Guest Post by Willis Eschenbach
Christopher Monckton recently put up a fascinating post entitled “The Final Nail in The Coffin Of “Renewable” Energy”. In it, he references the work of a man named Douglas Pollock who has proposed that there is a limit to the share of energy that a given renewable source can supply to the grid without battery backup. Further, Pollock says that the limit is the “capacity factor”, the fraction of the nameplate capacity that a renewable source can actually supply.
A rebuttal of this was put up as a post entitled “Sealing The Coffin Of “Renewable” Energy May Take A Few More Nails“, and Lord Monckton posted up a re-rebuttal entitled “Why Climate Skepticism Has Not Yet Succeeded“.
(Unfortunately, in the comments of the latter post I fear I waxed wroth when Christopher falsely accused me of being “openly and deliberately dishonest” … ah, well, I know that science is a blood sport, but I won’t take that from any man. However, I digress …)
While interesting, Lord Monckton’s posts are theoretical exercises. He has not provided any actual data to back them up. And when I looked at the data, I found a problem—most countries are well below the “Pollock limit”, and thus they can’t say anything at all about what happens when the windpower share of total electrical generation nears the Pollock limit.
Figure 1. Percentage of electricity from wind by country. Dotted line shows the global average wind capacity factor, which is the average fraction of the nameplate capacity that a wind turbine can actually supply in the real world.
However, a couple of countries have renewables shares that are above the Pollock limit. I picked Ireland as a test case. I got the annual information on the Irish electrical supply from the BP Statistical Review of World Energy. Here, on a year-by-year basis, is the annual windpower share of total Irish electricity versus the annual installed capacity (red/black line), as well as the annual capacity factor (yellow/black line).
Figure 2. Annual Irish wind share versus installed wind capacity, and wind capacity factor. 2022 values are from here and are preliminary.
There are several interesting insights from Figure 2. First, in contradiction to the proposed numerical value of the Pollock limit as being equal to the capacity factor, the wind power share of total Irish electricity is well above the wind capacity factor.
Next, it’s interesting how much the wind capacity factor varies year to year, swinging about ± 5% above and below the average value,
Next, in agreement with the concept of the Pollock limit, as the installed capacity has increased, the windpower share has moved more and more in parallel with the wind capacity factor.
Finally, the last four years are particularly interesting. From 2019 to 2022 Ireland added about 4 TWh of wind nameplate capacity … but the share of the total generated by wind only increased slightly. So it certainly appears as though it’s approaching some kind of limit.
These facts taken together suggest that there is a limit, as the Pollock limit states, but that in the Irish case, it’s higher than the capacity factor.
To visualize this in a different way, I looked at the annual windpower share as a percent of the annual capacity factor. Here’s that result (yellow line), along with theoretical calculations of what it should look like if the Pollock limit were 100% of the capacity factor (blue/black dashed line), and also what the real limit curve might be (red/black line).
Figure 3. Irish annual windpower share of total generation as a percentage of annual wind capacity factor (“Pollock limit”)(yellow), along with theoretical Pollock (blue-dashed) and possible real-world (red) limits.
In Figure 3, the curves show the situation when as the share of total generation approaches some physical li limit, each addition of wind capacity will make less and less difference as it slowly approaches the limit.
So … is there a limit?
The Irish data strongly implies that such a limit exists. And at least in the case of Ireland, it’s likely higher than the current value of 22.5% above the Pollock limit. Is it on the order of 40% above the Pollock limit as the speculative red curve illustrates? Perhaps. Perhaps not.
The problem is that we don’t really have enough data to say definitely what the limit is for Ireland, or even if such a limit exists. What’s shown in Figure 3 could just be a temporary slowdown … or not. A few more years should make things much clearer.
And that’s what I have found out about the Pollock limit. I have exactly zero idea why Ireland is able to exceed the Pollock limit. The claim is that, absent grid-scale batteries, the Pollock limit is a real physical limit equal to the capacity factor. But that is certainly not the case for Ireland. It’s already 22% above the capacity factor. Why? How?
