A few more observations on nuclear power

I thought I should respond to the latest suggestions from Department of Industry and others that nuclear power is an option worth considering for Australia. While I’m at it, I’ll add some updates on global developments.

* The most striking feature of recent Australian discussion, beginning with the Australian Energy Technology Assessment from 2011 is the claim that “small modular reactors” represent an appealing option for Australia. AETA listed these as being one of the cheapest options for 2020. with an estimated levelised cost of between $75 and $125/MWh. That’s both ambitious and remarkably precise for a technology that does not yet exist, even in prototype form. Leaving aside niche technologies like the Russian proposal to adapt nuclear sub reactors as floating platforms, the only serious contender in this field is the US, where the Department of Energy has provided grants for the development of two pilot plants. The target date (almost certainly over-optimistic) for these to begin operation is 2022. To get any idea of economic feasibility, it would be necessary both to undertake commercial deployment (in the US, obviously) and to to accumulate some years of operating experience. To get this done by 2030, or even 2035 would be an ambitious goal, to put it mildly. Again assuming everything goes well, Australia might be in a position to undertake deployment of SMRs by, say, 2040. But obviously, if we are going to reduce emissions on anything like the scale we need (80 per cent by 2050), we need to phase out most fossil fuel electricity well before that. Obviously, all these points apply in spades to proposals that exist only as designs, with no active proposals even for prototype development, such as the Integral Fusion Reactor. As I’ve argued before, to the extent that nuclear power makes any contribution to reducing CO2 emissions on a relevant time scale, it will have to be with current technology, most likely the AP1000.

* Talking of the AP1000, the builders four plants under construction at two sites in the US have just announced another 6 months delay, pushing the operations date out to 2017 or 2018 (release from FoE, but links to originals)

* Most interesting of all are projections released by the International Atomic Energy Agency last year for the period to 2050. Currently nuclear power accounts for around 11 per cent of global electricity. The IAEA “low’ projection has that falling to 10 per cent by 2030 and 5 per cent by 2050. The “high” projection, which includes steady growth in both North America and Western Europe as well as spectacular growth in Asia, has the share remaining roughly stable. So, even on the most optimistic projections of the world’s leading nuclear agency, nuclear power won’t play any significant role in decarbonising the electricity sector, let alone the economy as a whole.

I’ve come to the conclusion that nuclear power advocates, like climate delusionists (virtually all climate delusionists are nuclear fans, though not vice versa) are essentially immune to empirical evidence. So, I’d prefer no comments from our usual advocates (hermit, Will B etc) unless they have something genuinely new to say.

300 thoughts on “A few more observations on nuclear power

  1. John, I think you’re unduly pessimistic about the contribution nuclear power will make to carbon abatement. Even at the growth rates in the IAEA forecast you cited, it’s quite likely that nuclear will still be producing more low-carbon electricity than wind and solar will 30 years from now.

    The low-case IAEA estimate for nuclear generation in 2030 is 3426 terrawatt-hours, rising a bit to 3548 TWh in 2050; the high-case estimates are 5689 TWh in 2030 rising to 8971 TWh in 2050. But how about wind and solar? The data Quokka cited upthread (p. 3 # 6) from the IEA (different acronym!) for 2035 are 2251 TWh for wind and 802 TWh for solar PV and CSP. Putting those two estimates together implies that in 2035 nuclear will still be producing more low-carbon electricity than all forms of wind and solar combined.

    Wind and solar combined would likely outstrip the low-case IAEA nuclear estimate by 2050, but perhaps not the high-case IAEA estimate of 8971 TWh. So it’s quite possible that in 2050 nuclear will still be producing more low-carbon electricity than wind or solar separately, and perhaps even combined. Obviously there are great uncertainties in these estimates, but your conclusion that “nuclear power won’t play any significant role in decarbonising the electricity sector” seems unwarranted, unless you make the same conclusion about wind and solar as well. And these forecasts are all heavily dependent on political constraints, which can change. Green politics should fight for all forms of clean energy, nuclear as much as renewables.

    You also write on p. 1 # 17: “Solar PV installations are predicted to rise to 50 GW this year…(even allowing for lower availability) much more than we are going to see from nuclear any time soon, even disregarding closures. Wind is…likely to be around 40 GW.” Again, this is probably too pessimistic an assessment of the nuclear buildout. Nuclear capacity this year will probably grow faster in absolute (though not relative) terms than solar and possibly wind capacity, once you factor in the much greater productivity of nuclear gigawatts compared to solar and wind gigawatts.

    The World Nuclear Association anticipates about 15 gigawatts of new nuclear coming on line in 2014 (not counting the new Japanese reactor that definitely will not). (http://www.world-nuclear.org/info/Current-and-Future-Generation/Plans-For-New-Reactors-Worldwide/) Let’s assume that delays cut that to 10 GW. With an average global capacity factor of 85 percent, those 10 GW nuclear will have an “effective” generating capacity of 8.5 GW on average. Compare that with 50 GW of new solar capacity, with a global average capacity factor of 15 percent for an “effective” generating capacity of 7.5 GW. The new nuclear capacity will thus likely produce more low-carbon electricity than all the new solar capacity. 40 GW of new wind at a global average 25 percent capacity factor would have an effective generating capacity of 10 GW. If 12 GW of WNA’s forecast 15 GW of new nuclear get commissioned this year, then the year’s new nuclear capacity will generate more electricity than the new wind capacity.

    So new nuclear construction is ramping up and keeping well apace of new solar and wind capacity, once you factor in nuclear’s greater productivity. (And providing a much higher quality of power, because it is reliable and dispatchable.) It’s a mistake to disparage the contribution of nuclear to carbon abatement when it is clearly doing at least as much as wind or solar—and will continue to do so for the foreseeable future.

  2. Ronald Brak :
    … Nuclear power simply will not be built in Australia because it simply cannot compete with other generating capacity no matter what the carbon price is. There is simply no way it will not lose money. Australia’s average wholesale electricity price is around 5.6 cents a kilowatt-hour and it is not possible to build a nuclear plant that can produce electricity for less than that. Hinkely C in the UK has a minimum wholesale price of 15 cents a kilowatt-hour and that doesn’t include insurance. Even for a modern nuclear plant the cost of insurance alone could be much more than 5.6 cents a kilowatt-hour.

