David MacKay’s work was for me very important. It showed how one should do a systems analysis for a country to understand how one can achieve the change required. I was astonished when I talked to several of the leading people responsible for energy policy in several political parties in my country, to find they really had no clue about what they where doing on a systems level.
I think that updating this book is not pointless, but more importantly doing the same for more countries is really important.
It would be fantastic to turn the chapters into Templates where one has to plug in the relevant parameters for a country and see the calculations for that country.
For example, the Stanford/Jacobson study on powering the world with wind/solar/water.
Some results were famously silly, such as filing Finland with solar panels, or adding the cost of nuclear war to nuclear reactors. Jacobson sued other academics then backed down.
The point is, 5-10 years on, these models turned out to be over pessimistic! Australia doesn't need fancy modelling - simple roof top solar there is on target to meet full day time demand in 2-3 years.
Conversely we still dont really know how to meet the last 5% of outlier energy demand scenarios.
> Some results were famously silly, such as ... adding the cost of nuclear war to nuclear reactors.
Is that really silly? The incremental increase in the risk of nuclear war for each gigawatt of civilian nuclear power generation might be small, but the costs of a nuclear war could be quadrillions of dollars (for example if it set the world back by 100 years).
Obviously measuring that risk is difficult, but to give an example, imagine that civilian nuclear power were limited to just the countries which currently have nuclear weapons. That would greatly reduce the proliferation risk and the chance of nuclear breakout in countries like Iran.
The urgency of addressing climate change may take precedence over these concerns in the short term, but if we do manage to reach net zero with a mixture of renewables and nuclear energy, then it's not inconceivable that over-provisioning renewables might allow us to phase out nuclear energy later on.
Suppose it would take 50 years to build up that renewable energy capacity, and 10 years to build a new nuclear power station. That would mean the nuclear power station would only be useful for 40 years, which limits the amount of time they'd have to recoup their costs over.
> Conversely we still dont really know how to meet the last 5% of outlier energy demand scenarios.
Hydrogen w. combustion turbines. When a simple cycle turbine power plant is 5% of the cost of a nuclear plant of the same output, we can back up the entire grid and not have it be too expensive.
> It showed how one should do a systems analysis for a country to understand how one can achieve the change required. I was astonished when I talked to several of the leading people responsible for energy policy in several political parties in my country, to find they really had no clue about what they where doing on a systems level.
Massive +1. And i agree that perhaps "reusing" the book would be useful. E.g. take the key factual approach and update analyses for particular technologies and then plug those together - which to some extent is what the pathway calculator did which is why it is worth trying to get the source for that https://github.com/life-itself/climate/issues/2
> I think that updating this book is not pointless, but more importantly doing the same for more countries is really important.
While MacKay's work is excellent for a general lay audience to understand general scales of the issue, there are now far more advanced models for this that do things like replay historical weather and demand, down to 15 minute scales, to more least cost systems for zero carbon grids.
They are also so advanced as to take into current grid structure and generation resources, and the cheapest way to transition to zero carbon from our current resources.
Christopher Clack's models are probably the most advanced. The latest iteration has the very surprising finding that deploying lots of grid-edge solar and storage right now will save us a ton of money because it will use current transmission and distribution assets more effectively and reduce the need for future investment in these assets (T&D is the majority of our electricity bills, not generation!)
> Wind turbines are getting bigger all the time. Do bigger wind turbines change this chapter’s answer?
> Chapter B explains. Bigger wind turbines deliver financial economies of scale, but they don’t greatly increase the total power per unit land area, because bigger windmills have to be spaced further apart. A wind farm that’s twice as tall will deliver roughly 30% more power.
So, did wind turbine production actually increased by an order of magnitudes since the book ?
Annual additions gave gone from sub-GW to 70GW last year
And with turbine spacing of 1km you can actually do stuff with the land in between.
On the other hand, transmission projects still take 10-20 years, and new transmission projects are often over booked by 300%-400% before they are even started.
Electric cars are ridiculously efficient. A 100kwh tesla drives approx 500km. So 50km/day/car needs just 10kwh. The UK also has 2 persons per car (plus additional cats and dogs..)
So just the on-shore book projection is already enough AND only requires a tenth of the turbines.
Lets keep nuclear as well. Invest in efficiency. Lots of hard work ahead.
But we are probably closer to "problem solved" than to "civilization collapse". Cheers.
I am less concerned about the consumption, but the issue how much available wind energy we can harness in a best case scenario. Only half of transportation seems so low.
For what it's worth "In 2019, the average car in the UK drove 7,400 miles"[0] which works out to 33 km/day/car. It's probably much less now due to working from home (and fuel shortages, and the collapse of the supply chain...)
Okay, but GP's point was that these bigger turbines need to be spaced further apart instead, so we can place fewer of them. That we could build bigger turbines was not in question I think. Do you happen to know whether the MW per km² increased more than the 30% expected amount that GP cited?
And regarding doing things between the turbines, I wasn't "awake" yet in 2006 so I don't remember how it was then, but were they ever placed so close together that you couldn't have farmland in between turbines? The problem seems to be that people don't want to live near them and find nature filled with wind turbines ugly, not so much that you can't do anything else in between.
(Just to note, I don't find them that ugly (even if, of course, nature would be prettier without them... but that's not an option) and I also didn't find them to be very loud when I visited some nearby wind turbines. Perhaps it's different at night when it's all quiet, but personally I don't think I'd mind living next to one. The blades are a bit scary though, I can't help but imagine the consequences if one of them lets loose... but that's like being afraid of air travel I guess.)
> Okay, but GP's point was that these bigger turbines need to be spaced further apart instead, so we can place fewer of them.
What will be a problem if our energy consumption increases by an order of magnitude or two, and we take all of it from the wind. Currently, the entire question is about financial ROI.
> but were they ever placed so close together that you couldn't have farmland in between turbines?
Not since the modern turbines were invented. The first ones were placed in somewhat compact lines, with enough spacing between the turbine lines that you could raise a few lines of any crop. (The lines can't be too close anyway, as that would reduce the turbine efficiency and harm financial ROI.)
Because we have trouble finding spots where people don't already live and file complaints to prevent them being built near them? At least on-shore; off-shore is more expensive. If we can get more MW per km² then we can have more power with less trouble.
The goal of the book is not to find quick wins in reducing CO2 output, the goal is to give a complete overview over what is necessary to achieve carbon neutrality. And MW/km2 might be a bottleneck if we're to replace all oil & gas with electric power. If you read the book you can actually see what percentage of the seas around the UK would have to be fully covered, to achieve carbon neutrality. If that number really is outdated by an order of magnitude, perhaps MW/km2 is not a bottleneck.
Thinking about land required is more familiar to the public. But it does lead to these broad statements about feasibility that don't necessarily match up with reality. Real schemes are just going to be more complex and will look at lots of different factors. There is always a limited funds and you have to pick the best opportunities. That is true of very long running projects like nuclear also. Different sites will have different grid connections, environmental factors, ground conditions, access routes, labour force etc.
Yes, very good point. Renewable enegry generation really isn’t the struggle these days, it is transmission from areas with high solar and wind potential to areas with less viable solar and wind production.
They've gotten cheaper by an order of magnitude, but their basic physical characteristics haven't changed that much. A big point of the book is to show what it looks like to really deploy this stuff at the necessary scale.
The issue is that we now have real experience of deploying this stuff at scale and integrating intermittent sources into grids. Many countries in Europe have 30% or more of their electricity supplied by intermittent but it turns out that the integration problem isn't as difficult as envisaged. The consensus amongst engineers is that up to about 70% is manageable without any big tech breakthroughs.
Network reliability has actually increased in many European countries despite the increasing reliance on intermittent sources.
The current approach certainly depends on having back-up deployable generation capacity (natural gas) but that was always required for demand following even with thermal sources so most grids already have the infrastructure to incorporate more intermittent generation.
But assuming no technology advances in this area for 30 years time seems implausible. What has happened with price/efficiency of solar, wind and grid-scale batteries in the last 10 years provides a lesson. None of these were practical even 10 years ago and now they represent 90% of newly installed capacity globally (2020). And this is happening all-over whether with or without government support.
For example, new-build, grid-scale battery storage is now cheaper on an LCOE basis than new build open-cycle ("peaker") natural gas plants. This has only happened recently - about a year ago. I was completely skeptical that this could happen a few years ago.
I urge any/all hackernews readers to read more about _recent_ developments in de-carbonizing energy. It's fascinating and it has become clear to me that we're living through one of the great technology revolutions of human history. It genuinely is a Kodak/smartphone type moment with even bigger implications for human wellbeing.
It's amazing that by being cheap enough, all the disadvantages of a new technology become less relevant. A bit like how the PC displaced "real computers" (i.e. mainframes) even though at the time they were little more than microcontrollers integrated with a terminal.
Germany has a large share (50%) of renewables and their specific CO2 emissions per kWh are up to ten times higher as compared to France with 70% nuclear.
Volatile renewables don’t effectively help to reduce greenhouse gas emissions.
"Volatile" renewables absolutely do help reduce greenhouse gas emissions.
As the renewables share of generation has been increasing, the CO2 intensity of electricity generation in Germany has dropped from 542gCO2/kWh in 2000 to 296gCO2/kWh in 2020[1].
Btw, your link brings me to a table on agricultural CO2 emissions?
> Many countries in Europe have 30% or more of their electricity supplied by intermittent but it turns out that the integration problem isn't as difficult as envisaged.
True. It’s very easy when you can tap your neighbour’s power plants when you need. The problem is a bit different when all the neighbours have the same problems at the same time.
