the process of building the most expensive nuclear plant in the world at Hinckley Point C due to our insistence on regulating the best and safest plant ever*. Yet the guaranteed electricity price, while (too) high, is below strike prices for offshore wind. It is as though I am being commanded to pay a premium for a car that won't run in dull, windless conditions. I wouldn't volunteer a visit to that showroom.
Of course, usual reminder: on top of subsidies, we have additional costs of often inefficient backup generation and build out of grid infrastructure.
As per the old Irish navigational advice: "Well I wouldn't start from here".
* China and South Korea have shown that cost and time over-runs on nuclear plant are not a law of physics.
We need to cut a deal with Korea for the APR1400 reactors they are well proven and cost effective and constructionally efficient. Id be quite happy to give them £100/MWh for 4GW baseload.
Peak payments occur on windy days at weekends when demand is lower. This guarantees lower IMRPs and higher volumes of generation. In fact a windy Sunday leads to the highest overall cost to consumers outside of an energy crisis, because most of the dispatchable generation gets switched off, leaving it to be filled by CFDs costing an average of ~£150/MWh and ROC supported generation getting a subsidy of ~£100/MWh and there will be a high volume of curtailment payments which might add £25/MWh to overall costs over the day. These are the real costs to be paid: if a supplier is offering cheap power then the other bill payers will be making up the difference.
On 2. when they curtail they get paid out of the BM pot and as its zero generation no payment from LCCC. This is why the likes of Moray East has such a poor LF is because it offers a cheap price in the BM to disconnect because it has a much lower strike price than its predecessors.
OFGEM does publish data on the recycle payments but these are considerably in arrears, and come in two parts, or even three if mutualisation is triggered, and the announcements aren'teasy to find. Scheme years run to end March, after which ROC holdings can still be traded to square positions until September. Bear in mind that each retailer has to provide ROCs or cashout payments to cover almost half of their metered demand by that time. There is a grace period for late payments but these attract daily interest. The initial cash fund is divvied up less OFGEM expenses and paid out pro rata to ROCs submitted, with late payments being treated similarly later on.
The scheme is designed so that only if renewables production turns out to be higher than the forecast (usually slightly optimistic in practice) plus 10% is there no need for cash payments and the value of the ROC falls to the cashout price. Surplus ROCs can be banked for use the following year. This has only happened once in the scheme history. The average recycle premium in recent years has been about 13%, so I quote ROC values with that as a provisional uplift rather than wait until the following January for hard data. Occasionally, market traded prices for ROCs may be quoted in the energy press (doubtless if you want to spend on a price quotes subscription to the likes of ICIS you could get regular updates) if they are unusual.
For 2023-24 the recycle value was £5.95, just over 10% of the cashout price of £59.01. Here are the announcements
The weighted average number of ROCs per MWh doesn't vary very much since there are no big new units with the scheme being closed. As it winds down and older projects close that may change. A rule of thumb is that there are 4/3rds ROCs per MWh of ROC generation on average, so the total value per MWh currently including uplift works out just shy of £100/MWh based on the cashout value of £64.73/ROC.
You can download the data easily but its deep and wide and then needs some analysis and translation get to actual outputs. Mind you there is an even bigger horror story on subsidies with ROCs.
David, As I understand this article, the annual NZ conference cost to 2030 is circa £26bn - how does that reconcile with the possible £50bn cost suggested by your post of 5th January? Guidance appreciated…
David, I found Figure 9 particularly interesting. You explain the % of the producers’ revenues accounted for by CfD subsidies (56%, offshore wind; 48% onshore wind; 46% solar). Do you know whether renewable energy schemes enjoy taxpayer-assistance on decommissioning costs as do, say, operators of oil & gas fields?
I don't know for certain but I would think yes. Decommissioning is a business expense, so should be allowable against profits
…but, as I understand it, N Sea oil attracts extra taxpayer support which is why, I gather, Miliband’s mad decision has/will accelerate circa £25bn of costs.
Hi David. Thank you for the analysis. Could you do a future article on curtailment and how it is funded/reported on?
Presumably the generator either gets the CfD price and a curtailment bonus or the curtailment payment must just be above the CfD price, presumably it is the latter to be a market solution.