    I know you are not doing this deliberately, but this cost problem is simply only the case if and only if the various costs also fold in the additional burdens that have been (deliberately?) thrown on the nuclear industry. It is not an essential feature of nuclear power as such. I’ve mentioned CANDU reactors elsewhere, mainly because they are a proven technology; if they were run in certain special locations that wouldn’t suffer if there were ever any adverse consequences, reactors of that sort could be set up with much smaller costs. I only present this to show the engineering feasibility, since actually running them there would almost certainly not be worth it as the energy would be “stranded” – but submersible barge units could readily be built and maintained like that, i.e. at much lower cost, particularly if there were many units serviced there that could share overheads, while still being operated as sealed units at the more convenient sites. And that’s even before looking into other promising designs that are not yet proven (I personally like the possibilities of a fluidised sugar charcoal suspension homogeneous reactor).

  3. I don’t want to discuss nuclear power in a thread on GM food, so I’ll bring up a point here on continuing to use existing nuclear power plants. I’m very confident that if nuclear power plants had to pay the full cost of insurance to cover nuclear disasters then every single one would shut down in short order. In a purely “dollars and cents” evaluation it makes no sense to keep reactors operating in Europe, Japan, and I would think probably every country in the world. However, the moral case is different. I’m not saying the moral case is separtate from the economic case, but it does hinge upon the value one puts upon human life. Nuclear disasters can be incredibly costly, but outside of a disaster stricken nuclear plant radiation released is likely to kill slowly and so for the most part people can probably be safely evactuated from contaminated areas. So shutting down nuclear powerplants for for the most part protects property in rich countries and not lives. But carbon dioxide released into the atmosphere kills people through altering the climate and most of the people killed are in poor countries. So if one values human life equally regardless of wealth or origin then I would say it is probably better to keep nuclear plants that meet high standards of safety operating than to shut them down if that would result in higher greenhouse gas emissions. If the world suddenly became a fair and reasonable place I’m sure we would rapidly decarbonize electricity production to save lives and then rapidly shut down nuclear plants and replace them with other low emission capacity. But if it is not possible to do both these things then it would be better to decarbonise the electricity sector and keep the safer nuclear plants operating.

  4. @Ronald Brak

    I’d like to see if we are both on the same page or if you used a different process to arrive at your conclusion

    I used the same process as you which you would have realized if you had paid attention: https://johnquiggin.com/2014/01/18/a-few-more-observations-on-nuclear-power/comment-page-3/#comment-219779 . I’ll quote the comment here just for you, but I can’t be bothered reformatting it just for your benefit:

    “An Australian with rootop solar might pay a marginal cost of 30 cents per kilowatt-hour for grid electricity

    That’s pretty expensive. Mine this year will cost 22 cents per kWh.

    This means the batteries will save 44 cents per day or $160 a year and pay for themselves in 7.5 years.

    Or in my case, 28 cents per day or $102 a year and pay for themselves in 11.74 years, i.e. long after the lithium-ion batteries have failed.”

    But this particular hypothetical certainly seems to have got your blood up.

    Maybe you could stop tediously and pointlessly badgering me with it.

  5. @Ronald Brak

    total of over 2 gigawatts

    Gee, 2 GW out of 30 GW. Don’t hold your breath waiting for the rest.

    By the way, how much has the actual energy generation and Carbon emissions from coal generators declined?

  6. Are you okay, Chris? You seem bitter. Is there anything bothering you in your personal life that you’d like to get off your chest? If you need to talk, there are people here for you.

  7. @Hermit

    When Hazelwood and Bayswater have rusted away, all the gas has been sold to Asia and when only intermittents are allowed things will be crook. I imagine a still frosty night in Melbourne with people burning the floorboards for warmth.

    That solar electricity won’t look too cheap at 8.30 in the evening during the domestic peak. On the other hand, since there isn’t any, it’s cost will be zero. Just a pity that everything will be off.

  8. @Ronald Brak

    Ronald, you continue to ignore virtually all of the flaws I have pointed out in your arguments. To reiterate, they are:

    1. Your peak domestic electricity price is too high. A lot of people, e.g. everyone in Melbourne, can get it for substantially less than 30c.

    2. You assume Li-ion batteries last for the equivalent of 3,652 full cycles. They don’t.

    3. You completely ignore lost opportunity cost of capital employed.

    4. You assume that charge/discharge cycles are 100% efficient. They are not.

    5. You assume that Li-ion battery costs are going to rapidly decrease when the rapid decrease occurred years ago and the decrease has slowed right down.

    Is there any reason why I shouldn’t just consider you to be nothing more than a battery pollyanna?

  9. @ Ronald Brak
    I think estimated prices for future windpower should include the LGC subsidy of 3-4c per kwh omitted from the AETA study. That subsidy along with transmission, retail margins, GST etc goes into the retail price but is specific to commercial wind and solar. Nuclear electricity if it ever happens won’t get that particular subsidy.

    On insurance I think a government indemnity after the first few billion or so in liability would be standard practice as per the US Price Anderson Act. The WA and federal govts have indemnified Chevron in case any of the 120 Mt of CO2 to be injected under Barrow Island WA leaks out. If it all erupted due to an earthquake at the current price the carbon tax alone would be 120 Mt X $24.15/t = $2.9 bn. Chevron won’t pay a cent.

  10. @Will Boisvert
    Will: your numbers only work for the next couple of years. On the 10-year timescale for new nuclear plants like Hinkley Point, the relevant comparison is with solar and wind a decade out. Trade estimates for global solar are around 45 GW this year, up from 35 GW in 2013. By 2023, global PV installations will be running at >300 GW a year just projecting this historically typical one-year growth rate. This would on your parameters be equivalent to an annual addition of 44 GW continuous capacity or 52 GW of nuclear, which is clearly not on the cards. The scenario is just a thought experiment, but I challenge you to propose a historically plausible 10-year scenario in which solar doesn’t strongly dominate nuclear.

    Your 25% capacity factor for wind corresponds to the current US average. It’s much too low for new turbines, which are not only bigger and taller but have been tuned for higher capacity factors at the expense of peak output. 40% is not unusual.

