Saturday, September 24, 2011

French nuclear power pricing, and solar power pricing

Via Romm:
Pro-nuclear power France still has escalating costs for nuclear power. It's not American litigation driving these costs.

Also via Romm:

My opinion is that we should maintain and relicense most nuclear power plants (that's cheap), but nukes don't have an expanded role absent massive Republican governmental subsidies, with an unhelpful loan subsidy assist from Obama. A lot safer than coal, though.

UPDATE: probably should add that widely distributed wind power and grid charging off of power stored in plug-in car batteries could handle much of the night time load.


Anonymous said...

In Australia it seems that solar has already reached price-parity with (largely coal-generated) grid power.

Even if this is a slightly opimistic view, the reality won't be far away. I know that for me, personally, solar would be much cheaper than connecting to the grid when I build, even if I were to do so today.

Bernard J. Hyphen-Anonymous XVII

Anonymous said...


Permit me to wire up the linkie.

Bernard J. Hyphen-Anonymous XVII

Oale said...

As they say, one can't adjust the renewable energies fast or well enough, which I do nowadays accept, specially because of the altering amounts of hydropower available due erratic rains, I think there's always a call for some nuclear/other speedily adjustable energy sources. I'm imagining an energy policy where the continental-scale baseload would be from renewable sources, though call but if it would be easy that would have been done already.

Holly Stick said...

About relicensing nuclear power plants; wasn't one problem at Fukushima that the plant was old and maybe ready to shut down soon anyway? Not sure that recycling is a good idea in this instance.

John Mashey said...

0) MacKay's book/website is worth reading, see analysis of nuclear. I've heard him lecture and talked to him some, and he does interesting analysis.

1)While many places in the world (like right here) are unsuitable for nuclear power, and very suitable for renewables, that isn't true everywhere. In some places, it's hard to see how to get rid of fossil fuels without nuclear, just given the land footprints of many renewables.
(CA can do all right with solar, hydro, geothermal, a little wind. If we had better batteries, we'd get there a lot faster. If I had to pick a single thing that would get instant big VC funding, it would be a big leap in battery tech. Too bad it's hard.
Pacific NW has a lot of hydro. The midWest has good wind resource, and windfarms can coexist with agriculture OK there.)

It's unclear what places will do if they get less sunlight and lack big areas where wind work well, for instance, UK and US NorthEast come to mind.
[I.e., as always, do energy efficiency first, use renewables everywhere possible ... and if you shut down all the fossil plants, it's still hard to see how you power NYC, Boston, London. Maybe you can get more hydropower from Canada (seems relatively safe), but schemes to put solar in North Africa and run grids to Europe might be a little more nervous-making.]

2) While still early in R&D, Gen IV reactors seem well worth investigating, if nothing else, for the (only) hope of being able to clean up some of the existing long-lived radioactive waste and for closing down more of the fossil sources. Politically, unless there is a dense-baseload alternative to coal/gas *in a specific region*, people may want to close the coal plant, but "turn off the lights" is not politically attractive. Hopefully, if they can make Gen IV work and if there is less US administrative/governance screwup, they can solve two problems together. All this remains to be seen to be seen.

3) Of course, fusion research is worthy ... but counting on that as a solution might not be wise. When I was in high school (early 1960s) I was thinking hard about going into fusion research, since the claims made it look like it would get to commercial viability within 30-40 years and that seemed exciting to work on. Fortunately, I got sidetracked into computer science at end of senior undergrad year, where it was rather easier to actually build stuff and see it get used widely in a bit less time.)

HAUSMAUS said...

The fine print states that the graphs assume continued cost escalation for nuclear and continued cost depreciation for solar, on the order of +/- 4%/year (each).

That won't necessarily happen.

Note that the only reason we're still on the decreasing trend for solar prices is because the market collapsed. See the spike in panel prices in '07/'08 due to silicon shortage? The amount of PV installed in '07/'08 was trivial compared to total electricity generation. PV is effectively a commodity: a massive ramp up of solar energy will drive prices up, not down.

I respect Romm (his first books were excellent), but I get really irritated when he insists that nuclear can't overcome obstacles, yet solar, wind, etc. can scale by thousands of percent with ease. He's moving the goal posts.

If we had a strong carbon price, we could just leave the market to sort this out.

William M. Connolley said...

The figure you show is hard to read, and gives to my eye the impression that the US and France fit a pattern. Look at fig 12 instead ( and that disappears. I'm dubious that "negative learning" explains the huge jumps.

