Nuclear Energy is 50% Better than Solar for Lifetime CO2 Emissions

There was a recent updated analysis of various life-cycle energy impact measurements for fossil fuels, nuclear, solar and wind.

Lifetime CO2 Emissions per Kilowatt-Hour

Solar is 50% worse than nuclear for lifetime CO2 emissions per kilowatt-hour. Nuclear and wind tie for the best.

The study finds each kilowatt-hour of electricity generated over the lifetime of a nuclear plant has an emissions footprint of 4 grams of CO2 equivalent (gCO2e/kWh). The footprint of solar comes in at 6gCO2e/kWh and wind is also 4gCO2e/kWh. The best solar technology in the sunniest location has a footprint of 3gCO2/kWh, some seven times lower than the worst solar technology in the worst location (21gCO2/kWh). If we were running a civilization on ground based solar we would use a lot more bad locations.

In contrast, coal CCS (109g), gas CCS (78g), hydro (97g) and bioenergy (98g) have relatively high emissions

Energy Returned on Energy Invested

It is better to get greater amounts of energy generation that the energy needed to build and operate power plants. Energy Returned on Energy Invested (EROI) is the metric for this. Inverting EROI is energy embodied.

11% of the energy generated by a coal-fired power station is offset by energy needed to build the plant and supply the fuel, as the chart below shows. This means coal plants have an EROI of 9:1.

Nuclear power is twice as good as coal, with the energy embedded in the power plant and fuel offsetting 5% of its output, equivalent to an EROI of 20:1. Wind and solar currently are better at 2% and 4% respectively, equivalent to EROIs of 44:1 and 26:1.

Typical solar currently generates about 350 watts per standard panel module. Florida Power & Light wants to install solar farms with 30 million panels. This would be 10,500 megawatts. This would only generate the power of three gigawatts of nuclear power because solar generates about 25% of capacity versus 90% capacity for nuclear energy.

By 2050, a 2017 Nature study forecasts that EROI will get worse for new energy projects. It might be tougher to extract resources for fossil fuels and solar and wind would be built in inferior locations.

There is a video from China touting its energy megaprojects. However, the video also gives a clearer sense of the scale of materials needed for coal, solar, wind and nuclear.

Each solar unit might only generate 1 kilowatt for smaller 3 square meter mirrors or panels or tens of kilowatts for long troughs. However, tons of steel and cement are needed for each kilowatt.

Wind versus Nuclear

It would take about eight hundred five-megawatt wind turbines to generate the energy of one gigawatt nuclear plant. Each wind turbine has about 875 tons of steel and cement. This includes the foundation. Energy Skeptic had an analysis of a single wind turbine.

This is why wind power uses about 5 to 10 times more material to generate the same amount of power as nuclear. Solar also uses a multiple of the land and materials to generate the same power as nuclear.

Each wind turbine takes about 3 weeks to build from excavation to operation
40 to 100 geopiers installed for stability, weight unknown. These are part of the foundations and the amount of foundation various based on soil conditions.

Set 96,000 pounds of reinforcing steel rebar = 48 tons
53 concrete trucks pour foundations. If each truck can haul 8 cubic yards at 2538 lbs/yard * 53 = 1,076,112 pounds = 538 tons
Move 1,500 cubic yards of soil @ 2,200 lbs per cubic yard = 3.3 million pounds = 1,650 tons
3 blades : each 173 feet long and 27,000 pounds for 81,000 pounds = 40.5 tons
8 truckloads to deliver turbine components
Nacelle: weight 181,000 lbs = 90.5 tons with the generator, gearbox, and rotor shaft
Hub: weight unknown
Base tower height 53 feet 11 inches, weight 97,459 lbs = 48.7 tons
Mid tower height 84 feet 6 inches, weight 115,587 lbs = 57.8 tons
Top tower height 119 feet, weight 104,167 lbs = 52 tons
Final tower height to blade tip when fully extended 442 feet

Nature – Understanding future emissions from low-carbon power systems by integration of life-cycle assessment and integrated energy modeling.