Of course interconnection is a big help, but transmission isn't "free" or unlimited, the flows between individual European countries is typically a small fraction (less than 10%) of the national consumption and generation.
And there are cases like Ireland which has its own grid - with very limited external interconnection - and yet achieved 42% renewable share last year.
Techniques/engineering/theory for integrating intermittent sources has advanced considerably in the last ten years.
> Of course interconnection is a big help, but transmission isn't "free" or unlimited, the flows between individual European countries is typically a small fraction (less than 10%) of the national consumption and generation.
It’s more than a little help. Just imagine how a blackout involving 10% of the population or Germany would look like, never mind the EU.
> And there are cases like Ireland which has its own grid - with very limited external interconnection - and yet achieved 42% renewable share last year.
You cannot really compare Ireland with many of the other EU countries, either in term of population or industry, though.
> Techniques/engineering/theory for integrating intermittent sources has advanced considerably in the last ten years.
They have. And it is a good thing, because it really should not be a competition between nuclear and renewables. They still have several issues, which are very difficult to solve without either a massive reduction in consumption, or a massive increase in price.
I really struggle to understand how almost all mainstream “green” parties see fossil fuels as a lesser menace than nuclear. The truth is, if you read their manifesto, that they don’t care about particle pollution beyond “cars are bad” (which is true, but insufficient) or climate change.
It is not a competition, it is a struggle to get rid of fossil fuels, which are an existential threat to our societies. We can sort that out and talk about fusion once we’ve done it. In the meantime, getting rid of fossil fuels is a massive undertaking, and we would be stupid not to use every card in our hand.
We live in a world of finite resources. When considering the significant task of decarbonizing energy, there are lots of technologies potentially available. But some deliver more value than others.
Unfortunately nuclear is not _currently_ one of the options which delivers value in this regard.
Flamenville 3 - the newest French reactor will cost $22B to build 1.6GW of capacity and will have taken nearly 20 years from the start of construction to deliver a watt. The same amount of money could deliver over 20GW of wind capacity which could be deployed almost instantly as it's an easily parallelizable low-risk, relatively low-tech engineering task relying on mass-production for most of the installation. None of these benefits are available to nuclear reactor construction. Even with a capacity factor of 40% for on-shore wind vs 90%+ for nuclear, nuclear just isn't close.
I'm not against nuclear. If the nuclear industry can step up and deliver reactors on time and on budget which can provide power at competitive prices, then I'd be all in favour. It would be wonderful to have an alternative in the race to decarbonize energy. But all I see is a long trail of massively delayed or abandoned projects with incredible cost-overruns (Flamanville, Olkiluoto, V.C. Summer, Vogtle, etc). These failures are hard-engineering failures, nothing to do with politics.
With such a litany of recent failures, the incredible capital expense, the extra security required, the slowness of delivery, it's clear why resources have been directed away from nuclear toward other technologies which are quick and easy to roll-out and have already significantly contributed to reducing the the CO2 intensity of Europe's electricity (which is now less than half of what it was 30 years ago).
Btw, it isn't just nuclear that loses in this regard; bio-fuels, domestic-solar PV, green hydrogen, etc. all fail to compete in the new world of ultra-cheap solar PV and wind.
Nuclear is bottlenecked by the availability of properly trained engineers (and you really need those for such projects). The need in resources would not impact installations of either wind turbines or solar panels.
Nuclear+renewables is not more expensive than all renewables, particularly if you consider the fact that at the moment renewables benefit from an industry that is entirely dependent on subsidised fuel. And that nuclear reactors would be much cheaper per unit if we stopped building only a couple of each class. Also, a nuclear power plant is expensive, but so is the equivalent capacity in wind turbines plus the required storage.
A nuclear baseline mitigates the issues of wind turbines, and wind turbines are a good complement to nuclear. Both are mostly mutually exclusive for political reasons.
> Flamenville 3 - the newest French reactor will cost $22B to build 1.6GW of capacity and will have taken nearly 20 years from the start of construction to deliver a watt. The same amount of money could deliver over 20GW of wind capacity which could be deployed almost instantly as it's an easily parallelizable low-risk, relatively low-tech engineering task relying on mass-production for most of the installation. None of these benefits are available to nuclear reactor construction. Even with a capacity factor of 40% for on-shore wind vs 90%+ for nuclear, nuclear just isn't close.
The same reactor was built on time and on budget in China, and is currently operating. Hinckley point is also going more or less as planned. Areva/Orano’s abysmal performance is more related to their mismanagement than the reactor itself. You also need to factor stuff like storage, and the fact that we have a limited shore length to use.
> I see is a long trail of massively delayed or abandoned projects with incredible cost-overruns (Flamanville, Olkiluoto, V.C. Summer, Vogtle, etc). These failures are hard-engineering failures, nothing to do with politics.
Olkiluoto and Flamanville are all the consequences of mismanagement coupled with politics. They re-designed Olkiluoto whilst it was being built to adapt it to changing regulations, it is obviously a recipe for disaster. Most of Flamanville delays are due to poor quality control and issues with contractors and suppliers, i.e., mismanagement.
The problem with nuclear is that it requires a high initial capital investment, this is not a secret. This is also why private companies are not suited for this: they don’t price national interest and reliability of the energy supply at the scale of a continent. Nuclear programmes that do account for this (like France in the 1970s/1980s and China today) are massively successful.
Look at it this way: no more fossil fuel means that you can divert a sliver of the money that’s being spent on the military to secure oil supply to large infrastructure projects. You could build nuclear power plants with the yearly increase of the US military budget. It is only a matter of strategy.
> Btw, it isn't just nuclear that loses in this regard; bio-fuels, domestic-solar PV, green hydrogen, etc. all fail to compete in the new world of ultra-cheap solar PV and wind.
Well, good riddance to bio-fuels (which cause soil degradation, deforestation and dangerous monocultures) and green hydrogen (which is just greenwashed fossil fuel at the moment and does not make much thermodynamical sense in most scenarios).
Super-cheap wind turbines won’t be super-cheap anymore if they are the only tool we have.
I think this fact is under-appreciated. It has undermined most of the quantitative analysis done in the past. So most of the analysis available is stale.
It has also undermined a lot of business plans - even those based on renewables. For example the Xlink project in the UK - this involved building a huge solar PV plant (with complimentary wind and battery storage) in Morocco where panels are twice as efficient and an underwater HVDC cable to the UK. The problem is that panel prices have dropped so much, that it's now cheaper, easier and quicker to just buy twice as much PV panels and site them in the UK - compensating for the relative lack of efficiency in northern Europe by volume of panels.
This is why there's huge growth in utility solar PV in northern and western Europe in the last year or two despite the less-than-ideal conditions there - the panels are just so cheap that the lack of optimal efficiency becomes a non-issue.
My prediction is that wind, solar and batteries will dominate because they're riding on the benefits of mass-production - the capital cost for these is dominated by off-the-shelf component costs - installation is relatively trivial. This has delivered continuous and steep price declines while the prices for other forms of generation are stagnant (coal) or rising (nuclear). And there's no sign that the fall in prices is stopping at the moment so their price/value advantage will only grow in future.
I don't understand what you mean. Just because it's cheap, how is it suddenly not a consideration anymore whether we want to fill a large share of nature or farmland with PV, batteries, wind? There are more considerations than cost.
Also because one needs to keep in mind that renewable electricity is only about 10% of the energy demand. I sometimes see headlines like "61% of electricity in Germany came from wind last month!" which kind of miss the point, as that means we have only 94% to go. We really need a huge amount of surface area if we want to go for a monoculture of wind and solar power with battery storage.
Edit to be clear: I'm not against wind or solar or batteries. We will most definitely need it and must invest in both building them and innovating them further (in that order of priorities). But I also think we cannot just rely on them solving the whole problem. (Sometimes it is also assumed that this plummeting price curve will continue at the same rate for at least another decade, which I suspect is also not going to materialize.)
Getting rid of parking lots (and the energy- and ressources-hungry individual cars sitting on them) would certainly be a big step forward to a sustainable future.
I don't think they were suggesting getting rid of the the car parks, but rather putting solar over the top of them. Which would also reduce the amount of petrol used to AC the car when you get back in it too (though this effect is probably negligible)
I don’t think they did, either. Nevertheless, cars are a huge waste of ressources and energy and there is no realistic path to net zero in the next half century without a significant reduction in consumption. Electric cars or not.
Saying there is no realistic path to net zero without a reduction in consumption is somewhat funny to me. I don't think it is realistic at all to think that people will accept a reduction in living standards, which is what a reduction in consumption would result in.
I think switching consumption to forms that don't cause climate change is more realistic. Failing that, mass carbon capture is still more realistic than getting people to consume less, as far as I can tell.
Has anyone, anywhere ever had political success saying that everyone's living standards must decrease?
I don't think you're completely wrong, but it might be more nuanced than that. A decrease in consumption can come from less consumption, but also from more efficient consumption. Better second-hand markets for products that are fine to use a second time, longer-lasting products, less raw material needed perhaps, etc. But yeah I do expect the gains of that to be somewhat limited. Just look at how many people (especially in the USA, but also some Asian ones iirc) have an Apple device as a status symbol. They're probably going to put up with second-hand products as much as they are with Huawei phones. But that's a minority in most countries, i.e. far from everyone.
That depends on the form of consumption. For example, one large form of consumption is commuting from suburbs to the offices on gridlocked roads in SUVs with one person in it.
Everyone universally hates that, and working from home for office workers would greatly reduce consumption (of oil) while increasing living standards.
> Saying there is no realistic path to net zero without a reduction in consumption is somewhat funny to me. I don't think it is realistic at all to think that people will accept a reduction in living standards, which is what a reduction in consumption would result in.