As more wind and solar comes online presumably this cost will increase substantially.
If a wind farm is curtailed there is no CFD payment (or equally, no ROCs earned). CFDs and ROCs only pay out against actual generation. Curtailment payments are handled as a separate item by NESO, the grid operator, as part of the Balancing Mechanism. All generators submit prices at which they would be prepared to increase or decrease generation by a given amount. The control room looks at these bids and selects the cheapest way to adjust the generation portfolio to keep demand and supply matched while also meeting a variety of grid constraints (notably with wind generation, a shortage of transmission capacity between Scotland and England, which can result in other generation in England being fired up to cover the wind that can't be delivered because of the constraint). Constraint payments are paid as bid, not at a market clearing price, partly because some really only apply to handle a local problem rather than a national one which can lead to a large premium when NESO have no other real alternative - something that has become contentious with OFGEM investigations as to what rules they might impose.
Part of the problem with this is that these bids are driven by the subsidies that the wind farms would receive if they were to continue generating. Loss of subsidy is a big loss of income, so the most subsidised wind farms do not bid to curtail, which means that the consumer still gets to pay the full subsidy for all the costliest generation. Wind farms expect to be compensated for their loss of subsidy, plus any loss (or less a gain if the market price is negative) against market prices plus an amount for the insult (like a Bridge bonus).
The entitlements to subsidy vary significantly according to the terms under which the wind farm is operating. Those on ROCs may get anything from 0.7 to 4 ROC/MWh, with each ROC currently worth around £72, so ~£50-300/MWh. There are a number of wind farms that are operating on an unsubsidised basis (particularly those that bid aggressively low and unrealistic CFD strike prices which are waiting as long as possible before commencing the CFD and hoping that inflation will up the strike price to a more economic level in the mean time). All the AR3 wind farms were granted a judicial extension of the period before commencing their CFDs because of covid lockdowns delaying their progress.
There are a variety of terms under CFDs. The earliest CFDs pay out with a ceiling of the strike price when the day ahead market price is negative (IMRP is an hourly price). These wind farms only curtail when it is an operational imperative, and demand the full strike price to do so. The next tranche is on broadly similar terms, except with the wrinkle that if day ahead prices are negative for six or more contiguous hours (usually overnight) then no compensation is paid for the entire period of negative prices. This led to market games with prices being negative for 5 hours, then bid up to just positive for an hour, before being allowed to relapse below zero if necessary for the 7th hour onwards, thus ensuring that they maintained their subsidy throughout. When this gaming was discovered (I was actually the one who wrote to OFGEM with the evidence), the rules were changed for later CFD contracts so that any hour with a negative price pays no CFD compensation, so these wind farms then have the incentive to bid for curtailment payments.
That is actually not a sensible solution, because it means that all the wind farms with low strike prices will remain first in line to curtail, which means they will suffer the lion's share of curtailment on top of only getting a low uneconomic strike price when their CFDs come into operation. Rising curtailment volumes has seen increasing competition among wind farms who get no subsidy in surplus conditions, and so curtailment payments have eroded rapidly, as can be seen here:
https://www.ref.org.uk/constraints/index.php?tab=yr
The result is that they risk bankruptcy in due course unless other ways are found to subsidise them - which would leave us with a sudden shortage of capacity that DESNZ and OFGEM and the CCC simply haven't even thought about.
Curtailment volumes are now at the stage where they are increasing quadratically with increases in capacity: hours already in surplus get bigger surpluses, and there get to be more hours with surplus with each capacity addition. This effect eventually saturates as the extra capacity does very little to solve the problems of low wind hours, while simply piling up more surplus in hours already in surplus. The result is that a new wind farm could expect to see very little of its output earning a reasonable income: the effective price at which it must sell to cover its investment starts to multiply. If Miliband were to achieve his 2030 target a new wind farm would need perhaps 4-5 times the price to cover this. Something I explained already to the Energy Select Committee in written evidence several years ago now.
They seem to think that this will be solved by storage, but of course that adds enormously to costs. Storage needs volatile markets and frequent turnover to make money. There are already storage markets that are saturated: you only need so many batteries to act as online immediate backup if one of the interconnectors goes offline for instance. Wind surpluses are variable in size, and very intermittent. Whether you use batteries or make hydrogen in electrolysers you face the economics illustrated in this chart:
https://datawrapper.dwcdn.net/nZM72/1/
The economic capacity for storing surpluses will remain quite small: it will be cheaper simply to curtail instead. I think we can forecast that something will hit the fan!