SOURCES- Nature, Energy Skeptic, Youtube CGTN

Written By Brian Wang,

67 thoughts on “Nuclear Energy is 50% Better than Solar for Lifetime CO2 Emissions”

  1. I want nuclear power plants on the moon–the power going to make solar farms to make even more power–and so on. Both/And

  2. Current nuclear remains very inefficient leading to large amounts of waste. Next Generation, Gen 4 nuclear fission reactors are highly resistent to meltdowns. Options other than high pressure water and solid fuel rods with enriched uranium. Many Generation 4 designs exist. They can not be developed in the US. China will likely build US designs. This may be a vital option if other low carbon energy fails. We need more options. Currently prototypes are needed to test Gen 4 technology. Companies like Terrapower need to go elsewhere to develop this technology. Gen 4 reactors have other advantages. More efficient and producing less waste. Also other fuels apart from expensive and inefficient enriched Uranium. Renewable technologies are much more complex than most people think. China and India will dominate Gen 4 technology.

  3. All electric devices, from toasters to nuclear plants, are quoted at name-plate capacity, because they have to connect to house wiring or transmission lines. The connected wires have to have the same capacity for safety reasons. So everyone in the industry uses rated power.

    There is a “capacity factor” multiplier, in percent, which is the average power over the course of a year divided by the name plate capacity. The 25% and 90% numbers you quoted are capacity factors.

    An electric grid has to match supply and demand every minute of every day, but demand varies by time of day and season. Electrons are fungible. The consumer doesn’t care where their electrons came from, only that they show up when they want it.

    The job of individual utilities and a grid as a whole is to match supply and demand at the lowest average cost. No power plant runs 100% of the time, so they meed demand by having multiple power plants of various kinds on the grid.

    It doesn’t matter that solar only works 25% of the time. It only matters that it delivers that 25% at a low enough price that utilities will buy it when it is available. Since solar doesn’t consume fuel, it can usually deliver at a low enough price.

    California has the 7th highest electric rates in the US, and about 1/3 lower than Germany:

  4. The Department of Energy’s “electric power monthly” ( ) has the data to calculate that.

    Table 1.1 shows 2018 Utility + Small Scale generation as 4,207,500 GWh. Dividing by 365 x 24 yields average power of 480 GW.

    Table 6.2.A gives net summer capacity for Utility sources as 1082 and 1098 GW at the start and end of 2018, or 1090 GW average for 2018. To this we add 18.25 GW average small scale solar from the next table, for a total of 1108.25 GW average installed capacity.

    1108.25/480 = 2.309

  5. First of all Canada hasn’t licensed anything in a looooong time. They are in discussions with several dreamy startups. You know, the type of startup that has memorandums of understanding (MOUs) with real operators and equipment vendors in outreach and posts about it on LinkedIn. I was fed this same BS about processing oil sands with the mpower reactor that I dedicated 5 years of my life to. I distinctly remember reading exactly what I said about the Chinese Pebble beds and with a little bit of knowledge you too can understand why they will not be economical. my codes are in Fortran. My scripts are in Fortran wrapped in the shell. one of our core monitoring systems is on a Windows server and we interface/administer in SQL; it is a total freaking nightmare rube goldberg machine of layered virtual machines that craps out every week and requires the developer to patch it regularly. Mind you that the purpose of this windows server is to run a Fortran code every 30 minutes and read about 300 processed instrument inputs as real numbers every 15 seconds – something that could be done on a raspberriPI. I don’t have that issue on the Linux boxes on the PWRs, which I also maintain. Glad we’re both PEs; PE’s opinion is only valid in their field. This is my field.

  6. Yup lots of blowhard claims zero evidence. The real BS’er is you.

    But we should believe you and not the real experts at the CNSC that just gave a HTGR license. Have you called them yet?

    Those idiots at Terrestrial that have so impressed the experts at the CNSC have nothing on you. I guess those Tar Sands process heat fans that dumped a few hundred million into it, have shit fur brains. Shoulda called you first.

    Why would I believe because you never mentioned them and their 10 year service record . And yes the ARC Prism is a fast reactor SMR based on the IFR successful pilot just like Russia’s been running for a while and India is staking its nuke future on for service next year – real dimwits all of them eh.

    You seem unaware that China cancelled all nuke build for 3 years in their usual abundance of caution. We have no real idea how their HTGR is doing unless the CIA is briefing you.

    Actually I’m a PE EE with 40 years in the field but ya I can use SQL . Doesn’t seem a skill you have.

