This is another brilliant article, Chris. An exposé using easy to follow analysis of the utter futility of pursuing the 100% renewables path to 'net zero'/eliminating fossil fuel usage. It is, almost literally, "tilting at windmills".
I will have to re-read several times to get comfortable with some of the maths and science but I have already linked to this article on other social media I follow.
Amazing that the 'no nonsense', pragmatic analysis falls on someone like you, doing this non-professionally, when it should really be taken up by officially sponsored and supported engineering/science researchers in industry or academia.
Massive respect for your commitment to truth and integrity. Thank you.
Thank you, it's really good to receive such positive feedback.
For some very believable reasons this is not considered polite speech in many circles of society, and hence why most people who can be cancelled are not touching it, pls see "The Transcult: II" cross-post.
It discusses the recent spate of negative wholesale electricity prices in Europe caused by surplus solar generation and the prospect for a lot more surplus, and how it varies between countries. There is discussion of the need to curtail subsidies for surplus production, but no realisation that this means there would need to be higher prices for useful output to compensate. Also a rather bland assumption that variable demand (when you charge your EV) plus green hydrogen will solve it without looking at what that would take. The $64bn question you rightly ask. There is also an arrogant assumption that nuclear should give way to solar with no justification. They are still way behind the curve in their thinking.
Hello Idau, I've finally got around to responding.
I tend to avoid podcasts that don't have transcriptions, too much blather = too much time wasted.
"still way behind the curve in their thinking" I think applies to far too many policy-makers and commentators who only listen to other people in their 'bubble' or 'silo'.
You're right that hydrogen is not a viable way forward. Incidentally, I think the stability limit you identify is a bit higher than you suggest. It should include nuclear output that can't readily be turned off just because it's windy, and arguably even some of the import capacity that provides a steady input to the grid, although it doesn't offer inertia (and indeed interconnector trips are now the biggest source of larger frequency deviations, increasingly offset by batteries). As the nuclear shuts down we may need to run more gas to maintain adequate grid stability. National Grid has two approaches to this: one is taking bigger risks by lowering stability standards and hoping that battery response is fast and big enough when things start to go wrong, and the other is investing in additional stabilisation kit such as synchronous compensators. If it does go wrong the August 2019 blackout is likely to be vastly exceeded.
Back to hydrogen. I find it useful to look at renewables surpluses as duration curves. This highlights the fact that the interaction of varying demand and varying generation produces varying surpluses. It is never going to be economic to attempt to handle the largest surpluses, because you would have to build grid and electrolyser capacity to service them that would only be called upon very rarely. So you are still going to wind up with a large chunk of curtailment anyway. This chart gives a flavour of the duration curves you could expect from different levels of wind capacity (I made it when we had about 22GW installed, so the curves are for multiples of this up to 132GW).
It's a mouseover chart. If you select a point on one of the curves you get a readout of the grid and electrolyser capacity you need to install, and the utilisation it would get: everything to the left and above your point gets curtailed. Note that there remains a big proportion of the time when there is no surplus - especially if you only have limited wind capacity installed. If you use hydrogen to power generators when there is a deficit you will suffer the whammys of a very low round trip efficiency (say 60% for a PEM electorlyser, perhaps somewhat less in intermittent mode, and 50% for an intermittent gas generator, so no better than 30% overall) as well as higher costs from low utilisation of the generator.
Of course in practice these surpluses are going to vary significantly on very short timescales - as demand picks up ahead of the morning rush hour for example, or when a weather system moves through, or as the sun goes higher in the sky in summer. As you point out that's not good for efficient operational performance of electrolysers.
Clearly what you should not do is to try to keep the utilisation of the electrolysers higher by running them even when you have no surplus: you would be burning inefficiently made hydrogen to make hydrogen at a very low efficiency. The CCC idea of using floating offshore wind to run electorlysers is completely nuts. It takes the most expensive form of wind generation as the input to the process, and multiplies the cost due to the inefficiencies. It also (because there is no other choice) will end up with hydrogen being made instead of using directly produced electricity - exactly the scenario to avoid.
Idau, thank you again, and back to your comments, my responses in more detail.
Re: "the stability limit", I'm just going from what I see on the charts, which is a renewables (wind+solar) cut-off between ~1,400 MW and 2,500 MW - round number 2,000 MW. Nuclear is base-load, nuclear engineers have told me that the UK's reactors *could* follow demand fluctuations but that would do all kinds of bad things to the reactors, so they don't. (There's some speculation that French reactors' problems last year resulted from operating them in demand-following mode.) The 2,000 MW stability limit would nominally allow for the trip of one nuclear plant without unbalancing the wider grid.
