The last (I hope) extract from the climate change chapter of Economic Consequences of the Pandemic. I’m in two minds about whether this is really needed. The group of pro-nuclear environmentalists seems to be shrinking towards a hard core who can’t be convinced (and some of them, like Shellenberger turn out to have been concern trolls all along). But every now and then I run across people who seem open-minded enough, but haven’t caught up with the bad news on nuclear.
Debates about decarbonizing electricity generation inevitably raise the issue of nuclear power. Since nuclear power generates no carbon dioxide emissions (except in the construction phase) it is a potential solution to climate change, with a strong body of advocates.
Some of this advocacy may be dismissed as point-scoring. Rightwing pundits who oppose any action on climate change simultaneously promote nuclear power as carbon free, with the aim of embarrassing environmentalist. There is, however, a small but vocal group of nuclear power advocates who are convinced that a massive expansion of nuclear power is the only way to replace coal-fired power.
Nuclear power advocates point out that the health and climate risks of nuclear power generation are far less than those arising from burning coal. They are probably correct to say that it would have been better to continue building nuclear power plants in the 1990s and early 2000s than to undertake the massive expansion of coal-fired power that actually took place, although alternative strategies like a major push to improve energy efficiency might have been better still.
But that is irrelevant today. The choice is not between new nuclear and new or existing coal. It is whether to allocate investment to building nuclear plants or to accelerating the shift to solar and wind energy.
The key problem is not safety but economics. New plants are safer and more sophisticated than those that failed in the past, but they are also massively more expensive to build, and quite costly to operate. The capital costs of recent projects in the US, France and Finland (none yet complete) have been around $10/kw, compared to $1/kw or less for solar. And, whereas solar PV is essentially costless to operate, the operating costs of nuclear power plants are around 2c/kwH. Even when solar PV is backed up with battery storage, it is cheaper to build and to operate, than new nuclear.
The facts speak for themselves. Over the last decade, only two or three reactors have commenced construction each year, not even enough to replace plants being retired. This isn’t the result of pressure from environmentalists or alarm about the safety of nuclear plants. The slowdown is evident in countries like China, where public opinion has little influence on policy decisions, and in countries where public opinion is generally favorable to new nuclear power. China failed to reach its 2020 target of 58 GW of installed power, and currently has only about 15 GW of nuclear power under construction. That compares to 55 GW of new solar and wind capacity installed in 2019 alone.
It is clear by now that large-scale nuclear reactors have no future. The last hope for nuclear power rests on Small Modular Reactors. The idea is that, rather than building a single large reactor, typically with a capacity of 1 GW, smaller reactors will be produced in factories, then shipped to the site in the required number. The leading proponent of this idea is Nuscale Power, which currently has a contract with UAMPS to supply a pilot plant with a dozen 60MW modules.
It remains to be seen whether SMR’s will work at all. Even if they do, it is not clear that the reduced costs associated with off-site manufacturing will offset the loss of the scale economies associated with a large boiler, let alone yield power at a cost competitive with that of solar PV.
In any case, the issue is largely irrelevant as far as the climate emergency is concerned. NuScale’s pilot plant, with a total capacity of 720 MW, is currently scheduled to start operation in 2029. Large-scale deployment will take at least a decade more
If we are to have any chance of stabilising the climate, coal-fired power must be eliminated by 2030, and electricity generation must be decarbonized more or less completely by 2035. SMRs, if they work, will arrive too late to make a difference.
None of this means that we should be in a hurry to close down existing nuclear power plants. Whenever there is a choice between closing down a coal or gas plant and closing down a nuclear plant, the best choice is to reduce carbon-based generation. A properly operating carbon price would make this clear.
{ 54 comments }
Josh 12.07.20 at 11:49 pm
It’s really not apples to apples to compare nuclear vs. intermittent power sources like solar/wind. My first thought is “why not both”, since the characteristics complement each other.
That said, the point about nuke construction in China is well taken. The technologist in me just hopes that there is a “Fail-Safe” nuclear implementation ( I particularly like the concept of 20 year thermoelectric generators you bury and replace, and get rid of most of the grid entirely), and it’s just a chicken and egg problem.
Any recommendations on a book on why nuclear hasn’t been able to work?
Brian Hanley, PhD 12.07.20 at 11:55 pm
I’m sorry John, but this anti-nuclear stance is simply wrong on all counts. It’s not correctly analyzed, and at the base, the biology of radiation is not understood.
Aidan 12.08.20 at 12:49 am
Nice encapsulated John.
Also worth noting is the exposure of thermal power stations (including nuclear) to drought and low water levels.
https://qz.com/1348969/europes-heatwave-is-forcing-nuclear-power-plants-to-shut-down/
https://www.cleanenergywire.org/news/increasing-droughts-can-destabilise-power-supply-germany-and-beyond-wwf-report
You’d be a poor gambler to take the bet that global warming won’t exacerbate that problem.
