The books is good(ish) for it's time, but some of its analysis and forecasts are fundamentally flawed since it falls for the primary energy fallacy by comparing the chemical potential energy (in J) directly to electrical energy (also in J). The two are fundamentally different things and called be compared 1:1. E.g. To heat up your home with natural gas you need ~1J of chemical energy to get 1J of heat into the home, but with a electric heat pump you only need 1/6 J to get 1J of heat.
It is also a product of its time in terms of wind/solar vs nuclear. His forecasts of the impact of solar and wind is based on prices and performance from 2008. Prices have come down an order of magnitude since then, and performance and lifespan have increased drastically.
Heat pumps, and their efficiency (getting more than X Joules of heating for X Joules of electricity) are discussed in chapter 21 Smarter Heating, see for example the diagram and discussion on page 150: https://www.withouthotair.com/c21/page_150.shtml
I imagine the economics (and perhaps technical improvements) of some things may have changed more than the book could forecast. But heat pumps were definitely understood by the author.
I’m not saying he was unaware of heat pumps, I’m saying he’s comparing apples to oranges simply because they use the same units.
Look at “3 - Cars” on page 29. He says the typical car uses 40 kWh/day. 40 kWh of what? Chemical energy in the gasoline.
The go to page 33 where he looks at how much energy onshore wind could produce per days in the UK. His number is 20 kWh/d. 20 kWh of what? electricity
He then compares those two numbers directly and uses that comparison as the basis of his arguments: “Britain’s onshore wind energy resource may be “huge,” but it’s evi-
dently not as huge as our huge consumption.”
This is simply incorrect. A combustion engine converts less than half of the chemical energy in the gasoline into mechanical work that can move the car. The electric model converts >90% of it. So we don’t have to replace 40 kWh/day, we have to replace less than half of that since the electric process is more efficient.
This same issues, the primary energy fallacy, underpins large parts of the book.
He does correct for the efficiency of both electric cars, and heat pumps in later chapters. His line of argument is "here is the current energy supply and demand - now here are ways we can increase one and reduce the other".
The wind analysis was reasonable for the time, iirc the main error there was that he failed to foresee offshore wind dropping in price so much.
I disagree. The framing of focusing on primary energy and then sprinkling efficiency afterwards has two issues: 1) it pulls focus away from the thing we want (the end result) and 2) it makes it much easier to misunderstand (or to misrepresent) because the efficiency differences can be omitted.
And graphs like the one comparing wind turbine output to petrol car consumption are inherently deceiving. Two values are put side by side with the same units and then talk about directly as if they are comparable. But they simply are not. A kWh of chemical energy and a kWh of electricity have as much in common as a US dollar and a Jamaican dollar.
I agree with you. I read it cover to cover at the time and felt like a smart ass ("AksHuALLy renewables can't work alone!"), then wondered a few years ago how his calculations turned out in the real world.
1. Sad that he died from cancer
2. Noticed he compared primary energy and felt stupid for missing that
I never got a blanket claim of "actually renewables can't work alone" from the book. It seemed more like "actually you need to understand things quantitatively or you are doomed to talk nonsense".
MacKay's "renewables can't work alone" claim always seemed carefully scoped to the economics of 02008 (when solar modules cost 33× as much as they do now) and also his own densely populated, rather polar country. Renewables were already working nearly alone, at scale, in equatorial countries and more sparsely populated countries; I live in Argentina, whose grid was mostly hydroelectric at the time, and next door to Brazil, where a large fraction of the automotive fleet ran on sugar-cane-derived ethanol, which is a viable renewable energy source (unlike, apparently, corn ethanol).
I remeber the book differently - I remember it being primarily about getting exactly this right. Lot of space spent discussing efficiency and losses and comparing apples to apples.
(Admittedly, I read it many years ago, my memory may be off, though I strongly doubt it.)
Thermal energy and electrical energy really are the same thing, not two different quantities that happen to be measured in the same units, like grams of lead and grams of gold, or your example of US dollars and Jamaican dollars. When you convert less than half of the chemical energy in the gasoline into mechanical work to move the car, the other 50+% of the energy is converted into heat. Carnot gives us a reversible conversion factor between them, but it depends on the combustion temperature rather than being some kind of constant, as in your Jamaican-dollar example.