  11. An article in theenergycollective by Gail Tverberg makes the point that the emissions reduction from wind and solar is not as great as hoped. To paraphrase the key reasons are suboptimal operation of the backup power source and the emissions embodied in the construction and replacement of wind and solar. If a renewables quota obliges nonrenewables to throttle back at times there is not only duplicated capacity but poor use of fuel from stop-start operation. An example I’d give is running a combined cycle gas plant as a single cycle. Wind and solar also take silicon, rare earths, steel and cement which if sourced from China were probably made using a lot of coal. Plant lifetimes for wind and PV solar may be 25 years only half that of large thermal plant, therefore need more frequent renewal. Tverberg also points out that so far wind and solar have barely had any effect on the looming transport energy problem.

  12. @ James Wimberley,

    Do you have a source for your estimate of 300 GW of solar installations per year in 2023? I’m a bit skeptical of that. If you’re just assuming that the current yearly increment will compound forever, that’s doubtful. The difficulties of ramping up production to that level will be extreme, integrating that much solar into grids will be a nightmare; there are just a whole host of reasons to doubt such numbers, especially since solar subsidies in Europe are now being cut with the express purpose of slowing deployment.

    As for this decade, there’s little question that new nuclear will outstrip new solar. Currently there are 75 GW of nuclear under construction and due on line by 2020, the equivalent of 425 GW of solar. China alone will likely break ground on a further dozen gigs this year, and with its standard 5-year build times bring all of them online as well by 2020. Mass production really advantages nuclear.

    In fact, a deployment of 52 GW of new nuclear per year is not as outlandish as it sounds. In the 1980s the world deployed about 20 GW new nuclear per year. Given that the world is now richer, more industrialized, technologically sophisticated and motivated to build clean energy than it was in the 1980s, I don’t see why we couldn’t build nukes at many times the 1980s rate–provided that governments make a serious committment to doing so, as they have with renewables. Politics is the real barrier, not logistics.

    The 25 percent capacity factor is the global average for wind, the current US average is about 30 percent, and has been stuck there for several years. No national wind fleet gets 40 percent on average. It’s actually likely that global capacity factors for onshore wind will stagnate or decline in the future. New taller turbines are better at harvesting winds, but that has been partially offset by mass deployments resulting in sites of poorer and poorer wind quality being developed. Because geographic diversification will be an important strategy in coping with wind’s local unreliability, more turbines will be sited in regions of weak winds–sometimes wind power in Alabama will fill in for slumps in Nebraska, but the Alabama turbines will have dreadful CFs. And there’s the problem of diminishing returns; at high penetrations the grid will be unable to absorb all the wind power during peak generating periods, so much of that generation will be curtailed, further tanking CFs. We can expect offshore wind to be more productive than onshore, but it’s also much more expensive, so no cost gains there.

  13. Hermit, Large scale Generator Certificates do not increase the cost of building windfarms. If I built a wind farm without LGCs and then suddenly decided, “Wait a minute, I actually want them,” then once I obtained them it wouldn’t suddenly increase the price I paid to build the wind farm. How could it? Time travel?

    Hermit, do you believe that because TEPCO did not have insurance to cover the cost of the Fukushima nuclear disaster that disaster didn’t cost anything? Because there was no insurance to cover it the costs just disappeared? Because if you do that’s just plain nuts.

  14. @Ronald Brak
    I’m saying LGCs increase the retail price of wind powered electricity in a way that other forms of generation don’t. Therefore it’s disingenuous to exclude the LGC from the levelised cost. Interestingly pre-1997 hydro doesn’t get LGCs, perhaps a tacit admission that it’s reliable.

    I don’t what the insurance situation was with Fukushima. Some believe the evacuation zone was four times larger than really necessary. Surely the role of the Japanese government was to conduct an orderly response. Instead they stage managed fear and loathing like a kabuki play. Remember the mortality figures; missing or drowned 20,000, serious radiation doses 0. It affects us because since then the Japanese have bought a third of the world’s LNG, a price we will have to match from next year with our east coast gas exports. Japan like Germany and California has increased emissions since prematurely retiring nuclear plant.

  15. Indian 220MW reactors are the most cost effective small reactors in the world. One is likely to cost 500 million dollars or so. With supply of uranium from Australia to India now being negotiated, good business synergy is likely to develop.

  16. Chris, you say the electricity price I used in my estimate was too high. But I wasn’t trying to see if the Chris O’Neils of the world paying a marginal cost of 22 cents per kilowatt-hour for grid electricity would save money from home energy storage, I was trying to see if people who pay a marginal cost of 30 cents or more a kilowatt-hour, as many people do, would save money. So there’s not a problem here.

    I went out and found some more information on the Sunny Boy SMA and with a suitable rooftop solar system it is expected to have a storage throughput of over 8,200 kilowatt-hours in its lifespan of ten years, and a storage system efficiency of over 90%. So using these figures and your figure of $600 per kilowatt-hour of storage and assuming an eight cent a kilowatt-hour feed in tariff and that one fifth of the storage system loss occuring from solar to storage and four fifths from storage to AC gives a price of about 28 cents per kilowatt-hour of grid electricty consumption avoided using a 5% discount rate. Or to put it another way, a person paying a marginal cost of 30 cents a kilowatt-hour for grid electricity would make over a 5% return on their investment. In fact the return would be about 7%. This is very close to my original estimate.

    I assume you are okay to do the calculation yourself and confirm that I am correct.

    It is reasonable to expect a manufacturer to use the best figures they can for a product so the performance of the average installation may not be as good as they suggest, but it does seem likely that for people paying a marginal cost of 30 cents a kilowatt-hour, with a feed in tariff of 8 cents a kilowatt-hour, and a discount rate of 5%, home energy storage at $600 a kilowatt-hour is around the break even point. Of course, as I mentioned earlier, I expect home energy storage to be available in Australia for under $600 a kilowatt-hour before long.

  17. Hermit, pre 1997 hydroelectricity did not recieve LGCs becaue the purpose of LGCs is to encourage the building of new renewable energy generating capacity. Existing hydroelectric capacity didn’t need encouragement to be built because it was already built. I’ve told you this before. You must have forgotten.

  18. @Ronald Brak

    But I wasn’t trying to see if the

    people of Melbourne and other places

    paying a marginal cost of 22 cents per kilowatt-hour for grid electricity would save money from home energy storage

    Good, your hypothetical example is totally irrelevant to the people of Melbourne and probably lots of other people. At least we agree on that.