The reason cannot be economics, because Christopher’s mathematical derivation in his original post doesn’t contain any economics-related terms. (Or alternatively, if economics is the reason, then Christopher’s math must be incomplete.)
All thoughts on that question considered, although perhaps not replied to. So many drummers, so little time …
My best to all, and thanks to Christopher, Lord Monckton for highlighting Pollock’s most interesting theory.
w.
As Always: When you comment please quote the exact words you are referring to. I can defend my own words. I cannot defend your (mis)interpretation of my words. Thanks.
A Footnote Worth Noting: I am an honest man. I do my very best to tell the truth as I know and see it. Yes, I have been wrong, and more than once. And when I’m wrong, I admit it. Heck, I even have a whole post called “Wrong Again“, and to cap that off, another post called “Wrong Again, Again — who does that but a scrupulously honest man?
But despite being wrong at times in matters big and small, wrong far more times in my life than I’d prefer, I do my honest best to not ever lie, shade the truth, misstate facts, or deceive people.
So I’ll thank everyone to avoid accusing me of any of those misdeeds, as it angrifies my blood. And if that happens, it may well result in me conjecturing about the probable species and personal hygiene of some of your recent progenitors … and it also would involve a significant chance of me politely inviting you to engage in anatomically improbable acts of sexual auto-gratification and self-congress …
Willis,
your analysis is interesting, but you missed the more serious consequences of exceeding the capacity limit. In South Australia, they have reached 50% capacity for renewables, with some extreme consequences.
Firstly, the system only continues to work because of hundreds of megawatts of power imported from other states. It’s constantly teetering on the brink. Second, the price of electricity has skyrocketed. This is happening everywhere renewables reaches these giddy heights. Thirdly, SA suffers from what’s euphemistically called demand management. Industry has to shut down in periods of low wind/solar and residential consumers suffer rolling blackouts.
This is incorrect, and perhaps a source of the disupute. Perhaps my previous comment attempting to bring this to your attention was too verbose and it was buried. I’ve lead with it here, and tried to keep it short. My apologies for obfuscating, but I believe your statement to be in error.
Because it doesn’t. You’ve miscalculated Pollack for the reason stated above. These Irish (nor your previous Scottish wind data) cannot be used to compute Pollock. One must have hourly average data for all the power sources on the grid segment/system in question in order to compute the Pollock limit per Monckton’s description. Annualized and averaged data from public information sources is both likely to be misleading and not precise enough to be useful.
Irish data are available at half hourly resolution. They have very explicit rules and future assumptions about the maximum non synchronous generation they will tolerate, allied to investments in batteries, stations, extra grid capacity, synchronous condensers etc. to provide stability. If their is a weakness in their work, it is likely to be mainly in their assumptions about being able to use rising amounts of interconnector capacity to assist with grid balancing. They have already made clear to politicians that costs of higher renewables penetration are rising more and more rapidly and the latest political target of 80% is not known to be feasible. I have quoted from and linked to Eirgrid studies to that effect.
If the politicians don’t take notice they will kill their golden goose of providing huge numbers of data centres, which will move elsewhere if electricity is too expensive. Unfortunately Irish politicians have a track record of making a mess. They created the property boom and bust that ended with half built tumbleweed estates, vast numbers of unemployed and bankruptcy necessitating bailout after the financial crash. They are wide boys, thinking they can ignore economic reality. Their status as a special EU corporate tax haven (effectively granted so they could pay back their post financial crash International borrowing) is coming to an end. Add in mistaken energy policy, and there will be more hard times ahead.
Most assureadly true! I am completely confident that those data exist. Those were not what Willis presented, even if what he presented here was somehow derived from that same raw source. Any modern power generation system MUST have or develop such data as you describe in order to operate.
The questions therefrom usually being, are those data published, are the published data reliable, and are the data properly quoted or are they misused for propaganda or fraud?