Technically you can interconvert grams of lead and grams of gold one to one, too, but interconverting electrical and thermal energy is so easy that it happens all the time unintentionally.
> In the last chapter, we learned that electrification could shrink transport’s
energy consumption to one fifth of its current levels; and that public trans-
port and cycling can be about 40 times more energy-efficient than car-
driving. How about heating? What sort of energy-savings can technology
or lifestyle-change offer?
And then he goes into not just household heat pumps, and their achievable coefficients of performance, but also municipal combined heat and power, which take that ≈50% of the chemical energy "lost" from thermal power plants as waste heat and pumps it into your house.
So, far from being ignorant of the issue as you seem to be implying, he presents a more complete picture of the issues than you are presenting.
Specifically for household climate control, I suspect that both heat pumps and CHP are much less relevant now that we have cheap solar. You can think of a heat pump as a way to reduce the amount of solar-panel area that you need to heat your house. The trouble is that solar panels cost €0.13 per peak watt, while heat pumps cost closer to €1 per peak watt, so it may be cheaper to "waste" energy on heating your house resistively with a nichrome wire than to use a carefully engineered heat pump.
For the purposes of energy use, they are not. Heat is fundamentally worth less than the equivilent amount of electrical energy, due to these efficiency losses (which are not just a technology problem). It's misleading to compare them at any point in your analysis unless to highlight the problem with the comparison.
I think in 2008, it wasn't clear whether cars are going to be replaced by electric and what the final efficiency will be. It could also have been more because you would need to convert electricity to fuel first.
So I think as a conservative estimate, it kinda works.
Cost would depend on battery manufacturing but the original founders of Tesla created spreadsheets of well-to-wheel efficiency based on available data in 2002 and knew that electric came out as an obvious winner and so commited to starting the company based on that.
A lot of companies started in 02002 based on spreadsheets of available data went bankrupt due to the available data being wrong, or due to things not included in the available data at all. A quick check of Wikipedia turns up:
And those are cherry-picked from companies that got big enough to be "notable" by Wikipedia's guidelines. Far more companies never reach that level.
The conservative estimate is that any time someone tries to deploy a new technology, it will fail. Most innovations (new technologies or companies) do.
Take a look at the HN front page from 10 years ago: https://news.ycombinator.com/front?day=2016-09-28 There we see Uber's self-driving truck initiative Otto (an acquisition), deep learning startup Skymind, a new static Linux distro called Stali, and a bunch of things that weren't innovations. All three of those things failed; Stali hasn't been updated since 02019. LuaTeX, Coinbase, Linux, Wikipedia, and dyeing with indigo are other innovations that feature on that page, but they were already very old.
It doesn't change the physics though. I think the author under discussion actually gets this right in some chapters with regard to the superior efficiency of EVs, though not sure if that was a later revision or not.
That's true! What was holding up EVs 18 years ago, or for that matter 18000 years ago, was not physics, but the humans' ability to efficiently make things such as batteries and high-efficiency electric motors.
> In the last chapter, we learned that electrification could shrink transport’s
energy consumption to one fifth of its current levels; and that public transport and cycling can be about 40 times more energy-efficient than car-driving. How about heating? What sort of energy-savings can technology
or lifestyle-change offer?
and it goes on to talk about heat pumps, as today. So I think that even versions from 02008 got this right, though evidently that was at least the 13th numbered draft.
The average Brit drives under 20 miles per day. A Tesla will do 4 miles per kWh, meaning it requires 5 kWh per day. That's 1/8 of 40 kWh. I'm not sure where the discrepancy is.