I think you’re right. But I think Physics is going to show us the bill at some point and we won’t have a choice. When blackouts will be the norm, people will probably be very angry, but there won’t be much to do.
It is interesting to think about the perspective of e.g. a farmer during the fall of the western Roman Empire. I am certain that none of them wanted it, and it was actually quite violent at times, but there wasn’t anything they could do to stop it.
> Has anyone, anywhere ever had political success saying that everyone's living standards must decrease?
None (that I know of). It is already happening, though. Energy production in Europe is plateauing. Standards of living are already stagnating. Just like politicians cannot will away systemic technical problems in the economy or industry, they cannot change the laws of physics.
Our best bet is to use everything we can. Nuclear base production, as much wind and solar as we can, batteries and hydro to smooth out the peaks. We need to go all out. Even then, it would probably take a global industrial effort of the magnitude of the war efforts during WWI and WWII, just to get everything up and running. Let’s not kid ourselves: the world is burning, a mass extinction is under way, and there is already no way we meet our own bar of 1.5°C, and even 2°C requires immediate action.
I don't see why there have to be blackouts and that they would require reduction in consumption.
100,000 TW of sunlight hits the Earth's surface. Global primary energy demand is less than 20 TW. There is no global shortage of energy. If there are local issues, well, that's why we ship energy from one place to another, as is already done on a massive scale.
That's fair; I'm mainly looking at this from a western European perspective. If you're from a more southernly country with large unused deserts... that tends to help.
The very reason I'd like to see it updated is to see whether the conclusions change materially when we take into account technological progress.
As far as I understand it, whilst costs may be substantially lower that he envisaged, MacKay argued there are fundamental physical constraints on solar and wind that mean they still are unlikely to be able to provide all our energy needs (in the UK at least, without piping in energy from e.g. the Moroccan desert a la Xlinks).
If I recall correctly, he reached that conclusion by assuming a lot of biomass energy. Biomass is extremely inefficient at converting sunlight to usable energy (and burning biomass derived fuels in IC engines is very inefficient at converting that chemical energy to work.) PV + battery vehicles would drastically reduce the land area needed.
(I was David's Editor for "Sustainable Energy - without the hot air". There's been quite a bit of interest in updating the book, but there are obvious difficulties. If have suggestions, let me know.
Niall Mansfield
sewtha-2.0@uit.co.uk
)
I wonder if this is correct. Setting the nuclear slider to max single-handedly reduces emissions by 30%, decreases energy costs rather than increases, does not require land use change, improves air quality, and also reduces electricity demand by using the waste heat for district heating.
Now add: (1.) electrification of freight transport, (2.) tackling the (relatively) quick wins in industry electrification, and (3.) carbon capture, and one gets to 75% emissions reduction with less than one percent of cost increase and zero behavioural changes (such as eating less meat or flying less often) are required.
I can't find any other combination in this calculator that leaves nature so untouched, requires so little land use, requires no behaviour adjustment, or is so cheap while still reducing emissions meaningfully. (Well, geothermal is nearly as good but then its maximum potential is only a small percentage of energy demand and it's also a bit more expensive per Joule.)
Probably the most unrealistic part is public support for the choice, but in a perfect world? Would it be this simple?!
Agreed. I didn't use it at first because I thought it would be cheaper to just reduce emissions than undo emissions, at least up until this 80% goal that the site poses.
Without any carbon capture, you're still at 61% emissions reduction and the cost actually decreases by 0.8% from today rather than increasing by 0.6%. If you get to 61% with the only change being "build nuclear and make use of it" (no lifestyle changes, no cost changes, no landuse changes), that would be amazing.
Hence my main question is about the nuclear aspect. The argument against nuclear is usually that it's super scary and dangerous (easy enough to disprove that with numbers, so long as you're not talking to someone from germany) and the fallback argument is that it's so expensive now that PV+wind became so much cheaper. This calculator seems to show the opposite of that latter argument (when looking purely at price).
Surely there are studies of LCOE for both when trying to produce baseload capacity. Personally, I suspect nuclear construction costs are dominant, but that should also be addressible through improvements in technology and policy.
This is strange considering nuclear continues to be the most expensive new form of fuel and has only been getting more expensive with every passing decade.
There is new research in nuclear which if successful would bring costs down but it’s hard to see how maxing out on nuclear reduces average costs.
Edit: I see now. The original data is at least over 5 years old, at which point it’s possible that wind/solar may still have been more expensive than nuclear.
Editx2: The original book on which the original site is based was published in 2008/2009.
It’s not clear how much of the data on this new site has been updated.
2017 known supply is enough for 130 years at current usage (supplying about 10% of global energy). Each doubling of production halves the peak timeline. Basically no one expects discovery to outpace usage.
So, at best, nuclear is a stop gap and part of a different long term solution. Which could be fine! But the extremely low LCOE and fast build time for modern renewables suggests that nuclear is simply losing the race for relevance.
Fuel is very small part in cost of nuclear generation; As fuel price start increasing, more sources become viable, including sea water. There is practically unlimited supply once it happens, and this is before accounting for reprocessing. (and reprocessing happens in France, for example)
Right, but the fuel cost goes up over time as extraction gets harder (eventually requiring entire new types of facilities, which themselves take a decade to build), even as wind and solar continue their exponentially decreasing LCOE cost curve.
The exponentially decreasing cost curve is the part that I think people haven't actually internalized. The arguments on the pro-nuclear side are pretty much the same as they were ten years ago. (right down to 'we'll have thorium in ten years!') But things are very different now: It's like arguing for large scale investment in punch-card sorting machines circe 1960. Yes, the punch-card sorting machines can be used to manage all your bank data, but the digital processing are getting exponentially better with time and have been actually better for a couple years. But the proponents of punch card sorting machines haven't internalized the message yet.
Your meme pertains to exponential /growth/, not exponential /decline/. Exponential growth in a population hits a limit as a population saturates, causing the sigmoid shape. Exponential decline hits a constant asymptote and bottoms out. For populations, that asymptote tends to be zero. For cost curves, as in the solar panel case, it will be something more like capital cost of production divided by panel lifetime, which can be a very small number indeed.
> Right, but the fuel cost goes up over time as extraction gets harder
Uranium demand is declining, there has been almost no Uranium prosection in the last decades; Therefore it's no wonder reserves are diminishing.
If demand picks up, prospecting will renew, and a lot more uranium will be discovered.
This is not accounting for thorium reserves, which have never been seriously prospected.
Current reactors (light and heavy water reactors) all use solid fuel, which is very inefficient due to use of once-through fuel cycle.
Because xenon is the most common fission product, and it's a gas, the reaction must be stopped before all fuel is used, as otherwise the solid fuel pellets would crack. If you assume 3% enriched uranium (U-235) and 97% depleted uranium (U-238) (a typical fuel load), the fuel pellets are unusable after about 1/3 of enriched uranium is spent. So the fuel is removed and "thrown away" while it still has 2% U-235 in it.
Switching to liquid fuel reactors assuming no other changes would immediately increase supply 3 times (because xenon simply bubbles out). Switching to liquid fuel breeder reactors ~ 100 times. Switching to breeders based on thorium - around 1000 to 10000 times.
Peak uranium is only a problem because we're using a very inefficient nuclear reactor technology - water reactors. Why? Because they were chosen by the military and the technology/know-how was already there. Efficiency was really low on the military's list of priorities.
There's more to it than that. The long-term plan for nuclear fission always has been and remains to use breeder reactors after we have ~1000 GW-scale non-breeder reactors in place. Because they can utilize the majority isotope of uranium (U-238) and Thorium rather than just the small natural amount of fissile U-235, breeders enable us to make 100% of the world's primary energy for about 4 billion years (e.g. roughly until the sun burns out) [1].
Since this is roughly the remaining lifetime of the Sun, there's a very strong argument to consider nuclear fission breeder reactors just as renewable as the solar-derived energies (wind, solar, hydro, biofuel)
Which means that nuclear proponents should use the cost of breeder reactors with reprocessing, not burner reactors with a once-through fuel cycle, when calculating the cost of nuclear energy. This will likely increase the cost of energy from nuclear.
If we forced them to be honest we'd also force them to account for the huge carbon sink the plant represents until it has generated enough power to offset its construction footprint.
We need solutions that pay back as fast as possible, not stuff that will go neutral in 20 years.
Current nuclear supplies 10% of global ELECTRICAL energy, not total energy. If you scale burner reactors to supply the 18 TW of primary energy demand the world uses, projected cheap uranium runs out rather quickly.
Thanks Niall - updating the book is exactly what we are hoping to do and still thinking the best form for that. Re that tool i've been in correspondence with one of the other creators re trying to understand the source code etc - see https://github.com/life-itself/climate/issues/2
Niall - great to see you here. My mind goes towards helping others with "powers of 10 math" about climate. SWITHA for me embodied this concept of "how to think about the climate" more than anything.
That type of thinking is needed more now than ever. Even that line "2 billion years of energy reserves", so powerful.
Sir David Mackay [0] was a great mind who contributed major discoveries in Machine Learning, Information Theory, Coding Theory and Physics. His book "Information Theory, Inference, and Learning Algorithms" is a great resource for anyone learning ML and who wants to understand how ML, Information Theory and Statistics all fit together. The book is available for free download as a pdf [1] as well as for purchase.
His untimely death from cancer in 2016 at the age of 48 is a big loss to Science. His book on Climate Change "Sustainable Energy without the Hot Air" (also available as pdf) [2] is a testament to breadth and versatility of his scientific interests.