Omega Centauri 12.08.20 at 1:01 am
A minor correction:
“The capital costs of recent projects in the US, France and Finland (none yet complete) have been around $10/kw, compared to $1/kw or less for solar.”
Those units must have been watts not kilowatts.
In any case the issue with honest Nuclear protagonists, is not that they don’t get Nuclear, but that they don’t get renewables, and won’t accept that a mixture of overbuild, storage, and long distance transmission will ever be adequate to the task of keeping the lights on during stretches of unfavorable weather. It doesn’t help that the early papers by Jacobson made unrealisticly heroic assumptions about hydro storage, and simply dismiss the more recent studies by him and other authors as so much BS. So the debate quickly degenerates into both sides calling the other stupid.
I do think that small scale reactors may meet some niche needs, particularly in remote areas such as the Arctic that may not have decent renewable resources. I also suspect we will finally get workable fusion, but its likely to be just as expensive as fission, so probably won’t be a bug player either.
The problem with renewables plus storage, in my mind is that its largely a chicken and egg problem. Until we've made a commitment to very high renewables penetration, there just isn't much of a market for storage -particularly long-term storage, so its development is starved of sufficient funding. And then its hard to get buy-in for aggressive renewables, because the storage solutions have not been demonstrated at scale.
Phil Carlson 12.08.20 at 1:57 am
John,,
Cost comparison between renewables and nuclear is very complex. Yes, construction of nuclear is initially much more expensive but a nuclear reactor lasts 80 or maybe 100 years. Solar and wind need replacing every 30-40 years. Also, renewables will require backup which means costly batteries, natural gas, or both. It is just too simplistic and inaccurate to claim that nuclear is too expensive.
dilbert dogbert 12.08.20 at 2:28 am
Build enough nukes and you have to have a way to start counting atoms of plutonium and keep them sequestered.
The Dark Avenger 12.08.20 at 2:32 am
Considering you wrote a book on the subject, Prof Handley, I’m not sure I get where you’re coming from.
OTOH:
Understanding Radiation Biology: From DNA Damage to Cancer and Radiation Risk 1st Edition
https://www.amazon.com/Understanding-Radiation-Biology-Damage-Cancer/dp/0367255154
Jacques Distler 12.08.20 at 3:25 am
Not a single reactor on this list operated for more than 50 years. And most shut down well before the 50 year mark. Reactors operate in a punishing radiation and thermal environment. Expecting them to last 80-100 years is beyond fantasy.
When it come to a wind turbine or a bank of solar panels, the nominal 30-40 year lifespan is just that … nominal. There’s nothing in a photovoltaic cell that “wears out” after 40 years. Nor, with proper maintenance, is there any reason to expect a wind turbine to stop spinning.
You can’t forget the cost of nuclear fuel (which is cheap compared to fossil fuels, but adds up over the total lifetime of a reactor). Nor can you forget the cost of dealing with the nuclear waste produced. The total cost of that is something no one has really figured out.
lathrop 12.08.20 at 3:45 am
I’m skeptical about nuclear but willing to consider the possibility, given current conditions.
A chief issue for me is the problem that the costs of nuclear failures — disasters of various kinds — in the reactors, and waste handling, is potentially quite high, whether due to incompetence, profit-seeking, bureaucracy, laziness, malfeasance, inattention, terrorism, or systemic injustice. Who trusts capitalists or government to do it right?
I grew up with Three Mile Island and Chernobyl and was in the Livermore Action Group. Were we wrong?
Alan White 12.08.20 at 3:53 am
The county I live in here in Wisconsin (Manitowoc) on Lake Michigan is a literal embodiment of much of John’s post. At its north end is a still-functioning nuclear power plant (Point Beach) which is incidentally just about 5 miles south from one in the next county (Kewaunee) which recently starting decommissioning because of bad economics (and the decommissioning takes decades and lots of money too). But also literally across the road from the Point Beach facility is the now-nearly-finished Two Creeks solar facility with half a million panels on 800 acres–talk about a juxtaposition of contrastive energy producers. And Wisconsin plans at least one other massive solar array–and the economics and the carbon-savings are highly touted by the investors. FWIW
bad Jim 12.08.20 at 5:22 am
One of these days we will master fusion power and all our problems will be over. Well, not all; I will be dead, but that will only be a problem for me.
Radioisotope thermoelectric generators for deep space probes are definitely the way to go, at least for now. Photovoltaics only take one so far.
Maybe there’s some design for a small modular reactor that offers some hope. NuScale isn’t it.
nastywoman 12.08.20 at 7:20 am
”Is nuclear power the answer”
No.
bill tilles 12.08.20 at 11:57 am
Every idea here regarding nuclear is totally correct. The current iterations are only being built by those completely subsidized or financially insane. And SMRs will come too late to help. Nuclear only makes sense in submarines and other zero oxygen environments.
m 12.08.20 at 12:26 pm
I still fail to understand what this has to do with the pandemic.