  7. You’re including the Russian BN reactors as gen4? Not sure they qualify; depends on who draws up the list. Why would you imagine that I am not aware of 40+ year old Russian lead cooled submarine reactors and their metallurgical difficulties? News has come out of China that the HTR-PM units are not going to be economically competitive and that they don’t plan to build more of them in the near future – it’s a demonstration plant. These pebble beds are NEVER going to replace coal and the fuel product is hella expensive and very low in uranium content because it is mostly graphite. For instance, there is 1/50th the heavy metal per volume in a graphite pebble then is seen in LWR; at 20% enrichment the pebbles have only 1/12 the energy content of LWR fuel. That is simple math based on 9g Uranium in a 200g pebble. Also, like I said before, nobody is going to use a nuclear plant to process oil sands or refineries or some wacky iodine/sulfur water cracking process… You’ll see electrical generation and perhaps a district heater demonstration in China and some desalination in the desert. I know you’re a big dreamer and you don’t like to hear the pragmatic angle, but believe me when I say that I’ve forgotten more than you will ever know WRT this particular subject, which is full of NBF BS-ers like ThorCon, Transatomic, Terrestrial, etc.. We’ll be reading about them for some time, no doubt a couple more years… by then you’ll be a real SQL expert writing scripts and stuff.

  8. You might want to look at the inflation in the cost of hydro. Site C has about 1 GWh of storage for $15B – no pumping needed. About 15000/kWh – now there is cheap storage eh.

  9. I notice you have severe illiteracy issues. Why not learn to read before posting?

    When you can do that or get a friend to help you can read your own idiocy.

    “No industry is asking for the process heat” like you never heard of natural gas. If nukes can heat provide cheaper they’ll take it. In fact Terrestrial’s seed money came from an entrepreneurs desire for process heat in the tar sands.

    While I don’t doubt your expertise at running BWR’s apparently the distinguished nuclear engineers at the CNSC just licensed a pebble bed SMR for process heat and power. But hell they are really dimwits compared to a BWR expert like you. Maybe you can make some bucks offering your expertise?

    Then of course you seem unaware for the two Russian Fast reactors that have been working for a fair while as well as the Alfa sub that was powered by lead based unit over its service life.

    Actually as part of your reading course you’d discover that China’s HGTR will be used for process heat after they get some experience running it as a coal plant heat source replacement.

    Not often you find a engineer Luddite on this site spewing from ignorance. You no doubt along with ULA were having a great laugh at Musk’s expense when his Falcon 1 kept crashing – just another asshole dreamer eh.

  10. “Duh – you never heard of process heat -hydrogen production?”

    Oh, you mean the 850C sulfur-iodine cycle? Oh, that sounds really practical in itself, let alone when nuclear powered… Sounds good on futurist web pages to people who write SQL in SunnyVale for a living.

    “HTGR can use gas turbines…”

    Fun fact: Did you know that the Germans built/retired exactly 1 50MW helium turbine in the mid ’70s (Oberhausen 2) and that was the only example ever? That is to say that the General Atomics concept of the month is another paper reactor with paper turbine; although it seems intuitive (helium turbine) it is actually quite a distinct machine compared to a jet engine or steam turbine, which we actually have experience building.

    “Gen IV nukes don’t work the same as your BWR. Americans blah, blah.

    Funny, I’ve been hearing about Gen4 reactors for 20 years now and the only thing that has been built is the soon to be commissioned HTR-PM units in China that will use helium and blowers to boil water… If I had focused on paper reactors instead of real reactors then I would have, well, quite a bit of paper on my hands instead of a house with a pool and nice furnishings, etc. and real experience. So, I traded idealistic things for pragmatic things, made a lot of money, and continue to follow nuclear developments like a hawk – mostly to call BS. Believe me, I would be on the bullet train to Gen4 if it were actually taking passengers. You mention BWR… hands-down THE BEST tech.

  11. $100 a kwh might be a magic threshold for electric cars, but pumped storage is already around that figure, and the amount being built is miniscule – in fact Germany has been closing plants because they can’t cover their costs.

  12. “You can mostly avoid traffic jams by not traveling during rush hour.”
    Most of us can not work from home. Most businesses are open during daylight hours, if there is a pile up on the freeway at 1 AM you experience considerable delays, even here, at the arse end of Africa. If you pee into a soda can in a car, you tend to spill.

