I took a look at the latest plans for Cleve Hill. They estimate the solar farm at 373MW peak, and the battery is 150MW output, 700MWh of storage (though these figures are less certain). They have taken out a CFD under Allocation Round 4 covering 110MW of the solar capacity at a price that indexes currently to £56.99/MWh. I had forgotten that the major substation they will link to also serves the London Array wind farm, so they will be in a position to bid for wind surpluses to fill the battery in winter when solar output is low.
I found a local rooftop solar installation with excellent detailed historical data recorded here:
It's located outside Rochester, just a few miles away, and gives an excellent idea of the likely performance of the solar park. You can look at individual days to see how a weather front can dramatically change the output in the space of a few minutes. You can see how variable output is from one day to the next within a month, and look at the monthly seasonality and compare the annual performance back to 2015. These variations limit the extent to which the battery can be used profitably as a solar store.
The battery is designed to take less than half the maximum solar output as a charge rate. Indeed, it's only a little bit bigger than the capacity contracted to CFD. The way the CFD works, if there is a general solar surplus that pushes prices down on sunny summer days, is that consumers get to pay the subsidy for the MWh generated from the 110MW of CFD capacity, which can be used to charge the battery. So from the consumer point of view the cost of charging the battery in £56.99/MWh, but the battery gets its supply at somewhere close to zero cost. The battery then gets to sell the output in the evening when prices are rather higher, covering the losses in the round trip. Its interest is to discharge at the best prices it can, not to maintain any even flow. In winter solar output is low, and in fact not sufficient to charge the battery fully. However, it may still make sense to divert all the output to the battery (whose input capacity can easily handle the low input), and then discharge it at premium prices in the evening rush hour which will be enough to draw down the stored energy as much as is wise. The battery economics are quite complicated, especially if you include the potential for other sources and providing grid ancillary services in the mix. But the optimum economics are going to be some way away from the idea of creating an even flow of output.
I've just downloaded a bundle of info on Cleve Hill from the Kent planning website, but as it's all at 'application' stage I'm not going to spend too much time on it.
Your explanation of how it might work is illuminating/depressing. Here is a system of pricing; here are some rich developers who see an opportunity to make mega-bucks; here are the power consumers picking up the tab.
Hi again Idau, thank you for all your comments on this post and others.
I've been pretty busy recently and need to get some more stuff done over the next week or so, so won't be replying in detail for a bit, but I do greatly appreciate your inputs.
If you forced solar to come with storage you wouldn't have much solar, especially if you demanded even flow across a day. If you wanted even flow across the year there would definitely be none, because interseasonal storage is prohibitively expensive. 4 hour duration storage (MWh holding capacity divided by MW of maximum output) has challenging economics: it's what they're putting in at the giant solar farm Cleve Hill near Faversham, where the storage might have a secondary role in supporting stabilisation of interconnector flows from BritNed and NEMO if they trip out.
It's much cheaper to curtail. Trouble is, curtailment tends only to apply to commercial ventures. There are difficulties with organising curtailment for domestic rooftop solar, including ensuring that it doesn't result in energy build-up that leads to house fires. You can see the effect on a typical summer day in South Australia (where excess solar drives grid prices negative)
One of the reasons for my post was to emphasise the (in my opinion) absurdity of ever more solar *capacity* while completely ignoring the negative implications.
No, my proposed buffer storage would never be economic, because in reality solar PV in the UK will never be truly economic.
But that's all hidden by the subsidies/feed-in tariffs/opaque pricing.
Re: "rooftop solar doesn't see the negative prices " - but I thought *that* kind of functionality was one of the main purposes of 'smart meters', no? Which would further lead into providing safe functionality to make curtailment of domestic solar a design requirement. But retrofitting that functionality across the existing 'fleet' would likely not be simple nor cheap.
As far as I'm aware, not one had the intelligence to point out that due to their annual capacity factors, (broadly speaking) wind could feed approx 4x the annual quantity of electricity into the grid that solar could manage via an identical infrastructure capacity improvement. Consequently, nearly all solar projects should be relegated to the back of the queue.
Hello Ron, thank you and yes I saw that Beeb article.
In fact I complained to the Beeb that the graphic in the article headed "How the UK electricity system works" shows only renewables and thereby gives a misleading impression. 'Misinformation', perhaps?