Tim Worstall 12.08.20 at 1:29 pm
“that they don’t get renewables, and won’t accept that a mixture of overbuild, storage, and long distance transmission will ever be adequate to the task of keeping the lights on during stretches of unfavorable weather.”
That’s not quite the right point. What is is “At what cost?”
“compared to $1/kw or less for solar.”
Accepting the correction to W not kW that’s still not the right answer. Because that’s the cost of 1 W in situ. Without the “overbuild, storage, and long distance transmission”.
Personally I’m perfectly willing to accept that renewables are now cheaper. But I’d still want to insist that costs are worked out on a like for like and total basis.
“There’s nothing in a photovoltaic cell that “wears out†after 40 years.”
Eh? Then why does performance degrade over time?
Robert Zannelli 12.08.20 at 2:43 pm
General Atomic’s Energy Multiplier squared design shows some promise in terms of safety , sighting and cost IMO ( I am a retired Nuclear Engineer) . It’s true that nuclear has a long lead time but I don’t think it should be off the table given the advances in generation four designs And we definitely should not be shutting down operating plants to replace thes power they generate with natural gas or in Germany even worst , coal. We should of course expand solar and wind as fast as we can.
Matt 12.08.20 at 7:59 pm
Geophysical constraints on the reliability of solar and wind power in the United States Energy and Environmental Science, 2018
(Full paper from escholarship.org)
I originally saw this paper cited by a pro-nuclear activist who highlighted its “12 hour energy storage” scenarios as requiring implausible quantities of batteries.
However, it shows other scenarios including no battery storage at all. It models different combinations of over-generation, storage, and regional integration via new transmission capacity. If you look at Figure 2, the US can supply 65% of electricity demand from wind and solar without any storage or significant new transmission capacity, if it overbuilds solar/wind capacity by 50%. (That doesn’t mean 35% fossils, either; the residual 35% would come from existing nuclear, hydro power, geothermal, and some smaller residual of fossils.)
The mix that actually gets built will probably have less overcapacity and non-zero battery storage.
The rise of battery electric vehicles is undermining the case for nuclear power nearly as much as the improving levelized cost of wind and solar generation. Any scenario where BEVs come to dominate globally over internal combustion is also a scenario where batteries have become much cheaper and more abundant. Cheap and abundant batteries in BEVs undermine the there-is-no-alternative argument for nuclear power’s reliability in two ways:
1) Vehicle charging demand can be flexibly dispatched to match supply on a scale of hours to days, and the aggregate demand for BEV charging will be the largest single electricity use if ground vehicle fleets become majority-electric.
2) The same falling costs that permit BEVs to compete globally against ICE will also make it affordable and routine to build grid attached storage to firm up generation from intermittent wind/solar primary sources.
If you have a dour outlook on BEVs and believe that they cannot displace fossil fueled vehicles, that’s possible too, but it also means that any big nuclearization push is also going to leave a major portion of global emissions unaddressed.
Alex 12.08.20 at 11:35 pm
As noted by Omega Centauri, my pro-nuclear stance largely comes from an absence of belief in energy storage (and without storage, intermittent sources are nearly meaningless). Every time I look into it, and I admit I haven’t super recently I can’t come up with anything vaguely resembling a system design that actually converts solar and/or wind into a functional power grid.
The other major factor is speed. As noted in the post itself, we need to move, and move fast. Which means that we should go with the technology we have, rather than the technology that we might have a decade or two from now. There can be no debate that from engineering or financial perspectives that nuclear would work to decarbonize the grid (France absolutely proved that decades ago). If we build nuclear but I turn out to be wrong and energy storage is a solvable problem, well, no biggie. Nuclear plants don’t last forever, and will eventually need to be replaced. If we don’t build nuclear, but I turn out to be right, well, uh oh.
Re intermittent power sources: I don’t pay the electric company to make my meter spin. I pay them the ensure that when I flip the switch, there is enough power available that the light comes on: it is more accurate to say that I pay them for kWs, NOT kWHs. A better setup, and a good holistic mental model for the actual whole system cost would have the utility bid out kW of supply for an entire year. IE, a power plant would contract to guarantee the first (second, third, etc…) MW over a full year. In that regime, an intermittent power plant would need to hire someone else (storage, a coal plant, whatever) to provide the power when the sun isn’t shining or the wind isn’t blowing, as appropriate. Without something like that the market is seriously distorted by not paying for the necessity of guaranteed production.
Cranky Observer 12.09.20 at 2:34 am
Formalized markets such as you describe were introduced in the US (and to a certain amounts the portions of Canada and Mexico that choose to participate in them) in 1994 and made far more extensive in 2003. FERC has deputized Chicago School “market monitors” to “oversee” – that is, to control and discipline [1] – the nominally independent system operators and to stifle the nominally superior authority of state public service commissions over in-state transactions where such commissions still exist. FERC and the IMMs have been introducing more and more “markets” every few years in an attempt to patch together the whole system with epicycles; the last market introduction I was tracking was the 50 year (!) forward capacity market. Let’s just say that all this has not worked out the way the freshwater boys predicted in 1994, and while those changes (let’s not call them “reforms”) to the provision-of-electricity system in North American may have been driven by some need for change in the industry the post-2003 changes have been pure Enron-style arrogance and financialization of a public necessity.