Yes, I disagree with some of his takes but he was spot on regarding heat pumps:
> Let me spell this out. Heat pumps are superior in efficiency to condens-
ing boilers, even if the heat pumps are powered by electricity from a
power station burning natural gas. If you want to heat lots of buildings
using natural gas, you could install condensing boilers, which are “90% ef-
ficient,” or you could send the same gas to a new gas power station making
electricity and install electricity-powered heat pumps in all the buildings;
the second solution’s efficiency would be somewhere between 140% and
185%. It’s not necessary to dig big holes in the garden and install underfloor
heating to get the benefits of heat pumps; the best air-source heat
pumps (which require just a small external box, like an air-conditioner’s)
can deliver hot water to normal radiators with a coefficient of performance
above 3.
Which are you referring to as the primary comparison?
If I'm understanding the book's structure correctly, the comparison at the end of Part 1 (chapter 18 "Can we live on renewables?") is based on estimating existing demand. Deployment of heat pumps in Britain even today is woefully low, and when the book was written it was no doubt even lower, so of course it wouldn't make sense to include heat pumps as a significant factor when comparing actual demand when the book was written with potential supply.
Chapter 27 "Five energy plans for Britain" discusses large scale ways to make the energy budget "work", and those hypothetical comparisons do include use of heat-pumps. Maybe those numbers are inaccurate and a better hypothetical could be produced today, but I would first put that down to the difficulty of forecasting the future.
It's been a long time since I read the book fully though. Did I miss something?
When asking that question it's a mistake to assume no technological change/improvement. The process that would require the adaptation would drive changes in the technologies.
> His forecasts of the impact of solar and wind is based on prices and performance from 2008. Prices have come down an order of magnitude since then
This is going to sound like one-upmanship or nitpicking, but I think it's important to know that this understates the change. https://web.archive.org/web/20100722072720/http://www.solars... says that in May 02009, crystalline solar cells in China cost €2.17 per peak watt, a decline of 26.4% from January 02009, whose price is not listed directly but which we can calculate as €2.95 per peak watt. If the price had declined by only an order of magnitude since January 02009, it would now be €0.295 per peak watt.
But, according to https://www.in2013dollars.com/europe/inflation/2009?amount=1..., "€100 in 2009 is worth €145.57 today." So, if we adjust for inflation — as we should — if the real price had declined by only an order of magnitude since January 02009, it would now be €0.429 per peak watt.
So the price of solar modules now is actually 3.3 times lower than what a careful reader would infer from your remark. Solar module prices have declined since January 02009 by not merely an order of magnitude but 33×. That's slightly closer to two orders of magnitude than to the single order of magnitude you said. Now they are at 3% of the price they were at when MacKay wrote his excellent book.
(Which does, as others have pointed out, explain the efficiency advantage of heat pumps.)
one thing to note is that the solar industry has receded quite a bit in the last two years in China due to overproduction (or, as the Chinese state likes to call it, 'involution' or 'too much competition'). solar is cheap now because 1) Chinese subsidies are reduced, shrinking their installation market and 2) overproduction leading to essentially fire sales. once production stabilizes against the market, costs will go back up. it's really never been a better time to buy panels than now
>Some observers estimate 20-30% capacity reductions are required, but demand shrinkage could increase that figure. “I’m particularly worried about the downstream side,” says Reis. “Even as prices collapsed, demand remained robust for three years, but power market reforms are now upending the entire renewables market.”
combine this with the global backing-off of the Paris Agreement (and subsequent subsidy defunding) and power utilities in the US abandoning clean energy goals in order to immediately increase output for proposed hyperscale facilities and you've got something of an enduring demand shortage
that said, there is a lot of automation happening in that industry right now (which is likely going to cause a small recession in China due to half a million to a million workers being out of a job, depending on how much the parts manufacturers also automate) so it's possible production with continue improving as more R&D is dedicated there. only time will tell!
Risking the HN hug of death here, but my own prototype of a car miles model inspired by the David's approach can be found at https://trafred.cardiff.ac.uk/
Oh, very interesting. Not sure it's working correctly though: set everything to ambition level 4, which takes us to net zero minus 4. Then, the following knobs have no effect: efficiency, nuclear power, carbon intensity, solar (!) and wave & tidal. That can't be true, so I think it's buggy
Not commenting on the physics of this; I want to point out that I thought the narrative structure of this book was amazing. I have not read a hard-core piece of analysis in any field before or since that I thought was a page-turner the way this was. When I read it in 2008 I remember completing each chapter and feeling like "we're screwed" and then the next chapter feeling like "ok, we're going to make it." Presenting such analysis in the form of a race between supply and demand was narrative genius and I think should serve as an example of how ideas can be presented in a way that is both exciting and not-dumbed-down.