  13. “That’s why the US has 2.3 times the installed capacity as average demand. The extra covers peak”
    2.3 times, as much as that? I never thought it would be that much, I was thinking around 25 to 50%

  14. $5/kWh would be added to your power bill to go 100% wind/solar with the cheapest batteries envisioned.

  15. Duh – you never heard of process heat – hydrogen production. No industry is asking because they use massively subsidized gas.

    HTGR’s can use gas turbines not necessarily steam plant.

    Gen IV nukes don’t work the same as your BWR. No doubt safety regs will be upgraded.

    Americans have no clue how to pour concrete anymore – can’t build any civil works without massive cost overruns.

    But they are pretty at building MSR sized things in factories at a world competitive price.

  16. When you learn grade 3 arithmetic tell us how much it will cost at that magic $100 to levelize wind/solar over the 16 weeks needed.


    A variable renewables only solution also requires long duration energy storage. In a review of literature, Jenkins finds a seasonal storage requirement for 8-16 weeks worth of US electricity consumption.

    I get $5/kWh added to your power bill.

  17. The 500 avg MW Site C and Muskrat Falls dams are costing about half as much as the $24B 2.0 GW avg FOAK Voglte but they are still building them. No problem for state actors to finance.

    And state actors are building them – just not in the USA where politicians are owned by Big Oil.

    Solar/wind coupled with enough battery to level out over a bad year would add $5/kWh to your power bill. Only reason to build it is corruption – Big Oil gets to sell lotsa gas as backup power.

  18. That’s true in the States, but most countries have peak power demand in winter, when solar is least. If you look at energy demand, rather than electricity demand, it’s highest in the US in winter too. If electricity was used for heating, instead of gas, the demand in a polar blast would be higher than in a heat wave. Ground source heat pumps could do both jobs, and in cities with apartment blocks, district heating from nuclear waste heat would be carbon free.

  19. Yeah , I was wondering how Nuscale handled that – was just listening to Reyes last night. The salt merchants all plan to use a tertiary nitrate salt, outside the nuclear site boundary, and with no safety case to make. That would let them load follow, or sell heat to other industries. Other industries which are quite happy to burn coal or gas now, because there are practically no downsides to them putting it all into the atmosphere – except for their children, maybe.

  20. Two humps, bigger one at the end of day. Maximum demand in summer months when sunset is at 9pm. And solar is good for about 50% of this peak demand. After that you start adding batteries which are getting cheaper every year. Solar competes with peakers which are expensive source of power.

  21. “Batteries would need to be about ten times cheaper than they are now to compete with fossil fuels.”

    Maybe you haven’t looked at prices lately. Bloomberg NEF reports that grid batteries are already down to $187/kWh, and $100 is the magic threshold. Gigawatt-hours worth of grid batteries will be sold in the next few years, and across the next 10-15 years potentially hundreds of GWh.

  22. A soda drink cup can serve as a urine collector in a pinch, and then you can dump it out the window. That costs nothing, compared to hundreds of thousands for a flying car.

    You can mostly avoid traffic jams by not traveling during rush hour.

  23. …demand plus a reserve to cover plants down for maintenance or not producing. They are not going to change that plan of having extra power plants, only the mix of *which* plants they are using.

  24. Electric devices, from solar farms to toasters, are universally measured by peak power. That’s because the connecting wires, either transmission lines or house wiring, have to have matching capacity for safe operation.

    “Capacity factor” is the average annual output divided by rated capacity, as a percentage. For US coal and Combined Cycle Gas, the two largest electric sources, they are 54 and 57%. For Hydro, Wind, and Solar they are 43, 37.5, and 26% respectively.

    Electric power grids always use a mix of power sources, because demand varies by time of day and season. Some sources, like Nuclear, run nearly all the time (92-93% in the US). But even they have to shut down sometimes for maintenance and refueling. When they do, some other plants have to pick up the slack.

    Solar doesn’t have to run continuously to be useful. In the US Southwest, peak demand is on hot summer days. That’s exactly the time solar is working. It replaces gas peaker plants, which are much more expensive per kWh. There’s a secondary peak in the early evening, when people get home. The Sun isn’t shining then, but with a few hours of battery storage, you can save up the sunlight from earlier in the day.