[1] ‘discipline’ in the Republican sense of 5% guidance, 95% punishment
Cranky Observer 12.09.20 at 2:41 am
So honest answer persons of CT: since our last discussion of nuclear power here two weeks ago how many of you have searched out and submitted applications to work at nuclear power plants or the industries that support them (forging shops, heavy machine shops, nuclear waste disposal contractors, spent fuel pool welders, etc)? Long hours, on-call 365 days/year, high temperature and humidity, annual radiation dose, remote locations, high pay & decent medical benefits, no pension – what’s not to like for an out-of-work academic or adjunct lifer?
IIRC the last post spiraled down into a discussion of how “we” could provide incentives – presumably to someone other than ourselves – to devote their lives to those jobs. Staffing is a major problem for all heavy precision industries and very much so for nuclear as my generation of late Boomers approaches retirement and there is no large cohort behind us. But this is a policy theory site so… can openers for all.
John Quiggin 12.09.20 at 3:11 am
Cranky @20 I don’t remember this previous discussion. Can you link to it?
CHETAN R MURTHY 12.09.20 at 6:31 am
“Nor can you forget the cost of dealing with the nuclear waste produced. The total cost of that is something no one has really figured out.”
Frankly, I just don’t understand nuclear advocates. They wave their hands at this issue, and pretend that that’s an answer? Insane. Just insane. It’s like some caricature of capitalists-gone-mad: they pretend that every spillover cost is nonexistent, and then, well, sure, it’s great! Just Great!
And the same is true when it comes to accidents: we’ve already got a number of accidents that are basically uncontained catastrophes, and nobody, NOBODY has any idea how to prevent these things. All the fancy stories about how yeah yeah, we know how to prevent these things, founder in the basic fact that the technology is incredibly dangerous, and requires commensurate attention to detail in every step of its operation: something that we just cannot expect from operators, as we’ve seen to our chagrin.
It’s just ridiculous. And then they come up with these bullshit numbers like “nuke plants run 100yr”? It is to laugh.
What a bunch of gaslighters.
John Quiggin 12.09.20 at 6:59 am
m@14 Some of the earlier extracts explained links and analogies between the covid and climate crises
Hidari 12.09.20 at 10:47 am
@15
You seem to be implying that the renewable ‘intermittency’ problem has not been solved, or at least, not solved in a cost-effective way. Could you talk more about that?
You also seem to be implying that when all costs are taken into account it is problematic whether or not renewables are in actual fact cheaper than fossil fuels. Again I haven’t seen that claim before (I know you didn’t precisely state this but you seemed to imply that the ‘renewables are cheaper’ case is less of a slam-dunk than is usually taken for granted). Again, what evidence do you have for this? Genuine question.
Cranky Observer 12.09.20 at 12:21 pm
JQ,
Absolutely. Here is the <a href=”https://crookedtimber.org/2020/11/21/controversy/#comment-806349>link to my initial comment in that thread.
Mike Huben 12.09.20 at 12:43 pm
JQ @ 23:
Ah, the “consequences” in the title has mislead me.
Mark Uber 12.09.20 at 1:42 pm
I agree that nuclear power is not the isolated solution to provide carbon neutral power generation. However, that is not to say it does not have it’s uses. Nuclear reactor designs that are capable of using existing spent fuel from existing reactors would be capable of supplementing the power grid while converting our existing stockpile of radioactive waste ( with >30,000 year half life) to much smaller quantities of radiological material with half life of 300 years or less. Ideally, you convert one of the existing power plants where there is existing power grid infrastructure (and existing spent fuel) to a modular breeder reactor and processing plant design.
Chris Armstrong 12.09.20 at 2:17 pm
Since this is the question you actually asked, my gut reaction is that this doesn’t go in the book. It sounds more like something you’d publish independently, in whatever format, and point the interested reader towards.
Matt 12.09.20 at 6:44 pm
There’s nothing in a photovoltaic cell that “wears out†after 40 years.
Eh? Then why does performance degrade over time?
Most of the performance degradation in silicon PV systems, and their eventual failure, does not happen because cells stop working. The top failure mechanism is degradation of solder joints and connections between cells. It happens primarily due to thermomechanical cycling (day/night temperature changes, different coeffiicent of expansion between silicon and metals, metal fatigue). A significant runner-up for failures in metal connections is electrically accelerated chemical damage to internal soldering/wiring. The most common encapsulant material is ethylene vinyl acetate, a polymer that releases acetic acid when it breaks down under conditions of high temperature and UV exposure. The acetic acid is corrosive and electrically conductive. Damage to internal solder joints and wiring can cause local hot spots where too much current flows through an undersized conductive path. These hot spots can lead to open circuit failures, melting through the backsheet, cover glass cracking, and occasionally even fires.