Re: "import capacity that provides a steady input to the grid" - oh no it doesn't. The interconnectors are all 2-way and all switch from import to export and back depending on the relative power balances and prices UK vs Norway & Europe.
Re: "interconnector trips" - yes, in either direction. The interconnector capacities are up there alongside the nuclear stations; most 1,000 MW, but the "North Sea link Interconnector ... connects Kvilldal in Norway to Blyth in the UK. The cable is 730 km (450 miles) long, and has a capacity of 1400 MW."
Re: "synchronous compensators" = large rotating masses like generators but not able to generate - my understanding is Australia is planning some of these.
Re: "Clearly what you should not do is to try to keep the utilisation of the electrolysers higher by running them even when you have no surplus" - agreed 100%.
Hello Idau, thank you, you seem to know an awful lot more about our grid system than I do.
I looked at your duration curves and I don't fully understand what they're based on or exactly what they're showing... but I fully understand that large surpluses of renewables would have to be curtailed.
What I'm trying to do with my posts is illustrate how divorced from reality a lot of the renewables boosters are. More and more wind won't keep our lights on (unless the multiple is 20 or 30 or ... times current capacity).
More solar just adds to the huge surges the rest of the grid has to cope with or adds to the curtailment of wind if it coincides.
And hydrogen boosters have drifted even farther from their moorings. The idea that we can generate enough hydrogen from renewable energy to supplant natural gas is surreal. I didn't have the data to support my hunch that this was true: now I think I do.
So, I think we are back to battery tech if we truly think we need to stop burning fossil fuels.
Maybe Form Energy Inc with their iron-air batteries will be the solution...?
The duration curves aren't difficult to construct. Start by constructing the surpluses and deficits for a particular level of wind capacity, just as you have done: I used 6GW of nuclear as the stabilisation factor rather than your 2GW of gas. Then you can take the values of the surpluses and deficits and sort them in descending order from largest surplus to largest deficit. As we are looking at fuelling green hydrogen manufacture, we're only interested in surpluses, so all the deficits rank as a zero surplus. If you number the sorted values in an adjoining column you can calculate the percentage of the total number of time periods that each rank corresponds to.
Thus organised we can answer the question "What percentage of the time is the surplus greater than some given value?" The largest surplus is where the curve intercepts the vertical axis, and so it only occurred in one time period (ignoring unlikely ties). Where the curve hits the horizontal axis the surplus is zero, so the rest of the time there is no surplus.
An intermediate point on the curve has higher surpluses to the left of it. If you set a capacity limit on transmission or electrolysis at that level, higher surpluses cannot be used. So the roughly triangular area above that level and under the curve must be curtailed. At the same time, its percentage value shows the utilisation factor for the capacity chosen at the margin - that is, how much use do you get from building an extra MW of electrolysis if you fuel it from surpluses? Of course, the first bits of capacity you build will fare rather better - but even the very first MW can't do better than the percentage of time where there is a surplus - i.e. where the curve intercepts the horizontal axis.
An alternative way of getting the curves from the calculated surpluses (zero the deficits by using an IF function) is to use the spreadsheet PERCENTILE function, which allows you to use a column of percentages as inputs along with the array of surplus data (including the zero values) without having to sort it: the function does the hard work.
In the real world it is likely that the unusable surpluses will be larger, in just the same way as we are already seeing curtailment regularly even with wind output in total being below residual demand, because you have to look at each location rather than assuming a "copper plate grid" where any surplus can be transmitted to any electrolyser.
I have downloaded the demand, wind and solar data from gridwatch for 2022 into an Excel file to model UK Labour’s proposal to decarbonise the electricity by 2030 by quadrupling offshore wind, doubling onshore wind and trebling solar.
I calculate that for this 2030 proposal where the average power demand is 36 GW and maximum 56GW we will require :
Hydrogen storage : 19 TWhrs = 600K tonnes (taking a generating efficiency of 40%)
Battery : 10 TWhrs
I have also made some costings, but these were before the offshore wind industry, requesting a “budget” increase, failed to make any bids for AR5.
If anyone else is interested in seeing and checking my method & calculations please email me at jbxcagwnz@gmail.com
"Of course, until GB and other countries honestly account for embedded emissions in imports, we will continue to de-industrialise to the benefit of other countries. But that’s a complete whole other story apparently too complex for our ‘leaders’ to grasp, so they’ll continue to pursue policies that reduce CO2-only in-territory emissions that incentivise our economic decline."
Economic decline is the purpose of Net Zero not an unintended consequence.
This is another brilliant article, Chris. An exposé using easy to follow analysis of the utter futility of pursuing the 100% renewables path to 'net zero'/eliminating fossil fuel usage. It is, almost literally, "tilting at windmills".