Whereas if you follow the https://www.bbc.co.uk/news/business-63976805 link in the article and scroll down to the Electricity generated in the UK in 2022 bar-chart you see the true picture with 38.4% gas heading the sources.
They eventually replied along the lines 'oh yes, will do better next time'. I note they did not go back and edit the article to correct the misinformation.
Anyhow, to your point re: capacity factors.
My understanding is that, because Scotland has wonderful windy conditions, quite a lot of the constraints are within Scotland (wind farms given the go-ahead without considering local grid system requirements) as well as between Scotland and Rest of UK (where the majority of the Demand resides).
If this is correct and a lot of the large-scale Solar is south of the border, then maybe large-scale Solar actually helps relatively more...?
But I definitely think that continuing to add large-scale Solar without addressing the spikes of power it injects into the grid is just piling up the problems.
Which is why I believe smoothing buffer electricity storage as I propose would greatly help.
Large scale solar is concentrated in the SW where it tends to be sunniest. So much is expected to be built that they're having to build a new transmission line to export it to elsewhere in England. Otherwise it will be constrained off. For now they are relying on Hinkley B shutting down and Hinkley C being delayed forever. National Grid have been working towards curtailment options for larger solar installations for some time: the trouble is that they are (except for one, just opened) not connected to the National Grid transmission system, but instead are "embedded generation" in lower voltage distribution networks, normally read simply as a reduction in demand on the grid. This complicates control, because their metering is not visible to grid control.
Interesting and makes a lot of sense. Would each solar facility need its own storage or could multiple solar facilities in a compact geographical area use a combined storage facility (assuming there are economies of scale)? Alternatively, could "we" (the grid?) independently implement standalone storage, so that "we" do not need to pay for curtailment?
I imagine that, if we were being completely rational, we would design a system to minimise life-time costs. Which may mean some shared storage for adjacent solar farms, assuming joint ownership, liability, etc. etc. can be sorted out.
But the cynic in me predicts arbitrary targets would instead be set, coz politicians *love* announcing targets, resulting in costs being maximised and benefits/savings minimised.
May 26, 2023·edited May 26, 2023Liked by Chris Bond
I’m not an electrical engineer, but, as a mechanical engineer, I participated in the early years of the BIPV development.
To my none expert mind, maintaining 50Hz with rotating generators is a given. (Speed regulated thermal plant, not wind turbines).
PV generators are always DC, and are rectified by clever bits of kit to ‘emulate’ 50Hz, but they need to be synchronised with the grid, via some kind of signal.
If the grid has no or minimal rotating plant then PV and wind must be disconnected.
That’s the project developers financial risk.
Not customers.
No retail or wholesale customers should be charged. The investors in the wind turbines or PV farms should take a haircut, not a financial reward.
Thank you, Gareth, yes, definitely frequency has to be controlled to tight tolerances.
Currently I think GB has enough traditional rotating masses in the system to provide frequency signal and inertia. [DC inverters work by magic as far as I'm concerned :) ]
I think I read that some states in Australia which have shut down thermal plants and have high proportions of Wind + Solar are planning to install non-generating massive rotating windings to provide frequency inertia.
Problems that I presume power engineers can easily solve, and yes, to be funded by the renewables sector which causes the problem.
I’m not an expert but I’m told that the export meter on a PV installation has a device which measures the frequency and allows export.
My complaint, really directed against OFGEM, is that, for reasons I don’t understand, OFGEM makes it impossible or extremely difficult to independently monitor the value of individual house level exported electricity. You have to go through a local utility company.
However, with such a clever export meter, this is obviously nonsense.
I took a look at the latest plans for Cleve Hill. They estimate the solar farm at 373MW peak, and the battery is 150MW output, 700MWh of storage (though these figures are less certain). They have taken out a CFD under Allocation Round 4 covering 110MW of the solar capacity at a price that indexes currently to £56.99/MWh. I had forgotten that the major substation they will link to also serves the London Array wind farm, so they will be in a position to bid for wind surpluses to fill the battery in winter when solar output is low.