Dear god, this page is like an argument against every person who has ever said "graphic design doesn't matter".
The colors are gross, it took me ten minutes just to figure out what I was looking at (a book), and heaven help me if I actually wanted to read that book in that format!
The untimely death of David MacKay was a tragic loss for the scientific community. It's astonishing that he wrote the textbook on information theory for the ML community as well as his contributions to rational discussion about energy policy.
He had a direct personal impact on my life since my introduction to ML came from video recordings of his lectures on ML and information theory. I am grateful to have learned from such an excellent teacher. Like a Feynman of our times.
That book significantly decreased my appreciation of our statistics professor @uni, because it presented the same topics in a much more interesting and digestible way. Can recommend.
That book and Gershgorin circles (to work around all the stupid, constructed 3x3 matrix eigenvalue problems that they love to sprinkle into literally every early stem exam) helped more in university than anything else.
1. Gasoline powered energy metrics do not translate to electric vehicles, due to the much higher efficiency of electric motors.
2. There are about twice as many people as cars in the UK so the comparison between 40 kwh/d per car and 20 kwh/d per person for the wind energy is at minimum misleading.
Worth reading even with the dated numbers. MacKay's lasting contribution isn't any specific forecast - it's the discipline of expressing everything in the same units (kWh per day per person), which makes energy arguments commensurable instead of vibes-based. Once you've internalized that, most energy coverage in the media starts reading like category errors. The specific figures have moved - solar costs, EV adoption, heat pump uptake - but the arithmetic habit hasn't aged at all.
It is also a product of its time in terms of wind/solar vs nuclear. His forecasts of the impact of solar and wind is based on prices and performance from 2008. Prices have come down an order of magnitude since then, and performance and lifespan have increased drastically.
I imagine the economics (and perhaps technical improvements) of some things may have changed more than the book could forecast. But heat pumps were definitely understood by the author.
Look at “3 - Cars” on page 29. He says the typical car uses 40 kWh/day. 40 kWh of what? Chemical energy in the gasoline.
The go to page 33 where he looks at how much energy onshore wind could produce per days in the UK. His number is 20 kWh/d. 20 kWh of what? electricity
He then compares those two numbers directly and uses that comparison as the basis of his arguments: “Britain’s onshore wind energy resource may be “huge,” but it’s evi- dently not as huge as our huge consumption.”
This is simply incorrect. A combustion engine converts less than half of the chemical energy in the gasoline into mechanical work that can move the car. The electric model converts >90% of it. So we don’t have to replace 40 kWh/day, we have to replace less than half of that since the electric process is more efficient.
This same issues, the primary energy fallacy, underpins large parts of the book.
The wind analysis was reasonable for the time, iirc the main error there was that he failed to foresee offshore wind dropping in price so much.
And graphs like the one comparing wind turbine output to petrol car consumption are inherently deceiving. Two values are put side by side with the same units and then talk about directly as if they are comparable. But they simply are not. A kWh of chemical energy and a kWh of electricity have as much in common as a US dollar and a Jamaican dollar.
1. Sad that he died from cancer
2. Noticed he compared primary energy and felt stupid for missing that
MacKay's "renewables can't work alone" claim always seemed carefully scoped to the economics of 02008 (when solar modules cost 33× as much as they do now) and also his own densely populated, rather polar country. Renewables were already working nearly alone, at scale, in equatorial countries and more sparsely populated countries; I live in Argentina, whose grid was mostly hydroelectric at the time, and next door to Brazil, where a large fraction of the automotive fleet ran on sugar-cane-derived ethanol, which is a viable renewable energy source (unlike, apparently, corn ethanol).
(Admittedly, I read it many years ago, my memory may be off, though I strongly doubt it.)