    I went out and found some more information on the Sunny Boy SMA and with a suitable rooftop solar system it is expected to have a storage throughput of over 8,200 kilowatt-hours in its lifespan of ten years

    Of course that’s more than 2 kWh of battery if it’s coming through in ten years. Why don’t you give us a citation so we can find out what the facts are?

    assuming an eight cent a kilowatt-hour feed in tariff and that one fifth of the storage system loss occuring from solar to storage and four fifths from storage to AC gives a price of about 28 cents per kilowatt-hour of grid electricty consumption avoided

    OK, so you’re moving the goal posts again. You haven’t explained in detail how you came up with your 28c.

    Also, just because there are Li-ion batteries available for $600/kW doesn’t mean the Sunny Boy batteries are that cheap.

    Let’s see how the deficiencies in your argument now stand:

    1. Your peak domestic electricity price may be relevant to a few people. It’s irrelevant to lot of people, e.g. everyone in Melbourne. Probably everyone in Sydney and Brisbane too I guess.

    2. You haven’t cited Li-ion batteries that last for the equivalent of 3,652 full cycles. 8,200 kWh might be the Sunny Boy 5.5 kWh capacity system for all we know.

    3. You completely ignore lost opportunity cost of capital employed. I haven’t seen any indication that you understand what this means.

    4. I don’t know how you incorporate <100% efficiency in your calculations. But I'll agree that this would only make about 1c/kWh difference when the input cost is 8c/kWh.

    5. You assumed that Li-ion battery costs are going to rapidly decrease when the rapid decrease occurred years ago and the decrease has slowed right down. You assume an expected battery cost of $600/kWh soon without giving any citations but imply they'll last 3,652 full cycles equivalent.

    You're still a battery pollyanna.

  19. @Chris O’Neill
    It looks like you (and possibly Ronald?) are not factoring in the daily charge for electricity. If you are talking about using batteries to go off the grid, that should be added in.

  20. The latest news on Fukishima is not good. I leave people to google it. You will find as many or more disturbing news items as I did. The bottom line is that nuclear power is too dangerous and too damaging for humans to handle. The empirical track record shows this.

    In the era of industrial decay we have now entered, the dangers of nuclear power stations exploding or melting down will only increase. It will not be possible to safely build, maintain or properly decommision nuclear reactors.

    I would argue that we have entered the era of industrial decay. Peak energy and then decline of energy available as exergy (energy for useful work) from all sources will mean that the trend to greater disorder will predominate over the process of ordering (building and maintenance of complex structures and systems) which requires energy.

    Reaching peak energy is not really about reaching peak volumes of oil, gas and coal production (the main fuels in the world energy mix). There are two other factors involved. These are the decline in the quality (energy density and purity) of the fuel and the rise in the difficulty (energy expenditure) of obtaining the fuel. Thus peak production volume will not coincide with peak energy production. Peak energy production will precede peak volume production, possibly by as much as a decade. Simply counting barrels of oil, cubic feet of gas and tons of coal does not enable calculation of peak energy from fossil fuels.

    Reaching and passing peak energy from fossil fuels is a physical certainty. The first question is whether other technologies can be deployed fast enough to increase energy supply or at least stabilise it at a plateau sufficient for a modern world of say 8 billion or 9 billion people if we can stabilise population at that level. The second question is scalability. Can the solutions scale up without reaching material resource limits of their own? The third question is transition costs in energy terms. Is there enough dirty fossil fuel energy left to power the transition? Remember, getting a net energy return from renewables, solar and wind, takes a long time, typically at least 3 to 5 years. During a relatively rapid ramp-up transition period, say 20 years, we could be in constant “energy return deficit”. That is the energy investment in ramping up renewables fast enough would exceed energy returns until the system matured and reaced a critical production mass. In the meantime, while ramping up we would have to energy fund both the ramp-up and all other ongoing energy needs. It would not be simple or easy and it is by no means clear that it is ultimately feasible.

  21. @Jagdish
    Such is US ‘soft power’ I suspect the only reactors that would be considered are those approved by the US Nuclear Regulatory Commission. The US is saying they won’t have an SMR ready for export until 2022. By that time Australia will have burned or exported another 2-3 bn tonnes of coal. If a gas price shock comes early say 2015 coupled with extreme weather we will be looking for low carbon dispatchable power much sooner than 2022. Perhaps small reactor developments in India, Russia and China will force the US to get a move on.

    According to the WNA article on Finland that country (also a buyer of Australian uranium) has ordered a Russian reactor. I wonder if that was influenced by delays and cost overruns with the French designed reactor still offline years late. That shows the US and France can’t assume the West will automatically prefer their models.

  22. @Chris O’Neill #19:

    I live in Melbourne and pay 39.16c/kwh for peak (7am to 11pm). Anyone who has solar will be offered peak/off-peak. I don’t understand why you think 30c is irrelevant to the people of Melbourne. That is the lowest peak price you could possibly find.

  23. Ronald.
    Just because you call me lair does not make me one. If you ignore the efficiency of the generator you used to generate the power in first place you get the higher efficiencies. We are talking about generating power and then storing it verses generating it online. Every time you change the state of the energy you lose something. Every time you ship power any distance you loose something. …of course you do need water to do pumped hydro storage. Last I heard Australia is short on water. I should think evaporation alone would be a significant loss factor.

    As for costs of nuclear power . here is an URL
    from this site:
    Projected Levelized

    Cost of Power (2014-2043):
    4.7 – 5.2 cents/kWh
    Comparison Costs*:
    Natural Gas: 6 – 14 cents/kWh
    Wind: 7 – 10 cents/kWh
    Solar: 11 – 42 cents/kWh

    Now the cost you pay can be higher or lower depending on if your government will tax you or subsidize you on the energy generated.

    Nuclear power insurance costs are not subsidized. That is a misunderstanding of how it works. True they limit the liability but the company gives up a lot of their rights. its more like getting an insurance policy and then letting the insurance carrier ride with you. Then allowing him to fine you every time you drift over the speed limit.

    Now if the government subsidizes any energy source …does the costs somehow change? no… it just means someone else is paying the part of your bill.

    Wind and solar price does not include the cost of storage so you will be limited to how much the grid can handle. Unless you can live with you factory power going up and down with the wind.

  24. Hermit, Right now the US NRC can not approve a hole in the ground called yucca mountain that they have been designing for 30 years. So, do not look for the designs to come out of the US unless you see some political changes.