I found a local rooftop solar installation with excellent detailed historical data recorded here:
https://www.sunnyportal.com/Templates/PublicPageOverview.aspx?page=9ae1951a-36a3-439b-8e09-06697d5d1452&plant=71b0b704-3255-4883-82cf-26fb78a5e0ca&splang=en-US
It's located outside Rochester, just a few miles away, and gives an excellent idea of the likely performance of the solar park. You can look at individual days to see how a weather front can dramatically change the output in the space of a few minutes. You can see how variable output is from one day to the next within a month, and look at the monthly seasonality and compare the annual performance back to 2015. These variations limit the extent to which the battery can be used profitably as a solar store.
The battery is designed to take less than half the maximum solar output as a charge rate. Indeed, it's only a little bit bigger than the capacity contracted to CFD. The way the CFD works, if there is a general solar surplus that pushes prices down on sunny summer days, is that consumers get to pay the subsidy for the MWh generated from the 110MW of CFD capacity, which can be used to charge the battery. So from the consumer point of view the cost of charging the battery in £56.99/MWh, but the battery gets its supply at somewhere close to zero cost. The battery then gets to sell the output in the evening when prices are rather higher, covering the losses in the round trip. Its interest is to discharge at the best prices it can, not to maintain any even flow. In winter solar output is low, and in fact not sufficient to charge the battery fully. However, it may still make sense to divert all the output to the battery (whose input capacity can easily handle the low input), and then discharge it at premium prices in the evening rush hour which will be enough to draw down the stored energy as much as is wise. The battery economics are quite complicated, especially if you include the potential for other sources and providing grid ancillary services in the mix. But the optimum economics are going to be some way away from the idea of creating an even flow of output.
Thank you again, Idau.
I've just downloaded a bundle of info on Cleve Hill from the Kent planning website, but as it's all at 'application' stage I'm not going to spend too much time on it.
From https://www.kentwildlifetrust.org.uk/campaigns/planning-and-development/cleve-hill-solar-park I get an idea - it's huge and it's contentious.
Your explanation of how it might work is illuminating/depressing. Here is a system of pricing; here are some rich developers who see an opportunity to make mega-bucks; here are the power consumers picking up the tab.
Hi again Idau, thank you for all your comments on this post and others.
I've been pretty busy recently and need to get some more stuff done over the next week or so, so won't be replying in detail for a bit, but I do greatly appreciate your inputs.
If you forced solar to come with storage you wouldn't have much solar, especially if you demanded even flow across a day. If you wanted even flow across the year there would definitely be none, because interseasonal storage is prohibitively expensive. 4 hour duration storage (MWh holding capacity divided by MW of maximum output) has challenging economics: it's what they're putting in at the giant solar farm Cleve Hill near Faversham, where the storage might have a secondary role in supporting stabilisation of interconnector flows from BritNed and NEMO if they trip out.
It's much cheaper to curtail. Trouble is, curtailment tends only to apply to commercial ventures. There are difficulties with organising curtailment for domestic rooftop solar, including ensuring that it doesn't result in energy build-up that leads to house fires. You can see the effect on a typical summer day in South Australia (where excess solar drives grid prices negative)
https://i0.wp.com/wattsupwiththat.com/wp-content/uploads/2023/05/SA-Gen-5-Jan-23-1684966578.2893.png
Batteries don't really figure as a solution. The fact that rooftop solar doesn't see the negative prices is the start of the problem.
Thank you, Idau,
One of the reasons for my post was to emphasise the (in my opinion) absurdity of ever more solar *capacity* while completely ignoring the negative implications.
No, my proposed buffer storage would never be economic, because in reality solar PV in the UK will never be truly economic.
But that's all hidden by the subsidies/feed-in tariffs/opaque pricing.
Re: "rooftop solar doesn't see the negative prices " - but I thought *that* kind of functionality was one of the main purposes of 'smart meters', no? Which would further lead into providing safe functionality to make curtailment of domestic solar a design requirement. But retrofitting that functionality across the existing 'fleet' would likely not be simple nor cheap.
Hi Chris
Another interesting post. Thanks. 👍
Earlier this month (nearly) all the MSM were whinging about National Grid's extended time scales for connecting new renewables supplies to their grid.
E.g:
"Billions of pounds' worth of green energy projects are on hold because they cannot plug into the UK's electricity system, BBC research shows.
Some new solar and wind sites are waiting up to 10 to 15 years to be connected because of a lack of capacity in the system - known as the "grid"....