Degraded EVA encapsulant also discolors, which lowers module output by reducing the amount of light making it to cells.
The acetic acid from degraded EVA has also proven to be a major contributor to the premature cracking of polyamide backsheets and the so-called “snail trail” degradation pattern sometimes seen in damage to screen printed cell contacts. Finally, the acid’s mobilization of sodium ions from cover glass makes cells more vulnerable to potential-induced degradation (one of the few module degradation paths caused by semiconductor-level damage to the cell.)
There are so many module degradation mechanisms caused or exacerbated by EVA encapsulant that I believe modules can easily get an extra 10 years of useful life just by switching to thermoplastic olefin encapsulants instead. TPO encapsulants are more durable and don’t release any acid from degradation. Some manufacturers have already switched. But modules are still improving fast enough that buyers mostly don’t buy with an eye to maximizing module lifetime. The usual 20-25 years of useful life achievable with EVA are still satisfactory for most buyers.
Another few degradation mechanisms:
Cover glass can go hazy over time from mechanical scratching or chemical etching by alkaline dust plus water, which reduces the amount of light reaching cells.
Cell micro-cracks, initiated by e.g. people walking on top of modules or other mishandling, can grow over time due to thermomechanical cycling. The cells that are in series with the cracked cell will lose some or in extreme cases all of their power due to weakened/opened circuit.
Boron-oxygen complexes form over time in illuminated cells that were doped with boron and formed from Cz-type monocrystalline silicon. The complexes form recombination centers for carriers and reduce the cell power output after a year or two of field use. This mechanism was common in older mono-type modules but has mostly been mitigated today by additional cell processes or choice of different doping schemes.
I tried searching for a few minutes for a review article that mentioned all these different issues, but didn’t find one. Plug any of the individual mechanisms into Google Scholar if you’d like more details.
John Quiggin 12.09.20 at 6:46 pm
@Cranky, thanks, that had slipped my mind.
Michael Cain 12.09.20 at 8:40 pm
If we are going to electrify transportation, we will be deploying enormous amounts of battery storage as part of it. I am surprised that more time isn’t spent discussing how that might be used for grid stabilization.
John Quiggin 12.09.20 at 10:06 pm
@31 There is quite a bit of discussion about this. My summary: there are big stabilization benefits from G2V, that is, using excess electricity (at midday, typically) to charge car batteries. Much less from V2G, allowing car batteries to return electricity to the grid. That depends on making charging infrastructure available where cars are parked in the daytime.
Area Man 12.10.20 at 3:02 am
The solar panels adorning my garage roof, which are fairly new, are supposed to lose performance at a rate of 0.5% per year. Assuming this is linear, that would still leave them with 80% of their original capacity after 40 years. It’s not exactly a huge deal.
MP 12.10.20 at 10:49 pm
Solar panels lose about 3% of capacity each year. More in hot climates. The best German panels are predicted to last about 40 years but the cheap Chinese ones last about 20 and have high failure rates.
WWS is intermittent and batteries to fill the still, dark times are prohibitively expensive. Look at a Tesla powerwall price. Assuming Australians use about 8kW (university of Melbourne did a study called “Solar Energy Without the Hot Air†a few years ago) per person, 27/7 for all energy needs and we would need say 3 days of storage and there 25 million of us…That is 14,400GWh of storage. A powerwall2 stores about 13.5kWh and costs $10,000.
That would mean about $10 trillion to go 100% WWS in Australia using batteries at the retail price. Ok, economies of scale, blah blah blah… say a factor of ten cheaper. Still over a trillion, that is a thousand billion dollars just for the batteries. Add in the panels, the poles and wires, the installation and all the grid stabilisation technology that doesn’t exist yet and it is a lot of money.
Pressurised Water Reactors nuclear power plants cost about $6billion per Gigawatt. And That is enough to build 1,700 one Gigawatt nuclear power stations. And Australia needs 200 GW of generating capacity.
John Quiggin 12.11.20 at 12:04 am
@34 “University of Melbourne did a study called “Solar Energy Without the Hot Air†a few years ago”
Are you thinking of the David Mackay book “Sustainable Energy – without the hot air” published in 2009? Even allowing for the fact that any prediction made at that time is out of date, Mackay’s arguments (similar in kind to yours) look pretty silly now
More generally, I’ve seen heaps of sketchy calculations of the type you present, purporting to show that renewables can’t possibly supply more than x per cent of electricity needs, where x was 30 per cent a while back, then 50 and is now approaching 100.
Omega Centauri 12.11.20 at 1:03 am
Lots of problems with those -without the hot air studies. They assume energy comes from a point source, rather than a transmission system that spans the size of weather systems with WWS distributed across these different places. They always assume you have to replace “primary energy” with WWS electricity (a discussion we had a few posts back). Usually they assume absolutely zero WWS for two weeks, while demand is at its maximum.