But if you believe Romm that solar is cheaper, there is no problem.

quokka said...

For a far more comprehensive view see the IEA 2010 Projected Costs of Generating Electricity. Section 3.2 (starting page 59) provides a country by country assessment of the costs of nuclear, coal, gas and renewables. Spent fuel management and decommissioning costs are included for nuclear. Nuclear looks pretty good - competitive with coal and gas and cheaper than solar and wind by some margin judged on levelized cost of electricity alone. Nuclear has a further network cost advantage over intermittent renewables because it is reliable and dispatchable.

The UK Climate Change Committee, which has done some very good work, assesses nuclear as likely the lowest cost low emission technology now, and right through to 2040. The Renewable Energy Review makes it quite plain that in their view the dominant technologies for UK electricity generation must be nuclear and wind. Solar not so much.

Flavius Collium said...

Maybe it's the building rate - if you keep building new nukes, it stays as a "healthy" industry: the parts manufacturing can churn out more parts with little more capital investment, new people choose it as a career path, are motivated etc etc.

Once you reach a steady state, it's seen as a dead end and a liability. The best and brightest go elsewhere. If you only build one in a decade, it's going to cost a huge amount. The contractors have to keep the tooling in shape just for small production runs resulting in large unit costs, everything has to be relearned etc.

Anonymous said...

Off topic, but the sun seems to be waking up: solar flux 190
Rabid Doomsaying Little Mouse

Every other time I have said that Mr Sun has hit the snooze button

David B. Benson said...

I've now put some considerable effort in attempting to find the least cost way to energize a reference power grid using only scalable, low carbon and existing technologies. Turns out that NPPs remain the least cost way, although there may be room for some residential solar PV; that last is up to you and your utility.

David B. Benson said...

The reference power gird load is derived from actual data but scaled and simplified. The load over night is exactly 20 GW for 7 hours. The daytime load is exactly 28 GW for 17 hours, 6 am to 11 pm.

You are invited to cost out any existing low carbon and scalable technology you wish. As far as I have gone 100% NPPs is the least cost. [And no, batteries won't do it, not at all.]

quokka said...

A comparison of two projects:

1. The Moree Solar Farm in Australia. Said to be one of the largest PV projects in the world, it will be one axis tracking PV, 150 MW capacity and 30% capacity factor. Project cost is reported as $925 million. On that basis levelized cost of electricity will be something like $250/MWh or more depending on financing. Planned completion 2015.

2. The
VC Summer Units 2 and 3 AP1000 Nuclear Power Plants
in the United States. Total capacity 2,234 MWe. Unit 2 scheduled for completion in 2015 and Unit 3 in 2019. Capacity factor 90%. According to page 28 of the linked Analyst presentation, cost of electricity is estimated at $76/MWh.

Scaling the PV plant up to the equal of the nuclear in terms of amount of electricity generated would represent a capital cost of of ~$35 billion. Of course the cost of PV project may drop and of the nuclear project may rise, but there are no guarantees and those cost changes would have to be huge to achieve parity.

Moree is a reasonably good solar location. PV cost in say northern Europe would be much higher.

David B. Benson said...

Saudi Arabia is planning on building 16 new NPPs. I had rather thought it was fairly sunny there.

EliRabett said...

Costs are interesting esp for nuclear vs. house installed solar which beats down distribution costs. Insulation is best.

bluegrue said...

Off topic.

Eli, whatever you've done with your new google post, your entry and your blog home page both automatically take you to youtube.

Pinko Punko said...

What blugrue said- the Rabett's page is almost impossible to get to. This is not good!

Rattus Norvegicus said...

Yep, it's a mess. I got back here by using my history to land on something other than the home page.

Anonymous said...

Your blog has gone odd... Seems to link to YouTube for some reason... WEIRD

Rattus Norvegicus said...

Working OK now...

David B. Benson said...

Regarding the VC Summer NPPs, there are no subsidies but there may well be a US DoE loan guarantee.

dhogaza said...

"When I was in high school (early 1960s) I was thinking hard about going into fusion research, since the claims made it look like it would get to commercial viability within 30-40 years and that seemed exciting to work on."

And in 2060 there will be bright high school kids thinking hard about going into fusion research, as it will appear as though it will get to commercial viability within 30-40 years ...

Brian said...