One thing that feels a bit strange to me is that he uses timelines of 1000 years when calculating nuclear reserves. I don’t believe we need a plan for the next 1000 years - think of what our understanding of power generation and transmission was a millennium ago. Assuming that we will still be using the same technologies in 1000 years seems a bit pessimistic to me.
On that timescale I expect us to engineer antimatter-based energy storage. Or do something literally incomprehensible — a thousand years ago we didn’t have the language to describe the language to describe relativity.
Antimatter might conceivably be viable as a form of energy storage. But as with hydrogen or other synfuels, it would never be a primary energy source, and the losses in production of antimatter are likely to remain large.
Ultimately, antimatter is a battery or fuel, not an energy source. And if it ever were a viable energy source ... well, we'd likely have bigger problems.
It doesn't quite work like that. 1000 years ago Europeans (you may say) only traveled routinely within 1 continent. Today, they travel routinely to all 7 continents. You can't say in another 1000 years anyone will travel to 14 continents -- there's only so much Earth to discover :)
We can know Earth in better detail, but it's inherently limited. In the same vein, there's no guarantee we will continue to discover revolutionary engineering strategies, and certainly no guarantee of discovering (or not discovering) new physics. In fact, physics has seen a slower pace of discoveries since mid-20th century. This kind of (not necessarily justifiable) optimism is in fact quite a risk when evaluating some technologies on sustainability.
I get the point, I just don’t think we can justify that level of confidence.
We might be just about to reach a grand unified theory of everything and then science just becomes collecting butterflies or whatever that quote was; or it might turn out that uncovering dark energy reveals 10^80 new kinds of fundamental particles and a way to construct pocket universes we can engineer to extract Big Bang energy levels or dump equivalent levels of entropy.
I wouldn’t want to put money on which; and I say that as someone hopeful for life extension tech that would allow me to wait and find out.
The thing is, we can't avoid putting money somewhere -- we are forced to make bets all the time (do nothing is also a bet). I would put the likelihood of pocket universes at something like 0.00000001% chance. Not zero, but really low.
Sure, but I wouldn’t make a 1000 year bet, nor accept that not making one in this particular case is itself a bet as the infrastructure we make today isn’t likely to still be around in even a third of that time.
All the stuff that I am currently in favour of for global energy supply and environment are mainly focused on the 50-100 year horizon for capability. I only care about longer timescales than that for damage on the grounds that kicking cans down the road seems both unfair and, given both how governments and civilisations fall, likely to make rebuilding unnecessarily difficult.
(I believe, but do not know, that this will change if we get radical life extension)
Nearly 100% effective power generation from matter via hawking radiation from a nano-sized black hole that was grown from one formed in the upper atmosphere via ultra-energetic cosmic rays.
> One thing that feels a bit strange to me is that he uses timelines of 1000 years when calculating nuclear reserves.
Not really. Uranium and thorium are insanely energy dense and whilst it’s obviously impossible to predict energy use a millenium from now, it could handle more than a couple of centuries of exponential growth of energy consumption. Then there are things like fuels made of natural isotopes (so no enrichment needed) and breeder reactors that can run off fission products (aka “waste”), which are not used much today, but are closer to readiness than things like carbon capture.
> it could handle more than a couple of centuries of exponential growth of energy consumption.
That rather depends on the exponent and the time period. The minimum for “a couple of centuries” is 200 years: 1%/year for 200 years? Sure; 1%/year for 1000 years or 5%/year for 200 years? 2-2.5 times global solar irradiance. I think that’s enough to raise the temperature to boiling, and I don’t think there’s enough nuclear fuel to do that (but I could be wrong about both).
It‘s such a tragedy that he died so young. Prof Mackay was a role model who impacted our world in many positive ways (e.g. his work in information theory, Bayesian statistics, environmental science). I recommend his lecture on information theory which is available on youtube [1]. He also documented the progress of his cancer very detailed on his blog [2], which is a very interesting (and sad) read.
It sounds like this should have a date attached to it as the numbers are quite outdated for todays world. This did have some good data for back when it was written of course.
* An efficient EV will do less than 12kWh/100km these days.
* Utility scale solar pv cost declines have been dramatic. It is now the cheapest form of new energy to deploy.
* Developments in deep water off-shore wind have led to much larger turbines which are capturing more wind energy for longer periods of time.
* Large scale batteries are now viable as a storage mechanism for renewables and are going to rapidly replace things like peaker plants.
* Lots of new research is happening with Nuclear like thorium/molten-salt reactors which will still be important for baseload generation.
> Utility scale solar PV [...] is now the cheapest form of new energy to deploy.
I've been hearing this claim for a few years now, and I took it at face value, until a few days ago. I was having a discussion with a friend and told him that PV's share of electricity generation must by quite high, and then I checked the EIA website [1], and it turns out only 2.3% of the electricity comes from PV, while about 40% come from gas and 20% from coal.
Something does not add up. If PV is so cheap, why don't we see more electricity being produced by PV?
Only recently did PV becomes the least-cost source of new electric generation.[0] And new utility-scale projects take 2-7 years to develop. For example, TX is anticipated to add 10 GW of new solar generation in 2022.[1] As I write this (while the sun is up, noon in TX), the current demand in TX is 38 GW.[2]
Source: Have been developing solar power plants since 2008.
[0] The tipping point for solar is different in different regions based on many factors, but primarily the amount of annual sunlight and the cost of fuel on the margin, and regulatory policy.
Hijacking your comment to say that the fine print in the EIA chart cited by the person you're replying to says they're only accounting for new and utility scale generation. Utility installs dwarf residential, but if you add all non-utility generation it's quite a bit more substantial, and it was 'only' until early to mid 2010's that utility installations outpaced other installs.
Utility installs dwarf residential in most places, but a notable exception is Australia. This probably reflects dysfunctional regulation and regulatory capture by fossil fuel interests there.
It takes many decades for old generation to phase out, so even with a significant fraction of new capacity being solar, the total proportion will remain low.
Solar capacity is increasing dramatically, but it does have limits.
Namely, solar + storage bids are only competitive in some markets. PV generation is often the cheapest source, but without storage it is not practical for many applications.
Yep, wind was the cheapest form of energy starting around mid 2010's.
I think wind's problem is that wind installations get massive pushback in a lot of communities. I wish I were joking when I said a turbine near me was permanently shut down after a federal judge declared the plaintiff's objections for health reasons to be valid.
The court fight triggered a poison pill town bylaw that a turbine inactive for X amount of time has to be removed. All the idiots had to do was get an injunction and then tie the whole thing up in court for long enough.
1) lifetimes of utility assets are measured in decades, not years
2) the utility industry is not used to needing to pay attention to new technology and new information, and has a huge bias against renewables that goes back to the hard energy /soft energy split of the 1970s and 1980s
3) Utility planning models (IRPs, often for five years out) often use outdated info that is 3-5 years old when planning deployments for the next five years. So even if utilities used up to date info, it would take 5 years for them to shift strategy.
4) many utilities are not incentivized to install least cost generation, and their incentives and profit rates are different for things like installing transmission, distribution, etc
5) regulators of utilities are often even further behind the times than utility executives or completely captured (see for example Arizona)
> 1) lifetimes of utility assets are measured in decades, not years
Fair enough, but here's [1] how the picture looked just 4 years ago. Coal was at 33%, natgas at 33% an solar at 0.6%, wind at 4.7%. So a lot of coal was replaced by natgas, and a bit of wind. For some reason solar went up only by 1.7%
The reasons further down explain why utilities are very slow to react and why even after a better choice is available.
But even with your further comparison, you are falling prey to the same bad logic: solar can't be cheaper because we aren't already using it. You are looking at the decisions from 2010 to 2015 to evaluate the situation from 2021.
For changing technology, that's a really bad assumption. If nearly all the market for CPUs is Intel, then AMD comes out with a far better deal than Intel, do you evaluate how much better AMD is by the installed base across all computer? Of course not.
Check out what is being planned for the future. Some of the decisions are not great, because utilities don't yet know that storage is super cheap. But you'll find that in price responsive markets, nearly all new planned generation is wind, solar, and storage. Along with a few new gas plants proposed by people hedging against the dominant tech, or just through inertia.
Cheapest to deploy, meaning adding new energy sources.
For instance, even since the year or two that EIA has last updated their numbers to the numbers that just came out that include August 2021... the amount of total solar (far right column) generated (154TWh) vs total electricity generated (4087TWh) gives about 3.8% solar electricity for the last 12 months rolling.
https://www.eia.gov/electricity/monthly/epm_table_grapher.ph...
Because it has only recently become cheaper to invest in solar rather than "ride out" existing coal investments.
I believe that the natural gas generators are still not at the "ditch as soon as you can" moment that hit coal hit a few years ago, and even then there's only so much solar being made right now.
The other factor is that solar manufacturing isn't being full-bore developed because as soon as you get a plant up, you're at risk of a paradigm shift beating you out. The promise of perovskites might be holding back silicon panel investment.
The Chinese, for as much as US media has been portraying them as a big bad, did a massive solid for the world in pushing solar panel manufacturing and battery manufacturing from the top down.
Granted they wanted to push everyone out/monopolize manufacturing, but who cares WHY, the net effect is to change our possible future for decarbonization. The US would never do that with the dominance of petroleum lobbying.
South Korea should also get a fair amount of credit for supporting Samsung in this direction.
It didn't seem to mention hybrids much either. It mentioned a polo bluemotion as the lowest co2 emitting vehicle at 99g/km, but even my 10+ year old hybrid clunker (Toyota auris) officially gets 82g/km - modern hybrids must be way better by now, and they are super common now rather than just Toyota's selling point.
A recent report I read on the BBC suggests that public bus transportation is 90g/km/passenger so my ancient old hybrid is better than taking the bus it seems?