    The French had cost overruns in Finland but their reactor is more complicated. Look to the south Koreans. They seem to be big payers in nuclear power.

  25. Chris O’Neill, I apologise for being so snarky. I used a simple levelized cost of energy calculator to work out the cost of storage. If you tell me what method you used to determine that energy storage at under $600 a kilowatt-hour is a waste of money I will use it to demonstrate that I am wrong and apologise profusely.

  26. I don’t want to bring up nuclear power in the GM food comment thread, so I will respond to Wills Boisvert’s comment on nuclear power being cheaper than renewables in China. I particularly want to reply as John says he is going to write on this topic.

    Will, wind power is much cheaper than nuclear power in China. China can build nuclear power more cheaply than richer countries but it can also build renewable capacity more cheaply. Wind power in China now appears to be about $670 per kilowatt:


    The 22% capacity factor you give for wind power in China appears to include wind curtailment due to inadequate transmission which currently results in about 10% of Chinese wind farm production being forgone. Currently electricity equal to about 23% of wind power’s total capacity is utilised. But using the 22% figure you gave for wind and the $2500-$3500 per kilowatt price you give for nuclear, wind is still cheaper than nuclear power in China and this is the case even before the cost of insurance is included. Just how much cheaper depends on estimates of various expenses and, Will, please feel free to provide us with estimates of these expenses over the lifespan of the Chinese reactors. All else equal, in absolute terms it would be cheaper to provide insurance in China than in a richer country, but it will still be expensive and China is rapidly growing richer and is on its way to becoming a developed country. Now it is possible, Will, that you think the appropriate thing for the Chinese government and Chinese citizens to do in the event of a Fukushima type nuclear disaster is nothing, apart from perhaps avoid licking any burning chunks of reactor core that happen to land in their town. However, nothing is not what the Chinese government or people in China would do and there would be considerable costs involved in such a disaster and if insurance is not paid it does not make the risk go away.

    I don’t know what the cost of installing solar power is in China, but they certainly have the capacity to install it at a lower cost than Germany and at German installation costs and using the 14% capacity factor you gave, point of use solar provides electricity at a lower cost than the average cost of grid electricity (http://reneweconomy.com.au/2013/graph-of-the-day-average-electricity-prices-around-the-world-24207) which means it out competes nuclear power. That’s using a 5% discount rate but an interesting thing about residential solar in China is that most Chinese people have a negative real return on their savings so the correct discount rate to use for many people might be zero or even negative. Last year China installed 12 gigawatts or more of solar capacity, most of it utility scale in the sunny western end of the country where in some places it is possible for fixed solar PV to have a load factor of over 25%. With a 20% load factor those 12 gigawatts alone would produce as many kilowatt-hours as 2.7 one gigawatt nuclear plants. Next year’s target is to install 14 gigawatts.

  27. @Ronald Brak

    If you tell me what method you used to determine that energy storage at under $600 a kilowatt-hour is a waste of money

    Just using your method with a 22c/kWh peak domestic cost is enough to do that.

    According to the cited website from the wikipedia Li-ion article, the best number of cycles is 1200, so on the optimistic assumption that this is achievable from a $600 a kWh battery (the Sunny Boy ones cost more than this I guess), each cycle of such a battery costs 50c, i.e. 50c a kWh.

    The cost of the just the battery itself is at least 50c for each kWh it delivers. (Even this ignores lost opportunity cost of capital employed.) Add the net 9c/kWh for lost feed-in and the system will cost at least 59c for each kWh, still ignoring lost opportunity cost of capital employed.

    Pity you don’t even try to answer my questions. Not even supplying a citation for your claimed battery throughput. That is the behavior of a troll.

  28. @ totarum.. Why are you paying so much for power in Melbourne? With all of the coal and natural gas Australia has why is power so high? Is the Australian dollar so much lower then the us dollar? I think not. Have the power rates always been so high? I pay .056 kwh 24 hour a day 7 days a week.

    from this site:
    Projected Levelized
    Cost of Power (2014-2043):
    4.7 – 5.2 cents/kWh
    Comparison Costs*:
    Natural Gas: 6 – 14 cents/kWh
    Wind: 7 – 10 cents/kWh
    Solar: 11 – 42 cents/kWh

  29. @ totarum… sounds to me like someone is artificially keeping your energy prices high and telling you suck it up its for the planet.

  30. Chris, I have checked my method and I have made a mistake. I’m sorry. You were certainly right that I was wrong and I apologise for being wrong. It was a stupid mistake and I behaved stupidly in making it and then not catching it. I now see that using the latest figures I provided for the Sunny Boy SMA home storage would not make money for someone with a marginal grid electricity cost of 30 cents a kilowatt-hour and a 5% discount rate until the cost of storage per kilowatt-hour was about $550. I was much too arrogant in my assumption that I was correct and far too childishly defensive of my opinion that I was correct. I apologise for my mistake and for defending it without checking it and for assuming that you were wrong when you were not.

  31. Brent, you wrote: “Really the only economical storage for the grid right now is pumped hydro which is only about 17% efficient.” That is not correct for the efficiency of pumped storage is not 17%, as Fran pointed out earlier in the thread. If you happen to be talking about a system where lamas are used to carry water to the upper resorvoir then you need to point that out as that is not what is generally understood by the term pumped storage. (I get an efficiency of about 17% for lama ported storage.)

  32. as was implemented at Fukushima

    Sure. But this is a management problem, not a technological one: you want nuclear power to be safe, you have to design management systems that can, sustainably-long-term and reliably, manage low-probability high-cost outcomes, and will do so if implemented in any cultural and regulatory framework, you have to design management systems that aren’t susceptible to the sorts of management failures that happened at Fukushima.

    But we’ve got pretty solid evidence from a wide variety of fields of endeavour that building management systems that avoid — reliably and over the long term — the sorts of management failures that happened at Fukushima is an intractable problem. Designing systems to manage low-probability high-cost outcomes is not something we really know how to do. The systems we can build simply don’t work reliably. We have a pretty good idea of exactly how they fail, but we don’t know any way to avoid that failure mode.