    > it may not be wise to risk large populations to extrapolations.

    Utility operators are pretty conservative. They have to match demand every minute of every day with the same amount of supply. That’s why the US has 2.3 times the installed capacity as average demand. The extra covers peak

  25. The ASME pressure vessel code allow you to make pressure vessels out of rolled plate too; that is Holtec’s angle.

    MSR still has a steam plant and that is what is such a PITA to operate. No industry is asking for the process heat and no industry wants the responsibility of hardening their industrial process to the point where it is basically safety related so that it doesn’t trip the nuke, which would be a “reportable event” or “operational occurrence”.

  26. My point was that we could have had flying cars at any time in the last 40 years, but we have not built them and they are not the future because they are unsafe to operate.

  27. The high wind belt running from Texas up to the Dakotas, where renewables boosters want to put terawatts of wind turbines, is also the migration route of the whooping crane, the tallest bird in North America, and one of the rarest.

  28. The 15 reactors China is building is only a small portion of the generation they need. Most of the rest is been satisfied by coal power plants.

  29. Have you ever been caught up in a traffic jam for several hours with a full bladder? Oh if you could just fly over it and p***on the lot?

  30. I think it is somewhat underhand to compare name plate capacities of solar PV to nuclear. You need 4 nameplate gigawatt of solar to compare to 1 nameplate gigawatt of nuclear to produce the same TWh output over a year. That presumes that solar will provide 20 to 25% of the rated power and nuclear (in the US) 90%. You have not mentioned cost of storage, alternatively back up, either. If wind and solar were really that cheap, why do the countries with the highest penetration of said technologies have the highest power prices? Denmark – Germany – California (not a country, but bigger than Denmark anyway)

  31. ‘Solar is a much less risky play.’ Tell that to all the people in Spain, for example, who lost their shirts on solar investments. That was after the government rescinded overly generous feed-in tariffs, that were bankrupting the power sector for minimal extra value. Must-take provisions that current solar developers rely on will also be vulnerable to a glut of power when its not needed, combined with the need to keep fossil plants funded for the dark hours. Batteries would need to be about ten times cheaper than they are now to compete with fossil fuels. For all the talk about them on renewables forums, the actual numbers being built are far below what would be needed to oust coal and gas.

  32. Depending on the energy used in enrichment I suppose they could both be true. But, yeah, Brian desperately needs a tech savvy proof reader.

  33. They don’t need forged pressure vessels, which means you don’t have to join the queue at one of the few forges capable of handling ~500 ton ingots ( none in USA ) . And they run 200 degrees C hotter than a light water reactor, which opens a lot of niches for process heat.

  34. Concentrating solar – lenses or mirrors – doesn’t work for diffuse light. Photovoltaic still puts out some percentage of the value it gets from full sunlight, but solar thermal just dies. That’s why they only build it in pretty much desert areas. Even Ivanpah, in the Mojave desert, was reported in its first year of operation to be producing “..about half of its expected annual output…. The California Energy Commission issued a statement blaming this on “clouds, jet contrails and weather”.( Wiki)

  35. Peak demand is in the evening usually, not at midday. Solar plants are competing with each other – the more you build, the closer to zero the additional extra value of each unit is. ( That doesn’t apply to solar thermal, but that really only works in desert areas, and is much more expensive than PV even there.)

  36. The peer-reviewed papers show that other than a few bad cherry-picked examples, there have been no real price rises in nuclear. Expensive nuclear power is a particularly *American* problem due to their unique regulatory framework that cripples nuclear. Most countries are building far cheaper than they are.

    We find that trends in costs have varied significantly in magnitude and in structure by era, country, and experience. In contrast to the rapid cost escalation that characterized nuclear construction in the United States, we find evidence of much milder cost escalation in many countries, including absolute cost declines in some countries and specific eras. Our new findings suggest that there is no inherent cost escalation trend associated with nuclear technology.

    That’s not all. New nukes without high pressure reactor cores can be built on an assembly line and mass produced, and many think new assembly line reactors will come off the line cheaper than coal, as Brian has presented here many times.