Most of the buses I see around here (Sunnyvale through San Mateo is my typical stomping ground) are either "not in service" or very nearly empty. I haven't taken a bay area bus in fifteen years. I routinely outpace them on a bicycle.
I could imagine 8-15 passenger vans with on-demand routing would see much better utilization.
Yup. They should be easier to electrify, too. Mass produced electric passenger vans should be a thing. Considering transit vans are only like $50,000 apiece for 12 seats, they should be pretty dang cheap to acquire, too, compared to the $1 million 30 seat e-buses. Charging infrastructure should be the same as regular electric cars, too, which makes it cheaper and more flexible to deploy.
Except if you're not driving the bus at full capacity (i.e. vast majority of the time), you're not saving any labor costs by using the larger vehicle. In fact, the greater maintenance requirements probably cause greater labor costs.
Additionally, transit vans don't necessarily need a commercial license to drive, which can reduce labor costs (in part by expanding the availability of the labor pool).
Your typical location might place you on the route bus drivers use to return to a depot or to get to the start of a popular route. Or the time you're out and about might coincide with when a large number of busses are coming into, or going out of, service. For example, if you work offset hours you'd easily see the pre or post rush hour ramp up/down traffic.
Out of service busses also travel faster than in service ones; as you note, you outpace busses on your bike.
Perhaps look up research instead of relying upon personal anecdotal observations by one person in one city, subject to observational biases?
The research is likely to be highly accurate given most public bus operators have very good data about ridership numbers and fuel use per vehicle and route.
Those mitigating circumstances are good things to think about, but aren't applicable to my experience. If you've got hard data, please post it. I was unable to find utilization rates. Total ridership is easy to find, but not answering the questions at hand (utilization and uptime).
ReaRead that, and then quit wasting everyone's time advocating a stupid idea that keeps coming bcback. Good transit advocacy should aim to get people to not drive.
The article talks a lot about coverage but not utilization. Searching the domain (I'm not going to read all of its content to narrowly search for something) doesn't surface much data or discussion of bus utilization.
Is that "official" figure from Toyota, or from an independent testing group? A huge number of manufacturers have been caught or admitted to, cheating on their emissions figures - including CO2 levels. Claimed figures are basically worthless.
Why would you think that a compact hybrid would have only slightly better CO2 performance compared to a city bus per passenger? It doesn't even make sense for the vehicle as a whole. Googling around, I came up with this:
...shows that your figure is 90g per km for the bus as a whole, not per km/passenger.
Every chart I've seen on co2 or fuel use per mile per person, busses, bicycles, and rail are king. Ride shares are the absolute worst because of all the wasted miles....followed by private SUV and car use.
If you do a google image search for "co2 emissions per passenger mile" you can see endless charts from different studies.
So anywhere between 800% to 1500% worse than a decade-old hybrid.
Of course, if the bus has 24 to 45 passengers on it, then it is better on a per-passenger basis than me in my car with my wife and kid. I couldn't find any rider-ship data from TfL, but they often appear kinda empty outside of rush hours.
> Large scale batteries are now viable as a storage mechanism for renewables
They are economical viable for solar for 3-4 hours every night, which is where they are being deployed. The article talks primarily about the UK with wind being the primary energy source, and so rather than a solar cycle of 24hrs you have wind fluctuations with lulls. The article calculate this to about 1200 GWh.
The articles I have seen on solar + batteries is a capacity of 80% for 4 hrs. I have not seen clear numbers but I would guess that they do get a fairly high discharge rate each day, thus providing an return of investment when the price is at its highest point each day. For wind you would need to have as much capacity as the worst lulls, and they would only really return profits during the lulls (if we do not count subsidizes). There would be some profits from slews, but most of the cost would be going to capacity that isn't needed for most part of the year.
Before research into sustainable energy, David MacKay had rediscovered Low Density Parity Check (LDPC) codes in early 90s, perhaps the most common type of error correction codes used nowadays in commercial systems (alongside with Reed Solomon codes used for storage).
He then worked on sustainable energy long before it became fashionable.
His early death was a significant loss. He blogged cheerfully from hospital til few days before his death
I highly recommend Prof. Tom Murphy's new book "Energy and Human Ambitions" (freely available as well) that is sort of a deeper dive into some of the questions that Mackay explored:
I came here to write about Tom's book. I consider it Mackay's successor and called it The Science Book of the Decade in my review of it. He wrote it for non-science undergraduates.
The original of this book is a wonderful resource. I am very happy to see a community bringing it up-to-date because MacKay had the foresight to license the work using CC.
There was indeed a lot of hot air in sustainability thoughts in 2008, and probably still is, but Mackay's treatment is sometimes bordering the absurd.
For instance, the chapter on wind power starts by comparing a back of the envelope calculation on wind power with energy usage by a fossil car.
While there are occasional good points, I would not recommend reading this book unless you actually know something about the things the book is discussing. It's simply too misleading in important places.
In case anyone is wondering about this, there are real studies on the feasibility of 100% renewables by organizations that actually know this stuff and have done the modelling work.
Mackay's point is that you could get through the confusion relatively quickly, and showing in principle how to evaluate and fight the problem. I don't think it was meant as a final solution -- just to show that it was feasible, and approachable as a problem. Without getting into politics or ideology.
I got a quite good intuitive sense of the scales involved, and at least for me Solar clearly reveals itself as an outstanding source of energy (at say 150W/m2).
Didn't this book also assume substantial biomass energy grown in the UK? If I recall correctly, it was negative on solar due to land use, but that was mostly due to this biomass (which has extremely low energy capture efficiency, requiring a lot of land.)
137 comments
[ 3.1 ms ] story [ 175 ms ] threadSome technologies (PV, batteries, wind) have improved by an order of magnitude - and swamp out all other considerations.
EDIT: to be clear, i loved the original book, and have re-read it several times over the years.
I think that updating this book is not pointless, but more importantly doing the same for more countries is really important.
For example, the Stanford/Jacobson study on powering the world with wind/solar/water.
Some results were famously silly, such as filing Finland with solar panels, or adding the cost of nuclear war to nuclear reactors. Jacobson sued other academics then backed down.
The point is, 5-10 years on, these models turned out to be over pessimistic! Australia doesn't need fancy modelling - simple roof top solar there is on target to meet full day time demand in 2-3 years.
Conversely we still dont really know how to meet the last 5% of outlier energy demand scenarios.
Is that really silly? The incremental increase in the risk of nuclear war for each gigawatt of civilian nuclear power generation might be small, but the costs of a nuclear war could be quadrillions of dollars (for example if it set the world back by 100 years).
Obviously measuring that risk is difficult, but to give an example, imagine that civilian nuclear power were limited to just the countries which currently have nuclear weapons. That would greatly reduce the proliferation risk and the chance of nuclear breakout in countries like Iran.
The urgency of addressing climate change may take precedence over these concerns in the short term, but if we do manage to reach net zero with a mixture of renewables and nuclear energy, then it's not inconceivable that over-provisioning renewables might allow us to phase out nuclear energy later on.
Suppose it would take 50 years to build up that renewable energy capacity, and 10 years to build a new nuclear power station. That would mean the nuclear power station would only be useful for 40 years, which limits the amount of time they'd have to recoup their costs over.
Hydrogen w. combustion turbines. When a simple cycle turbine power plant is 5% of the cost of a nuclear plant of the same output, we can back up the entire grid and not have it be too expensive.
Massive +1. And i agree that perhaps "reusing" the book would be useful. E.g. take the key factual approach and update analyses for particular technologies and then plug those together - which to some extent is what the pathway calculator did which is why it is worth trying to get the source for that https://github.com/life-itself/climate/issues/2
While MacKay's work is excellent for a general lay audience to understand general scales of the issue, there are now far more advanced models for this that do things like replay historical weather and demand, down to 15 minute scales, to more least cost systems for zero carbon grids.
They are also so advanced as to take into current grid structure and generation resources, and the cheapest way to transition to zero carbon from our current resources.
Christopher Clack's models are probably the most advanced. The latest iteration has the very surprising finding that deploying lots of grid-edge solar and storage right now will save us a ton of money because it will use current transmission and distribution assets more effectively and reduce the need for future investment in these assets (T&D is the majority of our electricity bills, not generation!)
https://journals.ametsoc.org/view/journals/apme/56/1/jamc-d-...
But it isn’t particularly helpful when trying to explain policy imho.
> Wind turbines are getting bigger all the time. Do bigger wind turbines change this chapter’s answer?
> Chapter B explains. Bigger wind turbines deliver financial economies of scale, but they don’t greatly increase the total power per unit land area, because bigger windmills have to be spaced further apart. A wind farm that’s twice as tall will deliver roughly 30% more power.
So, did wind turbine production actually increased by an order of magnitudes since the book ?
Turbines have gone from 1-2MW to 10-16MW
Capacity factors have gone from 20% to 50%-60%
Annual additions gave gone from sub-GW to 70GW last year
And with turbine spacing of 1km you can actually do stuff with the land in between.
On the other hand, transmission projects still take 10-20 years, and new transmission projects are often over booked by 300%-400% before they are even started.
Electric cars are ridiculously efficient. A 100kwh tesla drives approx 500km. So 50km/day/car needs just 10kwh. The UK also has 2 persons per car (plus additional cats and dogs..)
So just the on-shore book projection is already enough AND only requires a tenth of the turbines.
Lets keep nuclear as well. Invest in efficiency. Lots of hard work ahead.
But we are probably closer to "problem solved" than to "civilization collapse". Cheers.