... A new wind farm or solar site can only start supplying energy to people's homes once it has been plugged into the grid."
https://www.bbc.co.uk/news/science-environment-65500339
As far as I'm aware, not one had the intelligence to point out that due to their annual capacity factors, (broadly speaking) wind could feed approx 4x the annual quantity of electricity into the grid that solar could manage via an identical infrastructure capacity improvement. Consequently, nearly all solar projects should be relegated to the back of the queue.
Hello Ron, thank you and yes I saw that Beeb article.
In fact I complained to the Beeb that the graphic in the article headed "How the UK electricity system works" shows only renewables and thereby gives a misleading impression. 'Misinformation', perhaps?
Whereas if you follow the https://www.bbc.co.uk/news/business-63976805 link in the article and scroll down to the Electricity generated in the UK in 2022 bar-chart you see the true picture with 38.4% gas heading the sources.
They eventually replied along the lines 'oh yes, will do better next time'. I note they did not go back and edit the article to correct the misinformation.
Anyhow, to your point re: capacity factors.
My understanding is that, because Scotland has wonderful windy conditions, quite a lot of the constraints are within Scotland (wind farms given the go-ahead without considering local grid system requirements) as well as between Scotland and Rest of UK (where the majority of the Demand resides).
If this is correct and a lot of the large-scale Solar is south of the border, then maybe large-scale Solar actually helps relatively more...?
But I definitely think that continuing to add large-scale Solar without addressing the spikes of power it injects into the grid is just piling up the problems.
Which is why I believe smoothing buffer electricity storage as I propose would greatly help.
Large scale solar is concentrated in the SW where it tends to be sunniest. So much is expected to be built that they're having to build a new transmission line to export it to elsewhere in England. Otherwise it will be constrained off. For now they are relying on Hinkley B shutting down and Hinkley C being delayed forever. National Grid have been working towards curtailment options for larger solar installations for some time: the trouble is that they are (except for one, just opened) not connected to the National Grid transmission system, but instead are "embedded generation" in lower voltage distribution networks, normally read simply as a reduction in demand on the grid. This complicates control, because their metering is not visible to grid control.
Thank you for your extra info, much appreciated.
Interesting and makes a lot of sense. Would each solar facility need its own storage or could multiple solar facilities in a compact geographical area use a combined storage facility (assuming there are economies of scale)? Alternatively, could "we" (the grid?) independently implement standalone storage, so that "we" do not need to pay for curtailment?
Thank you, Richard.
I imagine that, if we were being completely rational, we would design a system to minimise life-time costs. Which may mean some shared storage for adjacent solar farms, assuming joint ownership, liability, etc. etc. can be sorted out.
But the cynic in me predicts arbitrary targets would instead be set, coz politicians *love* announcing targets, resulting in costs being maximised and benefits/savings minimised.
I’m not an electrical engineer, but, as a mechanical engineer, I participated in the early years of the BIPV development.
To my none expert mind, maintaining 50Hz with rotating generators is a given. (Speed regulated thermal plant, not wind turbines).
PV generators are always DC, and are rectified by clever bits of kit to ‘emulate’ 50Hz, but they need to be synchronised with the grid, via some kind of signal.
If the grid has no or minimal rotating plant then PV and wind must be disconnected.
That’s the project developers financial risk.
Not customers.
No retail or wholesale customers should be charged. The investors in the wind turbines or PV farms should take a haircut, not a financial reward.
Thank you, Gareth, yes, definitely frequency has to be controlled to tight tolerances.
Currently I think GB has enough traditional rotating masses in the system to provide frequency signal and inertia. [DC inverters work by magic as far as I'm concerned :) ]
I think I read that some states in Australia which have shut down thermal plants and have high proportions of Wind + Solar are planning to install non-generating massive rotating windings to provide frequency inertia.
Problems that I presume power engineers can easily solve, and yes, to be funded by the renewables sector which causes the problem.
I’m not an expert but I’m told that the export meter on a PV installation has a device which measures the frequency and allows export.
My complaint, really directed against OFGEM, is that, for reasons I don’t understand, OFGEM makes it impossible or extremely difficult to independently monitor the value of individual house level exported electricity. You have to go through a local utility company.
However, with such a clever export meter, this is obviously nonsense.