Search for my own earlier, overly verbose comment on this thread. Others before Monckton on Pollock have pointed out the potential errors and obfuscation deliberately created when applying standard generation engineering capacity factor and nameplate capacity ratings to irregular and/or intermittent, unpredictable generators, mostly to wind and solar. Pollock, and the year-old post from Stein and Stacey on WUWT (link on earlier comment) have proposed different ways of correcting for the uncertainty of irregular (“unreliable” per Monckton is more than justified), and have seemingly arrived at similar results. (Seperate theoretical methods broadly in agreement sounds awfully sciency to me).
Repeating: Neither Pollock per Monckton nor Stein and Stacey’s MLQ ratings can be considered equival to annualized average electrical generation capacity factor or to electrical share derived from solar or wind generator name plate ratings. Such critiques are false premise (strawman?) argument.
I will re-read your previous comments re Eirgrid. Thanks for bringing those comments to my attention.
Coontinuing to follow this discussion, I suspect the problem is a very simple one. If I am understanding it!
There seem to be two quantities:
the capacity utilization percentage of wind, that is, the percent of faceplate that an operator generates over a year
the percent of total demand that is generated by wind from a combined wind and conventional installation
Christopher, and perhaps Pollock also, seems to think that there is a logical relation between these two quantities which means that under all weather patterns the percentages must be about equal.
In fact if there is a relation its empirical and only holds under some patterns of weather. If you have the right weather pattern (and this may be the most common one) then the rough identity will hold.
If this is right then when talking to policy makers the thing to do is not focus on the alegebra, but on the weather. Explain to them that in this particular place, with this sort of annual weather pattern, you cannot realistically get to more than about 30% of your supply from wind and solar.
To get beyond that you will have to overbuild to a ridiculous extent and also install some form of storage at unaffordable cost. But the thing that is driving this is the weather, when the wind blows ahd how strongly it blows over the course of the year. This is what drives the relationship between average annual capacity utilization and the percentage of total annual demand that the intermittent installations, with that capacity utillzation, can deliver.
I am not certain of having got this right, but this is how it looks to me, if its wrong please tell me.
Would imagine that this problem could be solved if you have a customer for the electricity with a highly adjustable consumption rate.
A customer that bought all the excess power to a somewhat reduced price.
One such customer could be someone into hydrogen electolyze.
Hydrogen is needed in many industrial processes.
When you start to look at the prospects for hydrogen by electrolysis they are not bright. Some will happen because governments are determined to waste large subsidies to make it happen. But electricity surpluses are going to be very intermittent, and very variable in size. That means it will not be viable to try to provide capacity to take all the larger surpluses, and that even the economics of the smaller ones will be stretched by low capacity factors and intermittency reducing plant efficiency. Producing hydrogen by electrolysis already starts at a big cost disadvantage to doing so by steam reforming of methane, which is how almost all industrial hydrogen is produced. Moreover, reduced prices on output (zero or negative?) just mean that consumers will have to pay a higher price for the power they do consume if the wind farms are not to go bankrupt. Alternatively, you increase the size of the direct subsidy for hydrogen production in order to ensure a more continuous supply of power at an unsubsidised price – but that does nothing to solve variable surpluses, and only serves to make them bigger because you are adding baseload demand which requires more capacity and hence bigger surpluses and deficits as the wind varies.
Lord Monckton’s proof of the Pollock limit is a mathematical card trick, not a proof.
Do you remember the riddle where a group of 10 traveling salesmen arrive at a motel that has only 9 rooms available? They all want a room to themselves in case they get lucky with the innkeeper’s daughter. The desk clerk puts two guys in the first room, promising one of them he’ll come back for him. Then he puts the third guy in the second room, the fourth guy in the third room, … until he puts the 9th guy in the 8th room. Then he goes back to the first room and gets the salesman who was temporarily doubled up and puts him in the 9th room. Voilà!
If you recite this story quickly enough, most people won’t spot the trick. Here’s where Monckton pulls off a similar sleight of hand:
“It follows that the minimum installed nameplate capacity N < C of renewables required to generate the fraction f of total grid generation actually contributed by renewables – the renewables fraction – is equal to f C, which is also f H / R ex-ante.”