For what it's worth "In 2019, the average car in the UK drove 7,400 miles"[0] which works out to 33 km/day/car. It's probably much less now due to working from home (and fuel shortages, and the collapse of the supply chain...)
[0] https://www.nimblefins.co.uk/cheap-car-insurance/average-car...
Okay, but GP's point was that these bigger turbines need to be spaced further apart instead, so we can place fewer of them. That we could build bigger turbines was not in question I think. Do you happen to know whether the MW per km² increased more than the 30% expected amount that GP cited?
And regarding doing things between the turbines, I wasn't "awake" yet in 2006 so I don't remember how it was then, but were they ever placed so close together that you couldn't have farmland in between turbines? The problem seems to be that people don't want to live near them and find nature filled with wind turbines ugly, not so much that you can't do anything else in between.
(Just to note, I don't find them that ugly (even if, of course, nature would be prettier without them... but that's not an option) and I also didn't find them to be very loud when I visited some nearby wind turbines. Perhaps it's different at night when it's all quiet, but personally I don't think I'd mind living next to one. The blades are a bit scary though, I can't help but imagine the consequences if one of them lets loose... but that's like being afraid of air travel I guess.)
What will be a problem if our energy consumption increases by an order of magnitude or two, and we take all of it from the wind. Currently, the entire question is about financial ROI.
> but were they ever placed so close together that you couldn't have farmland in between turbines?
Not since the modern turbines were invented. The first ones were placed in somewhat compact lines, with enough spacing between the turbine lines that you could raise a few lines of any crop. (The lines can't be too close anyway, as that would reduce the turbine efficiency and harm financial ROI.)
Network reliability has actually increased in many European countries despite the increasing reliance on intermittent sources.
The current approach certainly depends on having back-up deployable generation capacity (natural gas) but that was always required for demand following even with thermal sources so most grids already have the infrastructure to incorporate more intermittent generation.
But assuming no technology advances in this area for 30 years time seems implausible. What has happened with price/efficiency of solar, wind and grid-scale batteries in the last 10 years provides a lesson. None of these were practical even 10 years ago and now they represent 90% of newly installed capacity globally (2020). And this is happening all-over whether with or without government support.
For example, new-build, grid-scale battery storage is now cheaper on an LCOE basis than new build open-cycle ("peaker") natural gas plants. This has only happened recently - about a year ago. I was completely skeptical that this could happen a few years ago.
I urge any/all hackernews readers to read more about _recent_ developments in de-carbonizing energy. It's fascinating and it has become clear to me that we're living through one of the great technology revolutions of human history. It genuinely is a Kodak/smartphone type moment with even bigger implications for human wellbeing.
It's amazing that by being cheap enough, all the disadvantages of a new technology become less relevant. A bit like how the PC displaced "real computers" (i.e. mainframes) even though at the time they were little more than microcontrollers integrated with a terminal.
Volatile renewables don’t effectively help to reduce greenhouse gas emissions.
https://ourworldindata.org/grapher/ghg-emissions-by-sector?t...
As the renewables share of generation has been increasing, the CO2 intensity of electricity generation in Germany has dropped from 542gCO2/kWh in 2000 to 296gCO2/kWh in 2020[1].
Btw, your link brings me to a table on agricultural CO2 emissions?
[1] https://www.iea.org/data-and-statistics/charts/development-o...
True. It’s very easy when you can tap your neighbour’s power plants when you need. The problem is a bit different when all the neighbours have the same problems at the same time.
And there are cases like Ireland which has its own grid - with very limited external interconnection - and yet achieved 42% renewable share last year.
Techniques/engineering/theory for integrating intermittent sources has advanced considerably in the last ten years.
It’s more than a little help. Just imagine how a blackout involving 10% of the population or Germany would look like, never mind the EU.
> And there are cases like Ireland which has its own grid - with very limited external interconnection - and yet achieved 42% renewable share last year.
You cannot really compare Ireland with many of the other EU countries, either in term of population or industry, though.
> Techniques/engineering/theory for integrating intermittent sources has advanced considerably in the last ten years.
They have. And it is a good thing, because it really should not be a competition between nuclear and renewables. They still have several issues, which are very difficult to solve without either a massive reduction in consumption, or a massive increase in price.
I really struggle to understand how almost all mainstream “green” parties see fossil fuels as a lesser menace than nuclear. The truth is, if you read their manifesto, that they don’t care about particle pollution beyond “cars are bad” (which is true, but insufficient) or climate change.
It is not a competition, it is a struggle to get rid of fossil fuels, which are an existential threat to our societies. We can sort that out and talk about fusion once we’ve done it. In the meantime, getting rid of fossil fuels is a massive undertaking, and we would be stupid not to use every card in our hand.
We live in a world of finite resources. When considering the significant task of decarbonizing energy, there are lots of technologies potentially available. But some deliver more value than others.
Unfortunately nuclear is not _currently_ one of the options which delivers value in this regard.
Flamenville 3 - the newest French reactor will cost $22B to build 1.6GW of capacity and will have taken nearly 20 years from the start of construction to deliver a watt. The same amount of money could deliver over 20GW of wind capacity which could be deployed almost instantly as it's an easily parallelizable low-risk, relatively low-tech engineering task relying on mass-production for most of the installation. None of these benefits are available to nuclear reactor construction. Even with a capacity factor of 40% for on-shore wind vs 90%+ for nuclear, nuclear just isn't close.
I'm not against nuclear. If the nuclear industry can step up and deliver reactors on time and on budget which can provide power at competitive prices, then I'd be all in favour. It would be wonderful to have an alternative in the race to decarbonize energy. But all I see is a long trail of massively delayed or abandoned projects with incredible cost-overruns (Flamanville, Olkiluoto, V.C. Summer, Vogtle, etc). These failures are hard-engineering failures, nothing to do with politics.
With such a litany of recent failures, the incredible capital expense, the extra security required, the slowness of delivery, it's clear why resources have been directed away from nuclear toward other technologies which are quick and easy to roll-out and have already significantly contributed to reducing the the CO2 intensity of Europe's electricity (which is now less than half of what it was 30 years ago).
Btw, it isn't just nuclear that loses in this regard; bio-fuels, domestic-solar PV, green hydrogen, etc. all fail to compete in the new world of ultra-cheap solar PV and wind.
Nuclear is bottlenecked by the availability of properly trained engineers (and you really need those for such projects). The need in resources would not impact installations of either wind turbines or solar panels.
Nuclear+renewables is not more expensive than all renewables, particularly if you consider the fact that at the moment renewables benefit from an industry that is entirely dependent on subsidised fuel. And that nuclear reactors would be much cheaper per unit if we stopped building only a couple of each class. Also, a nuclear power plant is expensive, but so is the equivalent capacity in wind turbines plus the required storage.
A nuclear baseline mitigates the issues of wind turbines, and wind turbines are a good complement to nuclear. Both are mostly mutually exclusive for political reasons.
> Flamenville 3 - the newest French reactor will cost $22B to build 1.6GW of capacity and will have taken nearly 20 years from the start of construction to deliver a watt. The same amount of money could deliver over 20GW of wind capacity which could be deployed almost instantly as it's an easily parallelizable low-risk, relatively low-tech engineering task relying on mass-production for most of the installation. None of these benefits are available to nuclear reactor construction. Even with a capacity factor of 40% for on-shore wind vs 90%+ for nuclear, nuclear just isn't close.
The same reactor was built on time and on budget in China, and is currently operating. Hinckley point is also going more or less as planned. Areva/Orano’s abysmal performance is more related to their mismanagement than the reactor itself. You also need to factor stuff like storage, and the fact that we have a limited shore length to use.
> I see is a long trail of massively delayed or abandoned projects with incredible cost-overruns (Flamanville, Olkiluoto, V.C. Summer, Vogtle, etc). These failures are hard-engineering failures, nothing to do with politics.
Olkiluoto and Flamanville are all the consequences of mismanagement coupled with politics. They re-designed Olkiluoto whilst it was being built to adapt it to changing regulations, it is obviously a recipe for disaster. Most of Flamanville delays are due to poor quality control and issues with contractors and suppliers, i.e., mismanagement.
The problem with nuclear is that it requires a high initial capital investment, this is not a secret. This is also why private companies are not suited for this: they don’t price national interest and reliability of the energy supply at the scale of a continent. Nuclear programmes that do account for this (like France in the 1970s/1980s and China today) are massively successful.
Look at it this way: no more fossil fuel means that you can divert a sliver of the money that’s being spent on the military to secure oil supply to large infrastructure projects. You could build nuclear power plants with the yearly increase of the US military budget. It is only a matter of strategy.
> Btw, it isn't just nuclear that loses in this regard; bio-fuels, domestic-solar PV, green hydrogen, etc. all fail to compete in the new world of ultra-cheap solar PV and wind.
Well, good riddance to bio-fuels (which cause soil degradation, deforestation and dangerous monocultures) and green hydrogen (which is just greenwashed fossil fuel at the moment and does not make much thermodynamical sense in most scenarios).
Super-cheap wind turbines won’t be super-cheap anymore if they are the only tool we have.
It has also undermined a lot of business plans - even those based on renewables. For example the Xlink project in the UK - this involved building a huge solar PV plant (with complimentary wind and battery storage) in Morocco where panels are twice as efficient and an underwater HVDC cable to the UK. The problem is that panel prices have dropped so much, that it's now cheaper, easier and quicker to just buy twice as much PV panels and site them in the UK - compensating for the relative lack of efficiency in northern Europe by volume of panels.
This is why there's huge growth in utility solar PV in northern and western Europe in the last year or two despite the less-than-ideal conditions there - the panels are just so cheap that the lack of optimal efficiency becomes a non-issue.