The other issue I have, is that they want to maintain the current paradigm, which is any user can use maximum power at any moment, regarding any gap between availability and desire as a terminal event. Yet having to reduce demand in a once in a blue-moon pinch, doesn’t have to be a big deal. Reduction of many energy consuming activities to MOL (Minimum Operation Level) for a day or two wouldn’t be a big sacrifice for a society/economy thats ready for it. Lots of things we would barely notice: Street lights at half intensity we would barely notice. Delaying clothes drying for a couple of days and so on would be a small price to pay to save the climate. There is always a tradeoff between convenience/reliability and cost, and a tradeoff is always made. So expert simulations using real weather data can determine day many hours per year on average we need to curtail some consumption, and the economist can tell us the costs.
steven t johnson 12.11.20 at 1:51 am
Delaying clothes drying a couple of days can mildew the wash, which strikes me as a problem. Draping wet laundry over all available surfaces isn’t convenient either.
My town is not replacing street lights in my neighborhood and a large fraction of them are drawing zero, not half, power.
Dimmer lighting is also imposed by reducing the wattage on bulbs, though the rather more expensive LED still seem brighter. No doubt aging eyes will simply have to turn on more lights, earlier.
How large are the improvements in transmission losses? Do we need to chop down many more trees to run power lines through woods for a better transmission network?
Matt 12.11.20 at 3:04 am
Solar panels lose about 3% of capacity each year. More in hot climates. The best German panels are predicted to last about 40 years but the cheap Chinese ones last about 20 and have high failure rates.
The most comprehensive review I’ve read is Photovoltaic Degradation Rates — An Analytical Review from the National Renewable Energy Laboratory. It found median capacity loss of 0.5% per year.
The 2020 PV Module Reliability Scorecard from PV Evolution Labs is great reading. It mentions all the degradation mechanisms I talked about before, and more. It has pictures and history as well as the final disclosure of top performing modules.
Most of the top performing modules for 2020, like most of the modules now made in the world, are from China. 10 years ago there was a big quality gap between upstart Chinese solar manufacturers and the rest. I’d say that top Chinese manufacturers have caught up on durability and efficiency and are now leading the industry forward.
John Quiggin 12.11.20 at 3:55 am
“Draping wet laundry over all available surfaces isn’t convenient either.”
This is fairly typical of this mode of discussion. It slides rapidly from “without always-on baseload power, industrial civilization will collapse, so we can’t do anything about global warming” to “unless wind and solar are better on every conceivable dimension, we have to stick with coal and nuclear”.
jerlich 12.11.20 at 7:40 am
It seems that no one has mentioned that next-gen breeders use the nuclear waste, that is currently quite expensive to store, as fuel. Once you add that to the cost-benefit analysis is nuclear really more expensive than solar/wind?
MP 12.11.20 at 8:08 am
@John Quiggin. You say my earlier comment is wrong because The UOM article is old and so out of date and that is enough to discount the thrust of my comment. All I was relying on the UOM article for was the 8kW of energy use per capita 24 hours a day in Australia for all uses. That is electricity, transport, agriculture, mining, smelting, concrete process heat and industry. Use a better number if there is one and do your own numbers. I say 3 days of battery storage to stabilise the grid. I think that is aggressive and a conservative estimate might say three weeks of storage. What number would you use? I say a Tesla PowerWall stores 13.5kWh and costs $10,000. What number do you use?
Anyway: 8kW x 25 million people is 200GW time 72 hours in 3 days is 14,400GWh.
That is 1.1 billion Tesla PowerWalls at $10,000 each or a bit over $10 trillion. (A bit over five years of GDP) And I said that is maybe an order of magnitude out. I doubt it is that far out because there is not enough Lithium refining capacity to make that many batteries in short time and the price would rise massively.
To install 200GW of PWR nuclear generation capacity at $6billion per GW would cost about $1.2 trillion.
Rather than vague “your figures are out of date†ad hominem arguments and given you want to get to the truth please answer the specifics and show me I’m wrong.
John Quiggin 12.11.20 at 11:03 am
MP, why don’t you take this up with AEMO, who showed the feasibility of 100 per cent renewables for Australia years ago (as have many others).
Peter Erwin 12.11.20 at 11:18 am
There’s an interesting recent paper by <url=”https://www.sciencedirect.com/science/article/pii/S254243512030458X#bib2″>Eash-Gátes et al. (2020) that tried to figure out what has been going on with nuclear-power-plant construction costs, at least in the US.
From the Summary:
They argued that only about 1/3 of the const increases over the last few decades can be attributed to increased safety requirements. They also note that industry and government forecasts for future nuclear-power generation often assume naive and historically unfounded cost reductions for future construction.