Technically you can interconvert grams of lead and grams of gold one to one, too, but interconverting electrical and thermal energy is so easy that it happens all the time unintentionally.
MacKay does in fact cover the Carnot factor you're talking about; his Chapter 21 http://www.withouthotair.com/c21/page_140.shtml begins:
> In the last chapter, we learned that electrification could shrink transport’s energy consumption to one fifth of its current levels; and that public trans- port and cycling can be about 40 times more energy-efficient than car- driving. How about heating? What sort of energy-savings can technology or lifestyle-change offer?
And then he goes into not just household heat pumps, and their achievable coefficients of performance, but also municipal combined heat and power, which take that ≈50% of the chemical energy "lost" from thermal power plants as waste heat and pumps it into your house.
So, far from being ignorant of the issue as you seem to be implying, he presents a more complete picture of the issues than you are presenting.
Specifically for household climate control, I suspect that both heat pumps and CHP are much less relevant now that we have cheap solar. You can think of a heat pump as a way to reduce the amount of solar-panel area that you need to heat your house. The trouble is that solar panels cost €0.13 per peak watt, while heat pumps cost closer to €1 per peak watt, so it may be cheaper to "waste" energy on heating your house resistively with a nichrome wire than to use a carefully engineered heat pump.
So I think as a conservative estimate, it kinda works.
- https://en.wikipedia.org/wiki/MStar (patent infringement, the remains bought by MediaTek)
- https://en.wikipedia.org/wiki/Green_Flash_Brewing_Company (succumbed to competition from local craft breweries, lender foreclosed, the remains bought by private equity)
- https://en.wikipedia.org/wiki/Primaris_Airlines (bankrupt in 02008, unclear why)
And those are cherry-picked from companies that got big enough to be "notable" by Wikipedia's guidelines. Far more companies never reach that level.
The conservative estimate is that any time someone tries to deploy a new technology, it will fail. Most innovations (new technologies or companies) do.
Take a look at the HN front page from 10 years ago: https://news.ycombinator.com/front?day=2016-09-28 There we see Uber's self-driving truck initiative Otto (an acquisition), deep learning startup Skymind, a new static Linux distro called Stali, and a bunch of things that weren't innovations. All three of those things failed; Stali hasn't been updated since 02019. LuaTeX, Coinbase, Linux, Wikipedia, and dyeing with indigo are other innovations that feature on that page, but they were already very old.
Chapter 21 of draft 2.9.3 from 02008 https://web.archive.org/web/20080906132444/http://www.infere... begins:
> In the last chapter, we learned that electrification could shrink transport’s energy consumption to one fifth of its current levels; and that public transport and cycling can be about 40 times more energy-efficient than car-driving. How about heating? What sort of energy-savings can technology or lifestyle-change offer?
and it goes on to talk about heat pumps, as today. So I think that even versions from 02008 got this right, though evidently that was at least the 13th numbered draft.
Much less. Your typical gasoline IC vehicle converts maybe 1/4 of the chemical energy into work.
> Let me spell this out. Heat pumps are superior in efficiency to condens- ing boilers, even if the heat pumps are powered by electricity from a power station burning natural gas. If you want to heat lots of buildings using natural gas, you could install condensing boilers, which are “90% ef- ficient,” or you could send the same gas to a new gas power station making electricity and install electricity-powered heat pumps in all the buildings; the second solution’s efficiency would be somewhere between 140% and 185%. It’s not necessary to dig big holes in the garden and install underfloor heating to get the benefits of heat pumps; the best air-source heat pumps (which require just a small external box, like an air-conditioner’s) can deliver hot water to normal radiators with a coefficient of performance above 3.
If I'm understanding the book's structure correctly, the comparison at the end of Part 1 (chapter 18 "Can we live on renewables?") is based on estimating existing demand. Deployment of heat pumps in Britain even today is woefully low, and when the book was written it was no doubt even lower, so of course it wouldn't make sense to include heat pumps as a significant factor when comparing actual demand when the book was written with potential supply.