My prediction is that wind, solar and batteries will dominate because they're riding on the benefits of mass-production - the capital cost for these is dominated by off-the-shelf component costs - installation is relatively trivial. This has delivered continuous and steep price declines while the prices for other forms of generation are stagnant (coal) or rising (nuclear). And there's no sign that the fall in prices is stopping at the moment so their price/value advantage will only grow in future.
Also because one needs to keep in mind that renewable electricity is only about 10% of the energy demand. I sometimes see headlines like "61% of electricity in Germany came from wind last month!" which kind of miss the point, as that means we have only 94% to go. We really need a huge amount of surface area if we want to go for a monoculture of wind and solar power with battery storage.
Edit to be clear: I'm not against wind or solar or batteries. We will most definitely need it and must invest in both building them and innovating them further (in that order of priorities). But I also think we cannot just rely on them solving the whole problem. (Sometimes it is also assumed that this plummeting price curve will continue at the same rate for at least another decade, which I suspect is also not going to materialize.)
The battery land requirements are approximately 1 square mile.
And wind requirements are effectively 0 since wind can very easily co-exist with other land usages.
I think switching consumption to forms that don't cause climate change is more realistic. Failing that, mass carbon capture is still more realistic than getting people to consume less, as far as I can tell.
Has anyone, anywhere ever had political success saying that everyone's living standards must decrease?
Everyone universally hates that, and working from home for office workers would greatly reduce consumption (of oil) while increasing living standards.
I think you’re right. But I think Physics is going to show us the bill at some point and we won’t have a choice. When blackouts will be the norm, people will probably be very angry, but there won’t be much to do.
It is interesting to think about the perspective of e.g. a farmer during the fall of the western Roman Empire. I am certain that none of them wanted it, and it was actually quite violent at times, but there wasn’t anything they could do to stop it.
> Has anyone, anywhere ever had political success saying that everyone's living standards must decrease?
None (that I know of). It is already happening, though. Energy production in Europe is plateauing. Standards of living are already stagnating. Just like politicians cannot will away systemic technical problems in the economy or industry, they cannot change the laws of physics.
Our best bet is to use everything we can. Nuclear base production, as much wind and solar as we can, batteries and hydro to smooth out the peaks. We need to go all out. Even then, it would probably take a global industrial effort of the magnitude of the war efforts during WWI and WWII, just to get everything up and running. Let’s not kid ourselves: the world is burning, a mass extinction is under way, and there is already no way we meet our own bar of 1.5°C, and even 2°C requires immediate action.
100,000 TW of sunlight hits the Earth's surface. Global primary energy demand is less than 20 TW. There is no global shortage of energy. If there are local issues, well, that's why we ship energy from one place to another, as is already done on a massive scale.
The very reason I'd like to see it updated is to see whether the conclusions change materially when we take into account technological progress.
As far as I understand it, whilst costs may be substantially lower that he envisaged, MacKay argued there are fundamental physical constraints on solar and wind that mean they still are unlikely to be able to provide all our energy needs (in the UK at least, without piping in energy from e.g. the Moroccan desert a la Xlinks).
(Yes, I realise solar can do an awful lot if you're willing to 'go big' and have huge batteries. https://www.robinlinacre.com/fill_country_solar/)
(I was David's Editor for "Sustainable Energy - without the hot air". There's been quite a bit of interest in updating the book, but there are obvious difficulties. If have suggestions, let me know. Niall Mansfield sewtha-2.0@uit.co.uk )
Now add: (1.) electrification of freight transport, (2.) tackling the (relatively) quick wins in industry electrification, and (3.) carbon capture, and one gets to 75% emissions reduction with less than one percent of cost increase and zero behavioural changes (such as eating less meat or flying less often) are required.
I can't find any other combination in this calculator that leaves nature so untouched, requires so little land use, requires no behaviour adjustment, or is so cheap while still reducing emissions meaningfully. (Well, geothermal is nearly as good but then its maximum potential is only a small percentage of energy demand and it's also a bit more expensive per Joule.)
Probably the most unrealistic part is public support for the choice, but in a perfect world? Would it be this simple?!
Carbon capture and less than one percent cost increase looks suspect to me
Without any carbon capture, you're still at 61% emissions reduction and the cost actually decreases by 0.8% from today rather than increasing by 0.6%. If you get to 61% with the only change being "build nuclear and make use of it" (no lifestyle changes, no cost changes, no landuse changes), that would be amazing.
Hence my main question is about the nuclear aspect. The argument against nuclear is usually that it's super scary and dangerous (easy enough to disprove that with numbers, so long as you're not talking to someone from germany) and the fallback argument is that it's so expensive now that PV+wind became so much cheaper. This calculator seems to show the opposite of that latter argument (when looking purely at price).
There is new research in nuclear which if successful would bring costs down but it’s hard to see how maxing out on nuclear reduces average costs.
Edit: I see now. The original data is at least over 5 years old, at which point it’s possible that wind/solar may still have been more expensive than nuclear.
Editx2: The original book on which the original site is based was published in 2008/2009.
It’s not clear how much of the data on this new site has been updated.
2017 known supply is enough for 130 years at current usage (supplying about 10% of global energy). Each doubling of production halves the peak timeline. Basically no one expects discovery to outpace usage.
So, at best, nuclear is a stop gap and part of a different long term solution. Which could be fine! But the extremely low LCOE and fast build time for modern renewables suggests that nuclear is simply losing the race for relevance.
The exponentially decreasing cost curve is the part that I think people haven't actually internalized. The arguments on the pro-nuclear side are pretty much the same as they were ten years ago. (right down to 'we'll have thorium in ten years!') But things are very different now: It's like arguing for large scale investment in punch-card sorting machines circe 1960. Yes, the punch-card sorting machines can be used to manage all your bank data, but the digital processing are getting exponentially better with time and have been actually better for a couple years. But the proponents of punch card sorting machines haven't internalized the message yet.
What you think is an exponential curve is actually the left side of a sigmoid.
Uranium demand is declining, there has been almost no Uranium prosection in the last decades; Therefore it's no wonder reserves are diminishing. If demand picks up, prospecting will renew, and a lot more uranium will be discovered. This is not accounting for thorium reserves, which have never been seriously prospected.
Because xenon is the most common fission product, and it's a gas, the reaction must be stopped before all fuel is used, as otherwise the solid fuel pellets would crack. If you assume 3% enriched uranium (U-235) and 97% depleted uranium (U-238) (a typical fuel load), the fuel pellets are unusable after about 1/3 of enriched uranium is spent. So the fuel is removed and "thrown away" while it still has 2% U-235 in it.
Switching to liquid fuel reactors assuming no other changes would immediately increase supply 3 times (because xenon simply bubbles out). Switching to liquid fuel breeder reactors ~ 100 times. Switching to breeders based on thorium - around 1000 to 10000 times.
Peak uranium is only a problem because we're using a very inefficient nuclear reactor technology - water reactors. Why? Because they were chosen by the military and the technology/know-how was already there. Efficiency was really low on the military's list of priorities.
[1] https://whatisnuclear.com/blog/2020-10-28-nuclear-energy-is-...
The same point was made in the Hot Air book as well [2]
[2] https://www.withouthotair.com/c24/page_162.shtml
Since this is roughly the remaining lifetime of the Sun, there's a very strong argument to consider nuclear fission breeder reactors just as renewable as the solar-derived energies (wind, solar, hydro, biofuel)
We need solutions that pay back as fast as possible, not stuff that will go neutral in 20 years.
Current nuclear supplies 10% of global ELECTRICAL energy, not total energy. If you scale burner reactors to supply the 18 TW of primary energy demand the world uses, projected cheap uranium runs out rather quickly.
Niall - great to see you here. My mind goes towards helping others with "powers of 10 math" about climate. SWITHA for me embodied this concept of "how to think about the climate" more than anything.
That type of thinking is needed more now than ever. Even that line "2 billion years of energy reserves", so powerful.
His untimely death from cancer in 2016 at the age of 48 is a big loss to Science. His book on Climate Change "Sustainable Energy without the Hot Air" (also available as pdf) [2] is a testament to breadth and versatility of his scientific interests.
[0] https://en.wikipedia.org/wiki/David_J._C._MacKay
[1] http://www.inference.org.uk/mackay/itila/book.html
[2] http://www.withouthotair.com/
EDIT: typos
http://www.withouthotair.com/download.html
The parent's link was for his other book "Information Theory, Inference, and Learning Algorithms", which is a happy thing to stumble upon.
Edit: Just realised that I misread the parent comment and the link to the Info Theory book was intentional.
Ultimately, antimatter is a battery or fuel, not an energy source. And if it ever were a viable energy source ... well, we'd likely have bigger problems.
We can know Earth in better detail, but it's inherently limited. In the same vein, there's no guarantee we will continue to discover revolutionary engineering strategies, and certainly no guarantee of discovering (or not discovering) new physics. In fact, physics has seen a slower pace of discoveries since mid-20th century. This kind of (not necessarily justifiable) optimism is in fact quite a risk when evaluating some technologies on sustainability.
We might be just about to reach a grand unified theory of everything and then science just becomes collecting butterflies or whatever that quote was; or it might turn out that uncovering dark energy reveals 10^80 new kinds of fundamental particles and a way to construct pocket universes we can engineer to extract Big Bang energy levels or dump equivalent levels of entropy.
I wouldn’t want to put money on which; and I say that as someone hopeful for life extension tech that would allow me to wait and find out.
All the stuff that I am currently in favour of for global energy supply and environment are mainly focused on the 50-100 year horizon for capability. I only care about longer timescales than that for damage on the grounds that kicking cans down the road seems both unfair and, given both how governments and civilisations fall, likely to make rebuilding unnecessarily difficult.