I can thoroughly recommend the analyses in Jaberwock's Newsletter on the issue of storage needs and economics:
https://johnd12343.substack.com/p/the-fantasy-of-free-electricity-from
and
https://johnd12343.substack.com/p/green-hydrogen-is-a-crime-against
Wow, incredible - there's the article I wanted right there! I really appreciate it and all the work you have been doing to get some sense into energy policy.
It's counterintuitive that although curtailment may increase the prices would potentially decrease because of the new CfD contracts, v.interesting. And also v.problematic for future wind as curtailment increases, so we will see future CfD prices increase hugely without storage.
I've thought of storage as potentially effective for solar; it's more predictable and it would potentially be used every day. Storage used for wind would be so irregularly used to render it infeasibily expensive and like you say the peak to trough difference so large that it just doesn't make sense.
I'd be interested to hear what your ideal system would be considering where we are and the direction of travel regarding net zero.
I’ve looked, albeit some amateurishly, at the economics of storage, with particular reference to PSH. Obviously it’s an arbitrage on power prices between periods of low and high demand. It’s sensitive to capital cost, but also to round trip energy losses, how frequently it’s cycled and how fast you can recharge. By my reckoning PSH can work for peaking — which is just as well, as it is being used for that which gives me a little more confidence in my approach — but it’s difficult to see how to make large schemes work for the long term storage needs to counter periods of dunkelflaute.
It would be interesting to consider battery backing specific solar generators. They don’t generate at night, the time of lowest demand and price, which is a distinct disadvantage. PSH is slow to recharge, for every hour of generation you need al least an hour of pumping water uphill, but it may be possible to “fast charge” batteries (I don’t know). Were that the case, it opens the possibility to quickly completely charge the storage during the shorter periods of day time minima. I suspect that, again, it’s capital costs that will determine.
At least with current battery technology fast charge is possible but is less efficient (more losses) and reduced battery life (so higher CAPEX).
PSH makes a lot more sense for backup as the storable power is much greater compared to the charge/discharge rate. The problem with it is the sites available in which to do it are severely limited. Also, what would the environmental cost be to have a reservoir operating like this?
IIRC that was one of the reasons hydro on the Severn didn't go ahead despite the huge amounts of energy which could be harnessed.
Thats a very comprehensive response thanks for taking the time. In my viewconstraints are an absolute scandal which is only going to get worse over the next five years until the Eastern Green Link DC cables are commissioned. The cost to wind farms is bad enough but those cost are insignificant in the scale of cost of alternative CCGT generation. I don't blame the CCGT owners as the market is rigged against them now so they need to exploit every opportunity to generate. However, this is surely a flawed mkt now as the CCGT operators can see when wind is going to be too high and just stick in high offer prices to generate when NESO comes looking for substitute generation. Quite frankly NO more windmills should have been authorised where there are known transmission constraints its only worsening the situation and not lowering bills.
I don't think that CCGT gets a particularly good price when replacing constrained wind: there is far too much potential competition to allow that. They will be paid a small premium to marginal cost, and gas prices will tend to be dampened by the low CCGT demand. Look at this chart:
https://i0.wp.com/wattsupwiththat.com/wp-content/uploads/2024/05/Gen-by-Price-Jan-2023-1716328267.549.png
When it gets windy the volume of CCGT that gets to run is quite small - probably only the most efficient stations in the best locations. Interconnector imports get backed out (except that if Ireland too has a surplus we may be asked to dispose of it), and there may be exports, depending on the extent of Continental surpluses.
In reality, most of the CCGT that is run on the excuse of alleviating transmission constraints would be run anyway even absent the transmission constraints to provide inertia, but NESO can hide behind the transmission constraint to pretend that wouldn't be a factor - they get more grid that way.
For clarity any generator with a CfD still has to sell their output on the open market to which they could have received a better or worse price than the IMRP and given most of them do so via long term PPAs, which are commercial sensitive contracts, it then becomes a matter of whether they show their hand in the annual accounts.