Chapter 27 "Five energy plans for Britain" discusses large scale ways to make the energy budget "work", and those hypothetical comparisons do include use of heat-pumps. Maybe those numbers are inaccurate and a better hypothetical could be produced today, but I would first put that down to the difficulty of forecasting the future.
It's been a long time since I read the book fully though. Did I miss something?
This is going to sound like one-upmanship or nitpicking, but I think it's important to know that this understates the change. https://web.archive.org/web/20100722072720/http://www.solars... says that in May 02009, crystalline solar cells in China cost €2.17 per peak watt, a decline of 26.4% from January 02009, whose price is not listed directly but which we can calculate as €2.95 per peak watt. If the price had declined by only an order of magnitude since January 02009, it would now be €0.295 per peak watt.
But, according to https://www.in2013dollars.com/europe/inflation/2009?amount=1..., "€100 in 2009 is worth €145.57 today." So, if we adjust for inflation — as we should — if the real price had declined by only an order of magnitude since January 02009, it would now be €0.429 per peak watt.
In fact, today, "mainstream" solar modules are currently €0.130 per peak watt, according to https://www.solarserver.de/photovoltaik-preis-pv-modul-preis..., and those are Chinese monocrystalline modules.
So the price of solar modules now is actually 3.3 times lower than what a careful reader would infer from your remark. Solar module prices have declined since January 02009 by not merely an order of magnitude but 33×. That's slightly closer to two orders of magnitude than to the single order of magnitude you said. Now they are at 3% of the price they were at when MacKay wrote his excellent book.
(Which does, as others have pointed out, explain the efficiency advantage of heat pumps.)
https://dialogue.earth/en/energy/behind-the-layoffs-in-china...
>Some observers estimate 20-30% capacity reductions are required, but demand shrinkage could increase that figure. “I’m particularly worried about the downstream side,” says Reis. “Even as prices collapsed, demand remained robust for three years, but power market reforms are now upending the entire renewables market.”
combine this with the global backing-off of the Paris Agreement (and subsequent subsidy defunding) and power utilities in the US abandoning clean energy goals in order to immediately increase output for proposed hyperscale facilities and you've got something of an enduring demand shortage
that said, there is a lot of automation happening in that industry right now (which is likely going to cause a small recession in China due to half a million to a million workers being out of a job, depending on how much the parts manufacturers also automate) so it's possible production with continue improving as more R&D is dedicated there. only time will tell!
https://en.wikipedia.org/wiki/David_J._C._MacKay
$10/W to 0.30/W is closer to 2 orders of magnitude than 1.
The underlying fundamentals did not change since 2008. If Solar was good/viable back then, it can only get better if it got cheaper over time.
Same goes for the nuclear debate. Nuclear was losing back then, and now it has finally lost absolutely.
Risking the HN hug of death here, but my own prototype of a car miles model inspired by the David's approach can be found at https://trafred.cardiff.ac.uk/
https://itila.blogspot.com/2015/09/what-do-you-tell-children...
The colors are gross, it took me ten minutes just to figure out what I was looking at (a book), and heaven help me if I actually wanted to read that book in that format!
He had a direct personal impact on my life since my introduction to ML came from video recordings of his lectures on ML and information theory. I am grateful to have learned from such an excellent teacher. Like a Feynman of our times.
Note the .com->.org: its a version of the book whose numbers are maintained here <https://github.com/life-itself/without-hot-air/commits/main/>.
https://news.ycombinator.com/item?id=34618613 ("Information Theory, Inference, and Learning Algorithms (2003) (inference.org.uk)")
That book significantly decreased my appreciation of our statistics professor @uni, because it presented the same topics in a much more interesting and digestible way. Can recommend.
That book and Gershgorin circles (to work around all the stupid, constructed 3x3 matrix eigenvalue problems that they love to sprinkle into literally every early stem exam) helped more in university than anything else.
Two quick points:
1. Gasoline powered energy metrics do not translate to electric vehicles, due to the much higher efficiency of electric motors.
2. There are about twice as many people as cars in the UK so the comparison between 40 kwh/d per car and 20 kwh/d per person for the wind energy is at minimum misleading.