I will have to re-read several times to get comfortable with some of the maths and science but I have already linked to this article on other social media I follow.
Amazing that the 'no nonsense', pragmatic analysis falls on someone like you, doing this non-professionally, when it should really be taken up by officially sponsored and supported engineering/science researchers in industry or academia.
Massive respect for your commitment to truth and integrity. Thank you.
Hello Tim,
Thank you, it's really good to receive such positive feedback.
For some very believable reasons this is not considered polite speech in many circles of society, and hence why most people who can be cancelled are not touching it, pls see "The Transcult: II" cross-post.
All the best.
I found this podcast was quite interesting.
https://www.buzzsprout.com/257968?client_source=large_player&iframe=true&referrer=https://www.buzzsprout.com/257968.js?container_id=buzzsprout-large-player&player=large#
It discusses the recent spate of negative wholesale electricity prices in Europe caused by surplus solar generation and the prospect for a lot more surplus, and how it varies between countries. There is discussion of the need to curtail subsidies for surplus production, but no realisation that this means there would need to be higher prices for useful output to compensate. Also a rather bland assumption that variable demand (when you charge your EV) plus green hydrogen will solve it without looking at what that would take. The $64bn question you rightly ask. There is also an arrogant assumption that nuclear should give way to solar with no justification. They are still way behind the curve in their thinking.
Hello Idau, I've finally got around to responding.
I tend to avoid podcasts that don't have transcriptions, too much blather = too much time wasted.
"still way behind the curve in their thinking" I think applies to far too many policy-makers and commentators who only listen to other people in their 'bubble' or 'silo'.
You're right that hydrogen is not a viable way forward. Incidentally, I think the stability limit you identify is a bit higher than you suggest. It should include nuclear output that can't readily be turned off just because it's windy, and arguably even some of the import capacity that provides a steady input to the grid, although it doesn't offer inertia (and indeed interconnector trips are now the biggest source of larger frequency deviations, increasingly offset by batteries). As the nuclear shuts down we may need to run more gas to maintain adequate grid stability. National Grid has two approaches to this: one is taking bigger risks by lowering stability standards and hoping that battery response is fast and big enough when things start to go wrong, and the other is investing in additional stabilisation kit such as synchronous compensators. If it does go wrong the August 2019 blackout is likely to be vastly exceeded.
Back to hydrogen. I find it useful to look at renewables surpluses as duration curves. This highlights the fact that the interaction of varying demand and varying generation produces varying surpluses. It is never going to be economic to attempt to handle the largest surpluses, because you would have to build grid and electrolyser capacity to service them that would only be called upon very rarely. So you are still going to wind up with a large chunk of curtailment anyway. This chart gives a flavour of the duration curves you could expect from different levels of wind capacity (I made it when we had about 22GW installed, so the curves are for multiples of this up to 132GW).
https://datawrapper.dwcdn.net/nZM72/1/
It's a mouseover chart. If you select a point on one of the curves you get a readout of the grid and electrolyser capacity you need to install, and the utilisation it would get: everything to the left and above your point gets curtailed. Note that there remains a big proportion of the time when there is no surplus - especially if you only have limited wind capacity installed. If you use hydrogen to power generators when there is a deficit you will suffer the whammys of a very low round trip efficiency (say 60% for a PEM electorlyser, perhaps somewhat less in intermittent mode, and 50% for an intermittent gas generator, so no better than 30% overall) as well as higher costs from low utilisation of the generator.
Of course in practice these surpluses are going to vary significantly on very short timescales - as demand picks up ahead of the morning rush hour for example, or when a weather system moves through, or as the sun goes higher in the sky in summer. As you point out that's not good for efficient operational performance of electrolysers.
Clearly what you should not do is to try to keep the utilisation of the electrolysers higher by running them even when you have no surplus: you would be burning inefficiently made hydrogen to make hydrogen at a very low efficiency. The CCC idea of using floating offshore wind to run electorlysers is completely nuts. It takes the most expensive form of wind generation as the input to the process, and multiplies the cost due to the inefficiencies. It also (because there is no other choice) will end up with hydrogen being made instead of using directly produced electricity - exactly the scenario to avoid.
Idau, thank you again, and back to your comments, my responses in more detail.
Re: "the stability limit", I'm just going from what I see on the charts, which is a renewables (wind+solar) cut-off between ~1,400 MW and 2,500 MW - round number 2,000 MW. Nuclear is base-load, nuclear engineers have told me that the UK's reactors *could* follow demand fluctuations but that would do all kinds of bad things to the reactors, so they don't. (There's some speculation that French reactors' problems last year resulted from operating them in demand-following mode.) The 2,000 MW stability limit would nominally allow for the trip of one nuclear plant without unbalancing the wider grid.