    Unless our management technology — not our reactor design technology — improves, nuclear power will remain unacceptably dangerous. “As was implemented at Fukushima”. What happened at Fukushima made the reactor unsafe. So: what are you proposing to do to stop the sorts of management failures that happened at Fukushima? How are you going to stop TEPCO being TEPCO, the japanese regulators being japanese regulators?

  33. I’ll go out on a limb and predict changes to our energy mix by 2020. There will be no built nuclear just more of what energy technology we have already but more expensive. With 0.2m a year net population increase we will have over 25m people. Our oil import dependence will go from what I believe is currently 60% to over 70%. Petrol will be over $2 a litre but road traffic will be less. Our emissions will still be over 500 Mt net CO2e a year, meaning we’ve virtually achieved nothing.

    Due to reduced incentives we won’t get from 1.2m solar roofs to 2m. With or without the RET wind capacity growth will slow, less than 30% from now I’d guess. Reasons being subsidy fatigue, takeup of good sites and nimbyism. The smart meter rollout will head Australia wide. LNG exports will be restricted (contrary to current policy) to keep the gas price down.

    This is somewhat different to the wind and solar powered nirvana some dream about. The public mood will be bitter and resentful eg at the cost of food, of driving and of keeping cool in summer. That’s why I think many will welcome nuclear power.

  34. @Collin Street

    One of the difficulties is that decisions about initial placement of a reactor system, i.e. the choice of site, are made with a standard economic model of cost in mind. Low frequency, high severity, risks just don’t figure into the calculations. I’ll be bold and say they are too easy to ignore through ignorance or through unbridled optimism (and bonus-related greed); furthermore, how does one truly assess the cost of an accident like Fukushima, at the time of site selection?

    To be blunt, the cost of an accident like Fukushima only matters if it is on your watch, and for those in charge of the original decisions that lock in the risks, there is very little likelihood of them being on that watch.

    Having said all that, to build a nuclear reactor so close to the shore and so close to sea level, at a location known for tsunami risk, knowing full well that the whole area is earthquake prone; well, that takes some breath-taking denial, I’d say. Obviously they want to be as close as possible to the shoreline in order to minimise cost of water access, among other things: pumping water 40m up land to a safer site is much more costly in terms of building and in terms of energy use during the plant’s lifetime, although in the big scheme of things it would seem like very cheap insurance against tsunami risk.

    Human factors get us time and again.

  35. @Brent
    Thank you for that very interesting link to the Columbia Generating Plant, Brent. But if you think your electricity is cheap because of that nuclear power plant you’d be mistaken. Construction on it started in 1975 and it took nine years to build. It had 2 billion dollars in cost over runs which is about $4.5 billion US in today’s money. So with a 5% cost of money and a 90% capacity factor over a sixty year lifespan the the capital cost of the cost over runs alone would come to about 2.5 cents per kilowatt-hour. Add to that the plant’s extremely high operating and maintenance cost which apparently comes to 3.6 cents a kilowatt-hour and the cost per kilowatt-hour comes to 6.1 cents which is higher than the average wholesale price of electricity in Australia and that’s before the cost of waste disposal, decomissioning, insurance, and even the pre-over run cost of the plant is included. So it looks like you are paying the marginal cost of electricity from the nuclear plant plus about 50% and not the full cost.

  36. Google fukushima-daini-model-of-a-safe-shutdown for an example of things going right. I understand that power plant was slightly more elevated. Excuse acronyms but ZCA like the idea of a high capacity HVDC cable crossing the Nullarbor to join the NEM to the WA grid. In that case perhaps nuclear reactors could be located on the Nullarbor coast (ie Great Australian Bight) such as at Ceduna SA or Esperance WA. In a sense that would fulfill Rex Connor’s dream of unifying the energy resources of both sides of Australia. Unfortunately the resulting Khemlani affair brought down the government in 1975.

    Since the GAB is well away from the Pacific ‘ring of fire’ I would expect tsunami and quakes to be rare to put it mildly. The main problem might be lack of network redundancy with a single cable. In past heatwaves HVDC rectifier substations have overheated. An alternative is putting reactors in existing coal fired stations using as much already in-place hardware as possible. In future summers I expect Perth, Adelaide and Melbourne to all hit 50C. Reliable power in extreme weather is a matter of public health not just affordable electricity bills.

  37. I’m going to say one last thing on this topic and then remain silent on it until there’s some new reason for discussing it.

    If the only criteria for assessing low ecological footprint stationary energy infrastructure were questions of engineering, then nuclear power would be a fabulous tool. If, tomorrow or even over the next 15 years, every fossil HC station contributing or likely to be called upon to contribute to a grid, in the parts of the world that have per capita emissions above 2tCO2e per person or as a jurisdiction more than 0.5% of world emissions were replaced by equivalent nuclear capacity, there would be a massive cut not merely in CO2e emissions, but in a whole lot of other airborne pollutants as well. If all motor vehicles then sourced their energy directly or indirectly from the grid, there would be an even larger cut. And if bulk carrier ships all switched to nuclear-powered motors, the seas would also be massively cleaner as well as cutting the carbon miles in goods transport.

    But that’s really only one of the questions to be considered. Firstly, as a matter of practice, the nature of nuclear power technology would be accompanied by a massive ramping up of the security state. It doesn’t really matter how ‘at risk’ nuclear plants would be from ‘you know who’. In practice the mere perception of their risk profile would demand increasingly bonapartist measures ‘for the public safety’ and all manner of secrecy since this would fall very definitely under ‘national security’. I regard it as fair to ask whether this is a cost most people would want to bear. I doubt most would, and in any event, it would impinge on accountability for the plants, and therefore our confidence that prudent provision was being made. The record in Japan has not been good.

    Moroever, it is very doubtful that a conversion program on this scale would even be technically feasible on such a timeline. There’s a limited amount of industrial engineering capacity for nuclear plants, and of course the capacity for independent and skilled examiners to oversee progress.

    Then there are the economic challenges. Almost all the FHC assets in stationary energy are privately owned, often held by trusts or pension funds. You couldn’t close them all down and hand their market share to nuclear plants without compensating the equity holders. So each new plant would have to begin its development on a timeline when an FHC plant of comprable capacity was likely to close, or else, the cost of the new plant would have to include the cost of decommissioning the FHC plants. That radically affects the cost-feasibility of new plants and the rate at which they can be built.

    Very few states have shown a serious interest in building new infrastructure out of public funds, and certainly not on the scale of a 15-year transition to nuclear power.