On a straight average basis though the CfD strike price will drop as the AR3 sites and beyond actualise their CfDs or will they? Then we have ones like Viking which is operational now but has nearly two years before it needs to activate. Conversely Neart na Gaoithe should have been completed four years ago but is only just completing and is allowed to keep its CfD!
Also with the change to the rules on negative pricing from AR4 onwards there's going to be an increasing a number of occasions when the CfDs wont pay out so going to be interesting to see how the generators behave then. Its a foregone conclusion that with the daft amounts of solar now with a CfD let alone smaller commercial schemes that we will see negative pricing on more occasions in high summer. Those with PPAs will have to honour them but maybe able to source energy more cheaply on the wholesale mkt.
All in all a buggers muddle which is then magnified by all the other add ons that are needed to actually make even the expensive CfDs viable like capacity mkt (definitely not going to get cheaper as the decade marches on), BM costs (definitely not going down as wind generation will outrun grid capacity for the next five years) and then the increase in TNUoS charges to pay to remove those grid constraints.
In reality most CFD PPAs price off a basis similar to IMRP because the CFD provides the wind farm with a guaranteed revenue price if they sell on that basis. There are exceptions: Seagreen reports that its contracts with its shareholders protect against both volume and price variation, though of course it isn't operating its CFD. This is hard to interpret, and may mean no more than it gets curtailment payments on an agreed basis.
It is also clear that most operators of CFDs based on baseload prices do not hedge/sell on the basis of baseload prices, but rather decide on the basis of the subsidy or tax (CFD strike price minus BMRP) which is a constant for each summer or winter, and the day ahead market price compared with their costs whether to generate. This resulted in no output during the energy crisis except when prices went sky high during extreme peak periods.
You are quite right to draw attention to the implications of rapidly rising curtailment and negative IMRP levels. The economic logic is that those with no active subsidy will be first in line to curtail, followed by those on the lower CFD strike prices. Here's the consequence at Seagreen
https://x.com/twallin_james/status/1862443490992357872
The problem becomes that as there is more competition for curtailment payments as surpluses become more frequent and extensive with rising capacity the prices paid will erode rapidly. This is already happening as he data at REF reveal
https://ref.org.uk/constraints/index.php?tab=yr
That combination is going to hit the economics of more recent wind farms very hard: some risk bankruptcy unless markets are rigged in their favour again. Those not on CFDs do have at least the compensation of much higher prices if they are able to generate at times when we need dispatchable generation and imports. The corollary is that dar from reducing market volatility, it will increase rapidly. This is something being factored into the plans of those investing in grid batteries for whom returns are greatly enhanced by frequent and high market volatility.
Thank you David for another data-driven analysis, which deserves several readings. It has provoked some questions and observations. Can a modern energy system be built without subsidies? In particular, nuclear power, of which I am convinced on account of energy density, small land footprint, reliability and safety. Subsidies, which transfer resources (currency) from the general population to particular interests, are evil, but may be a necessary evil in order to get some things done, where there is a payoff to society, such as keeping the lights on. The UK is in the process of building the most expensive nuclear plant in the world at Hinckley Point C due to our insistence on regulating the best and safest plant ever*. Yet the guaranteed electricity price, while (too) high, is below strike prices for offshore wind. It is as though I am being commanded to pay a premium for a car that won't run in dull, windless conditions. I wouldn't volunteer a visit to that showroom.
Of course, usual reminder: on top of subsidies, we have additional costs of often inefficient backup generation and build out of grid infrastructure.
As per the old Irish navigational advice: "Well I wouldn't start from here".
* China and South Korea have shown that cost and time over-runs on nuclear plant are not a law of physics.
We need to cut a deal with Korea for the APR1400 reactors they are well proven and cost effective and constructionally efficient. Id be quite happy to give them £100/MWh for 4GW baseload.
Yes, we need a radical overhaul of nuclear regulation to bring down prices. I covered that here:
https://davidturver.substack.com/p/how-to-make-nuclear-power-cheaper?utm_source=publication-search
A couple of questions.
- is the CFD paid against actual generation, or against nameplate?
- when an offshore generator has to curtail, do they receive any payment for the lost electricty they generate but don't supply, or don't generate
- in days of low supply, (dunkleflaut), off shore wind generates very little electricity, so they get CFD payments on just what they provide?