You see how he specifies in bold that N must be less than C, i.e. no over-building. But there is no mathematical reason why this must be so. Economic, sure, but this is alleged to be a mathematical proof. After he does his arithmetic he gets fmax = R. But because of the specification that N must be < C, he just proves trivially that the maximum value of a variable constrained by the language of the problem to be no more than R is, surprise, surprise, R. The trick is in the statement that N < C.
There is, of course, no mathematical Pollock limit because N can in principle be any fraction or multiple of C that the grid designer is willing to pay for.
Francis Menton of Manhattan Contrarian identified the problem with that ambiguously worded N < C statement but I believe my modest contribution is the recognition that it’s just a trick. There is no proof at all.
Being so far last of a long line who have presented something like this as an argument, perhaps you can succeed where they’ve failed and help me to find where it is described how building more wind turbines of ever greater capacity causes precisely enough wind to blow over the ever expanding wind turbine “field” during, or better just before, those times when electrical energy is needed?
It seems to be a common belief that more and bigger wind turbines somehow do this.
Rainmakers were once a thing, taking money for the ability to bring rain to parched farms. Perhaps we should name the faithful of this church as Windmakers?
It has been done to death that wind power being intermittent and unable to be summoned when you need it can’t work as a sole power source no matter how many windmills you deploy and how big they are. But that’s not a mathematical constraint. It just reflects the reality that the wind doesn’t always blow. That’s not the point.
Monckton was claiming that Pollock had mathematically proved that you can’t deploy more windmills than limited by the average capacity factor. That is like the apocryphal proof that bumblebees can’t fly because they violate the laws of aerodynamics, or the mathematical proofs that a steamship couldn’t cross the Atlantic Ocean because it would burn more coal than it could carry, or that heavier-than-air flight is impossible. In fact he has not proved anything at all. It’s the empiric observations of what happens when you try to close the intermittency gap by building more that shows the folly. Not arithmetic.
Yes, I know Monckton has argued at great length that he really meant that it applied to the self-evident (to him, after the fact) that he was placing the constraint to avoid expensive and wasteful overbuilding. But in doing so he admits that it’s not an arithmetic proof, merely an empirically observed one.
I’m sorry I can’t help you find out how building more windmills solves the intermittency problem because they don’t. But it’s not because of some mathematical card trick that just proves R = R..
Wind power capacity factor is limited by how much the wind is blowing. This is unlike the capacity factor of reliable and dispatchable energy sources where capacity factor and nameplate rating are used for planning and engineering, and when appropriate fuel supply is assumed.
“can’t help you find out how building more windmills solves the intermittency problem”
It’s okay, wasn’t really worried. Pollock, per Monckton, has shown that it is theoretically impossible to “solve the inetermittency problem”. This was precisely the point. And be happy, you are in with a great congregation of the faithful.
Its because of the distribution of the wind strength. Plot wind strength by hour across the year, and the shape of this distribution will determine how much supply you can get.
I don’t know where to get the data, or even how to do the analysis if having it, but it seems likely that its the shape of the distribution that will be the driving factor of how high a percentage of load you can satisfy from the intermittent source.
If this is right, Irish wind should be a flatter distribution than another example case where you get a lower contribution, one nearer the Pollock limit.
Maybe the Pollock limit is valid, but only for a certain case of wind distribution?
Thanks Willis for providing data to this Pollack limit debate. In any case this is being tested in many countries so more data is on the way.
We can already see how well it is going with Europe’s industry for its low cost Russian energy supply to be cut and replaced by 4x more expensive US LNG. The sweetest irony of all is that this externally imposed energy bankruptcy has made it impossible for Germany to supply its Leopard 2 tanks to NATO in ukraine. Their massively increased energy costs mean that Germany can no longer make tanks at a price at which anyone including their own government will buy them. And these tanks were popular. Thus Germany have said no at Rammstein to any Leopards for ukraine. This condemns them to a (slightly earlier) defeat. You can’t chase green energy fantasies and also win wars.
(Oops did I mention the war??)