From the Introduction:
Petter Sjölund 12.11.20 at 1:32 pm
The main advantage of nuclear power is its huge popularity among the political Right. It is the only kind of large-scale COâ‚‚-reducing project that will easily pass through a right-wing majority parliament. This also means that nuclear power plants can be built near wealthy populated areas without triggering any of the protests that a wind or solar farm would.
Tim Worstall 12.11.20 at 3:07 pm
@24
“You also seem to be implying that when all costs are taken into account it is problematic whether or not renewables are in actual fact cheaper than fossil fuels. Again I haven’t seen that claim before (I know you didn’t precisely state this but you seemed to imply that the ‘renewables are cheaper’ case is less of a slam-dunk than is usually taken for granted).”
It’s important, at least I think it is, to have a price “to do what?” Only when we’re comparing that truly like with like – what’s the price to do this thing? – can we be said to be comparing.
“The price of solar” isn’t really a thing. To take an absurd example, that price in the Atacama, to power your lithium extraction, is different from trying to run McMurdo through the winter. The latter will require much more battery, like perhaps 6 months worth of consumption.
Sure, getting nuclear, or gasoline generators, or hydro, or whatever, to those two different places and uses also have their own, specific, costs. But we’re going to do a lot better in our sums if we say “what’s the price of daylight energy in the Atacama?” than trying to peg one cost to something as variable as “renewables”.
So, when we discuss such prices those renewables, which need overbuild, battery storage and all the rest. Fine, so, can we please be using the price of the renewable that includes the overbuild, the battery, etc, not just a per W number for a solar cell?
I’ll agree entirely that renewables are, now, cheaper for some to many uses. I’d be astonished if they were for all (McMurdo comes to mind). Our consideration of all this is going to be better served by using an all in and total price to do what in which place at what time?
After all, what we really want to know is, if we’re in Stavanger, is how much dispatchable power do we need in Stavanger over time? So, how much is that much dispatchable power going to cost us? If we do it this way with solar and wind and batteries and long distance transmission and by reducing consumption through insulation and so on, then OK, what’s the price? And if nuclear, then what’s that all in price – including radioactive disposal. And if fossil then what – including emissions damages.
A different example, using electricity to power lithium extraction – absolutely fine. So is using electricity to smelt aluminium. But that second requires much closer control of the volume of electricity over time. Any significant variation in power available is just a boring nuisance to the first process. At some level of interruption it kills off a few hundred millions dollars of plant.
How much does the power cost, in which place, for what purpose, when? That’s the important number.
It doesn’t bother me in the least what the actual answers are (for Stavanger, almost certainly hydro but), it’s that we must be counting correctly to come to a useful answer.
As @36 points out “There is always a tradeoff between convenience/reliability and cost, and a tradeoff is always made.”
Sure, so we want the right numbers to be able to consider the trade off, whatever it is.
steven t johnson 12.11.20 at 4:18 pm
The mildewed laundry example ignored by John Quiggin was drawn from experience during a power failure. So, when Omega Centauri blithely assumed one could simply let the wet laundry wait, it really seemed like Omega Centauri neither knew what restricting consumption means in daily life, nor cared to know. The impression was reinforced by the airy thought that street lights are just some frivolous luxury.
The thing is, of course, is that sharply raising prices on power and generally restricting consumption are processes well under way. They have been for some time now. Without budgeting, where I pay power bills ahead of time (and yes, that is exactly what happens) I would be facing (again) a choice between heat or other bills or even food in February.
Of course I pay more for the unearned privilege of having both. I am in effect offering an interest free loan to the power company for some portion of the year, though I doubt any analysis of interest in the economy includes this any more than it includes the interest on loans by workers to employers who delay their wages. The next step I suppose is to condemn air conditioning as moral decadence. Heat stress and exhaustion, even heat stroke, are no more significant than mildewed laundry?
Given that the long standing efforts to restrict consumption by the rabble and the long decline by the demon coal have not been sufficient to leave the world unchanged, much less restore it to the idyllic past, how much more do you want? Every indication so far is that the saviors of the Earth are like a Victorian factory owner wanting a fair day’s work for a fair day’s wage from his employees.
Matt 12.11.20 at 6:16 pm
MP: this post started out explicitly talking about decarbonizing electricity. You are talking about primary (thermal) energy: “All I was relying on the UOM article for was the 8kW of energy use per capita 24 hours a day in Australia for all uses. That is electricity, transport, agriculture, mining, smelting, concrete process heat and industry.”
That effectively counts the wasted energy from existing fossil-powered generators and industrial processes as something that needs to be replaced. It’s true that these other processes need to be reworked to use electricity instead of fossil combustion, whether the electricity comes from nuclear plants or otherwise, but Australia does not currently consume 8 kW per capita of electricity. (Determining how much electricity it would need once all processes are electrified would be well beyond the scope of one blog post. But it’s going to be less than 8 kW/capita.)
Australia generated 204.5 TWh of electricity in 2019. With a population of 25.2 million, that comes to 8.34 MWh per capita per year, e.g. 0.953 kilowatts annualized average electrical power.