(I believe, but do not know, that this will change if we get radical life extension)
Not really. Uranium and thorium are insanely energy dense and whilst it’s obviously impossible to predict energy use a millenium from now, it could handle more than a couple of centuries of exponential growth of energy consumption. Then there are things like fuels made of natural isotopes (so no enrichment needed) and breeder reactors that can run off fission products (aka “waste”), which are not used much today, but are closer to readiness than things like carbon capture.
That rather depends on the exponent and the time period. The minimum for “a couple of centuries” is 200 years: 1%/year for 200 years? Sure; 1%/year for 1000 years or 5%/year for 200 years? 2-2.5 times global solar irradiance. I think that’s enough to raise the temperature to boiling, and I don’t think there’s enough nuclear fuel to do that (but I could be wrong about both).
[1] https://youtu.be/BCiZc0n6COY
[2] http://itila.blogspot.com/?m=1
* An efficient EV will do less than 12kWh/100km these days.
* Utility scale solar pv cost declines have been dramatic. It is now the cheapest form of new energy to deploy.
* Developments in deep water off-shore wind have led to much larger turbines which are capturing more wind energy for longer periods of time.
* Large scale batteries are now viable as a storage mechanism for renewables and are going to rapidly replace things like peaker plants.
* Lots of new research is happening with Nuclear like thorium/molten-salt reactors which will still be important for baseload generation.
I've been hearing this claim for a few years now, and I took it at face value, until a few days ago. I was having a discussion with a friend and told him that PV's share of electricity generation must by quite high, and then I checked the EIA website [1], and it turns out only 2.3% of the electricity comes from PV, while about 40% come from gas and 20% from coal.
Something does not add up. If PV is so cheap, why don't we see more electricity being produced by PV?
[1] https://www.eia.gov/tools/faqs/faq.php?id=427&t=3
> Utility scale solar PV [...] is now the cheapest form of new energy to deploy
For China... The West has very few semiconductor grade silicon smelters, and no industrial scale ones for PV ingots.
I think all of PV cell makers in the West just dice Chinese boules
https://ihsmarkit.com/research-analysis/over-80-of-2020-glob...
Source: Have been developing solar power plants since 2008.
[0] The tipping point for solar is different in different regions based on many factors, but primarily the amount of annual sunlight and the cost of fuel on the margin, and regulatory policy.
[1] https://www.reuters.com/business/environment/texas-track-add...
[2] http://www.ercot.com/
According to an industry group, 20GW of solar was installed last year (3.1GW residential), representing 43% of new generation capacity: https://www.seia.org/research-resources/solar-market-insight...
According to EIA, new wind generation in 2020 was a bit over 14GW.
Walking back from 20GW being 43%, that means in 2020 we had 46GW installed overall, so wind was 30%.
By any measure, 73% of new generation being renewable is pretty impressive. Worldwide the number of solar installs looks almost logarithmic.
https://cleantechnica.com/files/2020/09/Cumulative-Total-US-...
Solar capacity is increasing dramatically, but it does have limits.
Namely, solar + storage bids are only competitive in some markets. PV generation is often the cheapest source, but without storage it is not practical for many applications.
I think wind's problem is that wind installations get massive pushback in a lot of communities. I wish I were joking when I said a turbine near me was permanently shut down after a federal judge declared the plaintiff's objections for health reasons to be valid.
The court fight triggered a poison pill town bylaw that a turbine inactive for X amount of time has to be removed. All the idiots had to do was get an injunction and then tie the whole thing up in court for long enough.
1) lifetimes of utility assets are measured in decades, not years
2) the utility industry is not used to needing to pay attention to new technology and new information, and has a huge bias against renewables that goes back to the hard energy /soft energy split of the 1970s and 1980s
3) Utility planning models (IRPs, often for five years out) often use outdated info that is 3-5 years old when planning deployments for the next five years. So even if utilities used up to date info, it would take 5 years for them to shift strategy.
4) many utilities are not incentivized to install least cost generation, and their incentives and profit rates are different for things like installing transmission, distribution, etc
5) regulators of utilities are often even further behind the times than utility executives or completely captured (see for example Arizona)
Fair enough, but here's [1] how the picture looked just 4 years ago. Coal was at 33%, natgas at 33% an solar at 0.6%, wind at 4.7%. So a lot of coal was replaced by natgas, and a bit of wind. For some reason solar went up only by 1.7%
[1] http://web.archive.org/web/20170322133856/https://www.eia.go...
But even with your further comparison, you are falling prey to the same bad logic: solar can't be cheaper because we aren't already using it. You are looking at the decisions from 2010 to 2015 to evaluate the situation from 2021.
For changing technology, that's a really bad assumption. If nearly all the market for CPUs is Intel, then AMD comes out with a far better deal than Intel, do you evaluate how much better AMD is by the installed base across all computer? Of course not.
Check out what is being planned for the future. Some of the decisions are not great, because utilities don't yet know that storage is super cheap. But you'll find that in price responsive markets, nearly all new planned generation is wind, solar, and storage. Along with a few new gas plants proposed by people hedging against the dominant tech, or just through inertia.
https://www.greentechmedia.com/articles/read/report-renewabl...
For instance, even since the year or two that EIA has last updated their numbers to the numbers that just came out that include August 2021... the amount of total solar (far right column) generated (154TWh) vs total electricity generated (4087TWh) gives about 3.8% solar electricity for the last 12 months rolling. https://www.eia.gov/electricity/monthly/epm_table_grapher.ph...
I believe that the natural gas generators are still not at the "ditch as soon as you can" moment that hit coal hit a few years ago, and even then there's only so much solar being made right now.
The other factor is that solar manufacturing isn't being full-bore developed because as soon as you get a plant up, you're at risk of a paradigm shift beating you out. The promise of perovskites might be holding back silicon panel investment.
The Chinese, for as much as US media has been portraying them as a big bad, did a massive solid for the world in pushing solar panel manufacturing and battery manufacturing from the top down.
Granted they wanted to push everyone out/monopolize manufacturing, but who cares WHY, the net effect is to change our possible future for decarbonization. The US would never do that with the dominance of petroleum lobbying.
South Korea should also get a fair amount of credit for supporting Samsung in this direction.
A recent report I read on the BBC suggests that public bus transportation is 90g/km/passenger so my ancient old hybrid is better than taking the bus it seems?
IMO most bus operators do a terrible job at serving people and so it is no surprise that people don't ride.
I could imagine 8-15 passenger vans with on-demand routing would see much better utilization.
Additionally, transit vans don't necessarily need a commercial license to drive, which can reduce labor costs (in part by expanding the availability of the labor pool).
Out of service busses also travel faster than in service ones; as you note, you outpace busses on your bike.
Perhaps look up research instead of relying upon personal anecdotal observations by one person in one city, subject to observational biases?
The research is likely to be highly accurate given most public bus operators have very good data about ridership numbers and fuel use per vehicle and route.
ReaRead that, and then quit wasting everyone's time advocating a stupid idea that keeps coming bcback. Good transit advocacy should aim to get people to not drive.
Why would you think that a compact hybrid would have only slightly better CO2 performance compared to a city bus per passenger? It doesn't even make sense for the vehicle as a whole. Googling around, I came up with this:
https://www.bbc.com/future/article/20200317-climate-change-c...
...shows that your figure is 90g per km for the bus as a whole, not per km/passenger.
Every chart I've seen on co2 or fuel use per mile per person, busses, bicycles, and rail are king. Ride shares are the absolute worst because of all the wasted miles....followed by private SUV and car use.
If you do a google image search for "co2 emissions per passenger mile" you can see endless charts from different studies.
tl;dr:
- single-deckers are ~900g/km,
- double-deckers are ~1300g/km
- hybrid are 675-780g/km
So anywhere between 800% to 1500% worse than a decade-old hybrid.
Of course, if the bus has 24 to 45 passengers on it, then it is better on a per-passenger basis than me in my car with my wife and kid. I couldn't find any rider-ship data from TfL, but they often appear kinda empty outside of rush hours.
They are economical viable for solar for 3-4 hours every night, which is where they are being deployed. The article talks primarily about the UK with wind being the primary energy source, and so rather than a solar cycle of 24hrs you have wind fluctuations with lulls. The article calculate this to about 1200 GWh.
The articles I have seen on solar + batteries is a capacity of 80% for 4 hrs. I have not seen clear numbers but I would guess that they do get a fairly high discharge rate each day, thus providing an return of investment when the price is at its highest point each day. For wind you would need to have as much capacity as the worst lulls, and they would only really return profits during the lulls (if we do not count subsidizes). There would be some profits from slews, but most of the cost would be going to capacity that isn't needed for most part of the year.
He then worked on sustainable energy long before it became fashionable.
His early death was a significant loss. He blogged cheerfully from hospital til few days before his death
http://itila.blogspot.com/2016/04/perhaps-my-last-post-well-...
Rest In Peace David MacKay.
https://escholarship.org/uc/item/9js5291m
Here's the review: https://joshuaspodek.com/the-science-book-of-the-decade-ener...
I can't imagine he would.
For instance, the chapter on wind power starts by comparing a back of the envelope calculation on wind power with energy usage by a fossil car.
While there are occasional good points, I would not recommend reading this book unless you actually know something about the things the book is discussing. It's simply too misleading in important places.
In case anyone is wondering about this, there are real studies on the feasibility of 100% renewables by organizations that actually know this stuff and have done the modelling work.
I got a quite good intuitive sense of the scales involved, and at least for me Solar clearly reveals itself as an outstanding source of energy (at say 150W/m2).