    Regrettably, such a speedy transition isn’t going to occur. There would be massive opposition, and a bill to make projects like the NBN seem like trifles. So it’s not going to happen.

    What we ought to do is to focus on things that can happen. By all means let us preserve whatever feasible nuclear capacity we have and welcome new and well-conceived nuclear projects if and when they arrive. Let’s give them all the R&D they need and do our bit (for a suitable fee) to accept waste and develop either reprocessing or storage. But let’s also build low footprint infrastructure that can replace FHCs in stationary energy in the next 15 years. Whatever that costs is going to be cheaper by far than nuclear, not opposed by many people, and of course, far quicker in doing the business of pushing FHCs off their patch.

  38. Earlier I stupidly typed in Colarado instead of Columbia, just another in a long list of mistakes I have made in my life, and in doing so I came across an article on the Fort St. Vrain nuclear reactor. It’s interesting because while some Americans are perhaps quite justifiably proud of the high load factor of their nuclear plants, there may be a strong survivorship bias. After eleven years of construction Fort St. Vrain came online in 1979 and shut down only 10 years later. In 1996 it was converted into a natural gas plant. And anyone who thinks that Australia can turn coal plants into nuclear plants should note that America can’t even manage to turn a nuclear plant into a nuclear plant. The plant implemented some new advances that were technically very successful except for various flaws resulting in the destruction of vital and expensive plant equipment. Perhaps this is a situation where the expression, “The operation was a success but the patient died,” could be is used? There were also problems not related to the new design such as contractors starting a fire that damaged control lines and then not telling anyone about it for five hours.

    Another US nuclear plant that shutdown long before its expected lifetime which I learned about today was the Trojan nuclear plant in Oregan. It came online in 1976 after only six and a half years of construction but was shut down for nine months in 1978 to fix construction errors and to improve its earthquake resistance. After a breakdown in 1992 the plant was shutdown for good as it was deemed not safe to operate. It was shut down after only 16 years service, or really 15 since it took a nine month break. The decomissioning of the plant is not complete and is estimated to cost $230,000,000.

    So it seems pretty clear that in addition to taking account of many other expenses and risks, the cost of nuclear power will need to reflect the very real chance of a plant having to shutdown well before its planned lifetime. And the risk of it suffering an early death is increased if it uses new technology. And some people have promised us an awful lot of new technology.

  39. @ Ronald… The Columbia generating station published power costs would include any cost over runs during construction. So the cost should have been less not more then published.

    As for ft st Vrain was a first of a kind gas cooled reactor. They had problems with the gas bearings in the turbine. Once they had the problems figured out the stockholders pulled the plug. So you had first of a kind reactor and first of a kind turbine. The reactor worked but the turbine didn’t. The lesson is do your home work and weight all the financial risks.

    The discussion has been about using small nuclear reactors in Australia. Using the same disign over and over. Larger mega projects are subject to mismanagement. Small mistakes grow into large ones.

  40. @Ronald Brak

    until the cost of storage per kilowatt-hour was about $550.

    This would reduce the minimum battery cost from 50c/kWh delivered to 46c/kWh delivered.

    You never give up, do you?

  41. @ Ronald Brak on the cost of nuclear, wind and solar in China,

    Ronald, no, you misunderstood the article you cited above. The price of $670 per kilowatt is the “bid-invitation” price for the purchase of a wind turbine from the supplier by the developer of the wind farm, to wit: “developers have realized that they can no longer so viciously cut the bid-invitation price for wind turbine generators.” That’s just the price of one piece of equipment, not the the total capital cost of the installed wind capacity, which also includes building the concrete foundation pads, transporting and erecting the turbines, building access roads and transmission lines to take the electricity to the grid, all the labor costs, site prep, etc.

    Total capital costs of new wind and solar projects are indeed in the range of $1500-1700 per kilowatt, as I reported. Bloomberg estimates that China is investing about $27.2 billion per year to build 16 to 18 gigawatts of wind. (http://about.bnef.com/press-releases/china-was-worlds-largest-wind-market-in-2012/) The 1.4 gigawatt Three Gorges Wind farm, for example, will cost $2.18 billion, or $1557 per kilowatt. (http://www.sustainablebusiness.com/index.cfm/go/news.display/id/24240). China’s build of 35 GW of solar by 2015 will reportedly cost $50 billion dollars, or $1425 per kw, but that doesn’t include subsidies. (http://www.scmp.com/business/commodities/article/1412862/solar-industry-bouncing-back-mainland-china-after-prolonged). Those prices may apply only to utility-scale developments. This article by the same author you cited puts residential rooftop PV installations at $2500 per kilowatt in China (http://www.renewableenergyworld.com/rea/news/article/2013/07/residential-solar-pv-systems-experiencing-slow-adoption-in-china?cmpid=rss), but another article puts a larger Beijing installation at $1506 per kilowatt, with a 12 percent capacity factor.

    I also high-balled the nuclear costs. The nuclear capital costs of $2500-3500 that I reported are for first-of-a-kind Gen III AP1000s and EPRs currently under construction. China’s homegrown Gen II+ CPR1000s cost less, about $2000 per kilowatt, but they are being phased out in favor of the Gen IIIs. But China’s mass deployment program aims to lower the cost of its AP1000s and derivative designs to about $2000 per kw and even less, so 20-40 percent declines in the nuclear capital costs I quoted are in the offing.

    Given the hugely greater productivity of nuclear, 88 percent capacity factors vs. 22 percent for wind and 14-15 percent for solar, there is simply no way wind and solar can compete on price. That is reflected in Chinese feed-in tariffs: nuclear gets 0.43 yuan (7 US cents) while wind gets 0.51-61 yuan (and still needs must-take rules to get grid managers to buy it). (http://www.chinadaily.com.cn/china/2013-07/02/content_16710593.htm) (http://www.reuters.com/article/2012/09/09/us-china-windpower-idUSBRE8880J720120909) Solar gets 0.75-1.15 yuan per kwh FIT. (http://www.chinadaily.com.cn/business/2013-03/14/content_16307608.htm) The disparities are even larger if you count the additional system costs of wind, like the $billions spent each year stringing power lines out to distant prairies, or factor in the many decades the nuclear plants will spend producing electricity at much lower rates after their capital costs are paid off.