  37. There seems to be a contradiction in the post.
    It is stated that the embodied energy in a nuke is a larger fraction of the lifetime output than for a wind generator. It is also stated that a lot more material is used in wind & solar plants for a given energy generation than for nuclear.
    I don’t see how both can be true.

  38. I assume they take into account repowering the solar and wind plants? Up-rating the turbines and solar in 20-60 years with minimal new structures, and adding battery back-up over time as it becomes cheaper will probably change the analysis.

  39. Agree on the financing. But… are the solar plants power ratings you list capacity factors? Is their power firm? If capacity factors only, then that is their peak potential output during peak sun, and actual full year capability is about 10-25% of that if their is a grid that can take it, or you have enough batteries to act as a buffer. If the solar parks are providing that power continuously, in a firm manner (backed by natural gas plants, thermal storage, battery), then they are truly impressive.

    My 2 cents, either way, if current trends hold on battery, solar, and wind price decline trends, then solar or wind will start to dominate in some regions in about 5-20 years. There are already some edge cases, such as islands where they can dominate already. Caution is warranted though, as no exponential lasts forever. Also, it may not be wise to risk large populations to extrapolations.

  40. As ive said before. Fresnel lenses to run closed loop steam powered generated electricity. Forever. FOREVER.

  41. +30k people die on roads every year because it’s too inconvenient to do anything about it. You actually think 3-4 times more deaths per TWH means anything to anyone aside from their families. Every death is a tragedy, you send the perfunctory thoughts and prayers and get on with business.

  42. It does not matter that nuclear energy is 50% better than solar for lifetime CO2 emissions. Solar is way better than coal, oil & gas for lifetime CO2 emissions, most importantly, you can actually get solar plants financed. Good luck trying to get financing for nuclear.

    Watts Bar Unit 2 1165MW Cost $6.1 billion
    Solar Star, CA 579MW +2 billion
    Benban solar park, Egypt 1.8GW $4 billion
    Villanueva solar , Mexico 828MW $650 Million

    No one is stopping anyone from building nuclear plants, it’s just no one wants to risk a large chunk of cash and infinite liability. Solar is a much less risky play.

  43. So we should compare today’s nuclear to the renewables of ten years from now? Doesn’t that sorta put a thumb on the scales?

  44. Next Gen 3 & 4 reactors are designed to be ‘walk away safe’.

    As in, once they are up and running and say, a zombie apocalypse wipes out the staff, the plant will keep on a runnin’ for long while. If things get unstable, it will shut down/scram all on its own because of how it is designed. Or, it CAN’T have problems simply because of how it is designed (i.e. Pebble Bed Reactors).

  45. What will solar/wind, with battery storage, cost by then?

    As in what? Your unproven assumption that they will cost less than today and way less than nuclear?

    Or how about this: It doesn’t matter until it does drop in price. Everything else IF/WHEN THEN is just pure anti-nuclear masturbation on an internet forum.

    Nuclear can be made economically. The French proved this in the 1980s.

  46. but from NIBMY POV you don’t care if some poor soul (or dumbass) dies while destroying the plant , while with nuclear you have the specter of leaks even if no one dies. Your property valuation is at risk.

  47. Sure, coal and gas are a lot worse, but 3-4 times worse is still nothing to sneer at; You want to save lives, you replace coal with nuclear, not windmills.

  48. I know it doesn’t sound like flying cars to you, but you just said the equivalent of: “we are all going to be driving flying cars in 10 years.”

    Like flying cars, the MSR is also of questionable merit, creating new problems while solving not one I can put a finger on.

  49. A severe disadvantage is that a lot more people die per TWH. 3-4 times as many for solar and wind.

  50. It’s looking like in ten years the first commercial molten salt reactors will be coming on line, with much better economics than conventional nuclear. They’d be fast to build, too.

  51. The U.S. is pretty backwards about it but China has 15 reactors under construction and plans to start more soon.

  52. One advantage solar and wind has over nuclear is that once it is built it can be operated less responsible people without much of an oversight. Also if something goes wrong it is easy to be fixed. (people still can die from electrocution, falling from the towers, etc)

    NIMBYs can complain about noise or eye sore but that’s all.

  53. Half of all fossil power plants are peaked which only run for part of the day. Solar power compete against fossil peakers and not against fossil baseload and so the analysis should be against peakers.

Comments are closed.