So I guess what I am asking about is what are the conditions that result in the peak CFD daily payments?
- windy days with high gas prices
- quiet days with high demand
Peak payments occur on windy days at weekends when demand is lower. This guarantees lower IMRPs and higher volumes of generation. In fact a windy Sunday leads to the highest overall cost to consumers outside of an energy crisis, because most of the dispatchable generation gets switched off, leaving it to be filled by CFDs costing an average of ~£150/MWh and ROC supported generation getting a subsidy of ~£100/MWh and there will be a high volume of curtailment payments which might add £25/MWh to overall costs over the day. These are the real costs to be paid: if a supplier is offering cheap power then the other bill payers will be making up the difference.
1. Subsidies paid against actual generation.
2. Yes. But it's complicated, and I am pretty certain those payments do not get recorded in the CfD data in the above article.
3. Yes.
4. Moderately windy days with low gas prices and low demand.
On 2. when they curtail they get paid out of the BM pot and as its zero generation no payment from LCCC. This is why the likes of Moray East has such a poor LF is because it offers a cheap price in the BM to disconnect because it has a much lower strike price than its predecessors.
Thank you so much for your hard work and sharing of factual information. I have learnt such a lot.
Is there a database for ROC payments?
Yes. But the data is published months in arrears and only covers the certificates issued and not the buyout payments.
https://renewablesandchp.ofgem.gov.uk/Public/ReportViewer.aspx?ReportPath=%2fDatawarehouseReports%2fCertificatesExternalPublicDataWarehouse&ReportVisibility=1&ReportCategory=2
OFGEM does publish data on the recycle payments but these are considerably in arrears, and come in two parts, or even three if mutualisation is triggered, and the announcements aren'teasy to find. Scheme years run to end March, after which ROC holdings can still be traded to square positions until September. Bear in mind that each retailer has to provide ROCs or cashout payments to cover almost half of their metered demand by that time. There is a grace period for late payments but these attract daily interest. The initial cash fund is divvied up less OFGEM expenses and paid out pro rata to ROCs submitted, with late payments being treated similarly later on.
The scheme is designed so that only if renewables production turns out to be higher than the forecast (usually slightly optimistic in practice) plus 10% is there no need for cash payments and the value of the ROC falls to the cashout price. Surplus ROCs can be banked for use the following year. This has only happened once in the scheme history. The average recycle premium in recent years has been about 13%, so I quote ROC values with that as a provisional uplift rather than wait until the following January for hard data. Occasionally, market traded prices for ROCs may be quoted in the energy press (doubtless if you want to spend on a price quotes subscription to the likes of ICIS you could get regular updates) if they are unusual.
For 2023-24 the recycle value was £5.95, just over 10% of the cashout price of £59.01. Here are the announcements
https://www.ofgem.gov.uk/publications/renewables-obligation-certificates-presented-and-redistribution-buy-out-fund-2023-24
https://www.ofgem.gov.uk/publications/renewables-obligation-late-payment-distribution-2023-2024
Data on ROC generation and ROC entitlements are published monthly in arrears by DESNZ. Scroll down here for the spreadsheets
https://www.gov.uk/government/statistics/energy-trends-section-6-renewables
The weighted average number of ROCs per MWh doesn't vary very much since there are no big new units with the scheme being closed. As it winds down and older projects close that may change. A rule of thumb is that there are 4/3rds ROCs per MWh of ROC generation on average, so the total value per MWh currently including uplift works out just shy of £100/MWh based on the cashout value of £64.73/ROC.
https://www.ofgem.gov.uk/publications/renewables-obligation-ro-buy-out-price-mutualisation-threshold-and-mutualisation-ceilings-2024-2025
In April, RPI indexation will add about 3.6% (we await the official calculation) to the values.
You can download the data easily but its deep and wide and then needs some analysis and translation get to actual outputs. Mind you there is an even bigger horror story on subsidies with ROCs.
This site attempts to aggregate data
https://energymap.co.uk/default.asp
also
https://www.ref.org.uk/energy-data
have a good database but often a few months in arrears.
Not to mention the ecological devastation caused by offshore wind in particular.