    No question about it: Chinese nuclear is cheaper than wind and solar.

    –On capacity factors, Ronald, the 22 percent capacity factor for wind is after factoring out curtailments and the fact that much of China’s wind capacity is not connected to the grid. The raw production figures are much worse. In 2012, China produced 100.4 TWh of electricity from 62.7 GW average of grid-connected turbines for a capacity factor of 18.2 percent. As for solar, MIT says 13.6 percent CF, Scientific American 15.2 percent. (http://globalchange.mit.edu/files/document/MITJPSPGC_Rpt242.pdf) ((http://www.scientificamerican.com/article.cfm?id=china-scales-up-solar-power-50-percent) Can you provide a source for your claim of 20-25 percent capacity factors for Chinese solar?

  42. @ Fran Barlow,
    Fran, the polities with the highest proportions of nuclear generation include France, Sweden, Switzerland and Ontario, all of them very open liberal democracies. Your argument that nuclear power fosters an over-weening security state that impinges on freedom is clearly without foundation.

    I don’t understand your argument that deploying nuclear plants would have to wait for the slow retirement of existing fossil plants, or else require compensating their owners. I don’t think we do need to do that, but in any case why wouldn’t that principle also apply to replacing fossil plants with renewable sources?

    Your belief that renewables can accomplish a faster decarbonization than nuclear is also unfounded. Just look at Germany. This past week the new Social Democratic economy and environment minister announced yet more subsidy cuts and deployment slowdowns for the Energiewende, citing insupportable costs. In 2023, when the last German nuke closes, Germany will be generating about 38 percent of its electricity from renewables, providing it meets its targets, which is iffy. In 2010 its low-carbon fraction of electricity was 38.8 percent–so 13 years with no progress whatsoever on decarbonization.

    By contrast, the places I cited above all achieved 60-90 percent decarbonization of their electricity generation by prioritizing nuclear and hydro, with deployments that were much faster and cheaper than Germany’s program and with no resort to Bonapartism. Decarbonizing an industrial democracy using nuclear power is so easy and routine that it’s positively banal.

  43. @Will Boisvert

    Erm, what about all the oil they are still burning, especially in the transport sector? How do you de-carbonize that?

    Also, have you heard of peak uranium? Fissile uranium will run out. Breeder reactors are highly dangerous.

    As at 2010, 80.6% of world total power was provided by fossil fuels, 16.7% by renewables and 2.7% by nuclear. Nuclear power is all but neglibible. Renewables are killing nuclear, producing over 6 times as much power.

    Note: It is misleading to just talk about generation of electricity. It is total energy use that counts and the figures and facts show that nuclear is dangerous and near-useless in the total mix.

  44. @Ikonoclast
    I think you’ll find that half the renewables can be accounted for by hydroelectric dams built between 1950 and 2000. Sure wind and solar have grown strongly funny what quotas and subsidies can do. In Australia’s case there are shortfall charge penalties if you don’t meet the interim quota. Germany and other countries seem to be showing us it will get a lot harder integrating wind and solar from now on.

    The big white hope with transport energy other than inner city battery runabouts may have be hydrogenated fuels of which there are a number of types. Unfortunately none cheap. The big question is whether we can eventually adapt to fuels costing several times what they do now. We’ll need a low carbon energy source that can not only keep the lights on but get us to work, power tractors etc. We’ve flogged underground carbon to excess maybe the future energy source that can do the heavy lifting is based on E = mc^2.

  45. I’d just like to apologise once more. I’ve been making a string of stupid mistakes lately, although to be fair on myself I think I’ve mostly confined them to the period ranging from now back to the time I started breathing.

  46. Brent, if you look at the site you linked to you’ll see that for the Columbia nuclear Generating Plant they give the, “Projected Levelized cost of Power (2014-2043): 4.7 – 5.2 cents/kWh” when the plant started operating in 1984. That’s a neat accounting trick. You’ll find most forms of generation are pretty cheap if you work out their levelize cost from halfway through their lifespan.

    You also gave us their comparison costs which were: “Natural Gas: 6 – 14 cents/kWh Wind: 7 – 10 cents/kWh Solar: 11 – 42 cents/kWh” Now I don’t know how much these things cost in Colarado, sorry, Columbia, but new wind and solar are quite a bit cheaper here. Our Macarthur wind farm in the state of Victoria cost one billion dollars, has 420 megawatts of capacity and has a capacity factor of 35%. The Snowtown II wind farm under construction in South Australia will be 270 megawatts and will cost $436 million Australian and will have a capacity factor of about 42%. Is it reasonable to include a wind farm that’s still under construction? Well, since they haven’t even started construction on any nuclear power plants in Australia, I’d say yes. The lifespan of both windfarms is 25 years. I’ll let you do the maths to work out the cost od electricity per kilowatt-hour as I’ve been rather maths challenged lately. With regards to solar thanks to our high retail electricity prices rooftop solar is the cheapest source of electricity for Australian households and would be cheaper for consumers than nuclear power even if its cost was zero cents per kilowatt-hour. Rooftop solar, without any subsidy, outcompetes any form of grid based generation. As for the cost of natural gas here, you can discuss that with Hermit if you are interested, as it is something he is terribly concerned about.

  47. @ Ikonoclast,

    The 16.7 percent of renewables in the total energy supply is almost entirely hydro and, mostly, biomass. That latter is mainly people in the third world burning wood and dung for heating and cooking, which is horribly unhealthy, polluting and destructive to the environment. The world will be a better place if most of that 16.7 percent goes away. Nuclear produces many times more clean energy than solar and wind combined. Think a little harder about the statistics you cite, Ikonoclast.

    As for transport we can electrify trains with nuclear (old hat) we can run ships with nuclear reactors (also old hat) we can power automobiles with batteries (old hat in the 1920s) charged by nuclear power; we can distill liquid fuels from the atmosphere with electric power. Lots of ways to skin that cat. And no we’re not going to run out of uranium, there’s billions of tons of it in the oceans, in ores we haven’t yet mined, in phosphate deposits that we used to mine for uranium until uranium got too cheap, and elsewhere, with lots of good technical developments on getting it. People were saying peak uranium in the 1940s, its just gotten cheaper and more available as the years pass. Billions more tons of thorium, which is a great nuclear fuel. Breeders are not dangerous, some of them have been running for decades.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s