Re: "import capacity that provides a steady input to the grid" - oh no it doesn't. The interconnectors are all 2-way and all switch from import to export and back depending on the relative power balances and prices UK vs Norway & Europe.
Re: "interconnector trips" - yes, in either direction. The interconnector capacities are up there alongside the nuclear stations; most 1,000 MW, but the "North Sea link Interconnector ... connects Kvilldal in Norway to Blyth in the UK. The cable is 730 km (450 miles) long, and has a capacity of 1400 MW."
Re: "synchronous compensators" = large rotating masses like generators but not able to generate - my understanding is Australia is planning some of these.
Re: "Clearly what you should not do is to try to keep the utilisation of the electrolysers higher by running them even when you have no surplus" - agreed 100%.
Hello Idau, thank you, you seem to know an awful lot more about our grid system than I do.
I looked at your duration curves and I don't fully understand what they're based on or exactly what they're showing... but I fully understand that large surpluses of renewables would have to be curtailed.
What I'm trying to do with my posts is illustrate how divorced from reality a lot of the renewables boosters are. More and more wind won't keep our lights on (unless the multiple is 20 or 30 or ... times current capacity).
More solar just adds to the huge surges the rest of the grid has to cope with or adds to the curtailment of wind if it coincides.
And hydrogen boosters have drifted even farther from their moorings. The idea that we can generate enough hydrogen from renewable energy to supplant natural gas is surreal. I didn't have the data to support my hunch that this was true: now I think I do.
So, I think we are back to battery tech if we truly think we need to stop burning fossil fuels.
Maybe Form Energy Inc with their iron-air batteries will be the solution...?
The duration curves aren't difficult to construct. Start by constructing the surpluses and deficits for a particular level of wind capacity, just as you have done: I used 6GW of nuclear as the stabilisation factor rather than your 2GW of gas. Then you can take the values of the surpluses and deficits and sort them in descending order from largest surplus to largest deficit. As we are looking at fuelling green hydrogen manufacture, we're only interested in surpluses, so all the deficits rank as a zero surplus. If you number the sorted values in an adjoining column you can calculate the percentage of the total number of time periods that each rank corresponds to.
Thus organised we can answer the question "What percentage of the time is the surplus greater than some given value?" The largest surplus is where the curve intercepts the vertical axis, and so it only occurred in one time period (ignoring unlikely ties). Where the curve hits the horizontal axis the surplus is zero, so the rest of the time there is no surplus.
An intermediate point on the curve has higher surpluses to the left of it. If you set a capacity limit on transmission or electrolysis at that level, higher surpluses cannot be used. So the roughly triangular area above that level and under the curve must be curtailed. At the same time, its percentage value shows the utilisation factor for the capacity chosen at the margin - that is, how much use do you get from building an extra MW of electrolysis if you fuel it from surpluses? Of course, the first bits of capacity you build will fare rather better - but even the very first MW can't do better than the percentage of time where there is a surplus - i.e. where the curve intercepts the horizontal axis.
An alternative way of getting the curves from the calculated surpluses (zero the deficits by using an IF function) is to use the spreadsheet PERCENTILE function, which allows you to use a column of percentages as inputs along with the array of surplus data (including the zero values) without having to sort it: the function does the hard work.
In the real world it is likely that the unusable surpluses will be larger, in just the same way as we are already seeing curtailment regularly even with wind output in total being below residual demand, because you have to look at each location rather than assuming a "copper plate grid" where any surplus can be transmitted to any electrolyser.
Thank you re: the duration curves, all now clear.
I happen to think my spiky charts demonstrate the issue more clearly, but then, I *would* think that :)
In case of interest:
I have downloaded the demand, wind and solar data from gridwatch for 2022 into an Excel file to model UK Labour’s proposal to decarbonise the electricity by 2030 by quadrupling offshore wind, doubling onshore wind and trebling solar.
I calculate that for this 2030 proposal where the average power demand is 36 GW and maximum 56GW we will require :
Hydrogen storage : 19 TWhrs = 600K tonnes (taking a generating efficiency of 40%)
Battery : 10 TWhrs
I have also made some costings, but these were before the offshore wind industry, requesting a “budget” increase, failed to make any bids for AR5.
If anyone else is interested in seeing and checking my method & calculations please email me at jbxcagwnz@gmail.com
"Of course, until GB and other countries honestly account for embedded emissions in imports, we will continue to de-industrialise to the benefit of other countries. But that’s a complete whole other story apparently too complex for our ‘leaders’ to grasp, so they’ll continue to pursue policies that reduce CO2-only in-territory emissions that incentivise our economic decline."
Economic decline is the purpose of Net Zero not an unintended consequence.