Peter Erwin 12.11.20 at 9:20 pm
The capital costs of recent projects in the US, France and Finland (none yet complete) have been around $10/kw, compared to $1/kw or less for solar.
I think those numbers are off by a factor of a thousand (i.e., in both cases those would be the capital costs per watt, not per kilowatt), no?
MP 12.12.20 at 1:10 am
@John Quiggan There is no doubt that it it technically feasible to go 100% renewable. The science and the engineering is certainly achievable. But at what cost?
Here is one possible roadmap of what the USA would need to go 100% renewable:
… 328,000 new onshore 5 MW wind turbines (providing 30.9% of U.S. energy for all purposes), 156,200 off-shore 5 MW wind turbines (19.1%), 46,480 50 MW new utility-scale solar-PV power plants (30.7%), 2,273 100 MW utility-scale CSP power plants (7.3%), 75.2 million 5 kW residential rooftop PV systems (3.98%), 2.75 million 100 kW commercial/government rooftop systems (3.2%), 208 100 MW geothermal plants (1.23%), 36,050 0.75 MW wave devices (0.37%), 8,800 1 MW tidal turbines (0.14%), and 3 new hydroelectric power plants (all in Alaska).
And the authors go on to say:
That will meet average demand. Then you need 1,364 additional new CSP plants and 9,380 50 MW solar-thermal collection systems (“for heat storage in soil”) “to produce peaking power, to account for additional loads due to losses in and out of storage, and to ensure reliability of the grid.”
Just the first item of 328,000 on shore, 5MW wind turbines is gobsmacking. So can 100% renewables be achieved… probably but nothing like cost effectively. And when renewables advocates say “… and it will create so many jobs.†What they really mean is it will cost a bomb.
This lecture by Professor Kelly is a useful reality check:
https://www.thegwpf.org/content/uploads/2019/11/KellyWeb.pdf
Let’s not bet on discovering unicorns.
By the way, nuclear power is discounted because of cost when compared to say gas generated electricity. Many planes have gas turbines that produce megawatts of power… these engines are mass produced and that is the engineering learning curve.
By the same token, Nuclear reactors, especially molten salt reactors (MSR) could also be mass produced cheaply if the regulatory environment were adjusted. For example many of the cost drivers in a PWR come from proving the containment vessel is safe in the event of a failure. You can’t build any reactor unless you submit the paperwork and assessments take years. But an MSR has no water, cannot explode, is fail safe and produces virtually no waste. In fact they consume existing wastes.
But as Professor Kelly shows, WWS is trying to concentrate very low density energy and is necessarily resource hungry. There is no engineering learning curve available. That is down to the laws of physics however inconvenient.
I will read what I can find from AEMO. Audite et alteram partem as they say.
Over and out.
John Quiggin 12.12.20 at 2:29 am
Are you aware that GWPF is a denialist outfit? https://en.wikipedia.org/wiki/Global_Warming_Policy_Foundation
MP 12.12.20 at 3:40 am
I don’t care if they are a “denialist outfit†if the information they publish is accurate. Argue the facts not the people please. Saying that they are a “denialist outfit†is not an argument.
What about this one: https://euanmearns.com/the-dream-of-100-renewables-assessed-by-heard-et-al/
Maybe these guys are a “denialist outfit†too but they report a study that comes from an acceptalist outfit, Heard et al, and they give the, or at least an, AEMO study a score of 3.5 where at least a 7 is needed to address the major impediments to 100% renewables.
Omega Centauri 12.12.20 at 3:55 am
I never meant to imply that one needs to let wet laundry sit, but rather that when the power authorities say we need to conserve for the next three days, that one put off doing the laundry if possible. We already do this during extreme heatwaves in California. My sister plans when to do laundry so she can take advantage of good outdoor drying weather. With a little planning a lot of consumption can be flexed. And many industries with high industrial demand have been doing this for decades, they are offered lower priced power in return for the requirement to curtail some of the demand when asked by the grid operator. Even for the example of the Aluminum plant, the MOL (Minimum Operating Level) is that which keeps the electrolytes from solidifying, but produces little output. The ratio of MOL to power for regular production is the important number here. And of course the economic tradeoffs, between cheaper power and the occasional loss of production capacity would go into the decision making. Lulls in renewable production can be predicted by weather forecasts, and flexing actions are plannable.
John Quiggin 12.12.20 at 5:20 am
“I don’t care if they are a “denialist outfit†if the information they publish is accurate.” That’s your problem, right there.
Seriously, I’ve wasted enough of my life playing whack-a-mole with denialist talking points. As it happens, I responded to Heard et al when their report came out, and I’m not going to repeat myself https://johnquiggin.com/2017/04/10/burden-of-proof/
Go and read the AEMO reports, then come back if you have something serious to say.
steven t johnson 12.12.20 at 3:46 pm
Thanks to Omega Centauri for clarifying the wet laundry issue.
Comments on this entry are closed.