Couple of comments:

Generally, I agree that an adequate price for carbon could fix all this. Also that it's somewhat dangerous to take a trend line and extend it indefinitely. OTOH, the trend line for solar pricing has a long history of moving down.

Bernard - seems like they're including large infrastructure costs for coal, but if the infrastructure's already there, it's a sunk cost.

John Mashey - Interesting book link. I would somewhat grudgingly support government subsidies for nuclear research. My marginally informed sense of these new reactor types is they've been just over the horizon for quite a long time now. And it doesn't end the role that nuclear power has had in nuke weapons development.

Belette - sorry for the graphs, I tried to squeeze them on the page. Your link's better, but even there it's undeniable that the trend line is up in both countries. Here in the US the Republicans say it's American litigation artificially driving costs, but that doesn't work for France. Honest figures from China would be very interesting....

Quokka - that is a present day graph from the IEA, not a forward projection. Biomass did better than I expected. Biomass (if done correctly) plus carbon sequestration is one of very few carbon negative processes available that might save our hides several decades from now.

John Mashey said...

1) Back when I read the ~1956 book Project Sherwood, they thought commercial fusion was ~30 years away, with luck. A few years ago (Nobel physicist) Burt Richter thought it was at least 50. He might or might not be right, but so far history seems with him.
2) I go to Stanford energy seminars, including one where Burt brought in one of the leaders in the Gen IV work.

3) Most people have little famialiarity with the nature of serious long-term R&D, which is why I wrote

One has to run research pipelines that run decades, and the energy business is tougher than what we were doing, since Moore's Law only helps a little.

4) Some if the advanced reactor types were proposed decades ago, but the combination of TMI and Chernobyl broke the research pipeline badly for a few decades, and contrary to "Breakthrough" foolishness, it is difficult to make up for lost time in R&D pipelines. As I noted in R2-D2, "never schedule breakthroughs" and never deploy anything widely unless you are sure it works... and all that takes time.

5) The real problem is that neither of the two extremes is very useful:
A) Nukes are THE answer, this renewable stuff us fantasy, and if only the evil environmentalists had let's us do our work, we'd have gotten rid if coal long ago. (Most commonly heard from babyboomer nuclear engineers who thought they were entering a growing field.)
B) No nukes anywhere, ever.

Realistically, conditions vary geographically, whether or not one it more of the Gen IV technologies works is still unknown, although my sense, talking to serious people, is that at least one of them can be made to work within the next decade or two. (I think that very unlikely for fusion).
Likewise, the economics remain to be seen. I am certainly attracted by the idea of processes that extract much more energy per input and waste output, especially if the can convert existing long-lived wastes.
I still recall SF stories from the 1950s where tens of thousands of years from now, hereditary priesthoods gave evolved to take care of these places they don't understand, but where people get sick if they get too close.

Flavius Collium said...

John Mashey, are you familiar with the molten salt reactor that was run at Oak Ridge? (LFTR = liquid fluoride thorium reactor is the more modern equivalent design family concept.) Invented by Eugene Wigner and championed by Alvin Weinberg no less.

It would be interesting to read some wide overviewing book about these only very partly explored designs which have potential for much smaller fuel demands (Say 1 t thorium for 1 gigawatt-year or a tennis ball for your whole lifetime usage) and less waste and very little long lived transuranics.
Reactors are such a rich design space as compared to something like combustion where there's just a few simple kinds that can operate.

EliRabett said...

Apologies, Willard was the first to point out the problem. It was a messed up url. . . .

John Mashey said...

Flavius: yes.

The seminar I mentioned was this by Jacques Bouchard, who has been the Chairman of GIF since 2006. I talked to him at the mixer afterwards and he seemed pretty reasonable, i.e.e, optimistic that at least one of the technologies would work, but clearly not expecting this overnight or to be trivial.

Those new to this might start with history.

This page links to the various technologies, of which one is the MSR.

Of course, I was also thinking of Argonne's IFR, which may or may not have worked out ... but might well have been instructive had it been completed instead of being cancelled.

quokka said...

@John Mashey,

I think you are rather overstating uncertainty about whether Gen IV nuclear technologies will work. One of the Gen IV Forum technologies are sodium cooled fast reactors. Sodium cooled fast reactors are not new and there is something like 400 reactor years operational experience. The Russian BN-600 has reliably run for decades and the larger BN-800 is under construction. The GE-Hitachi PRISM is a full commercial design for a passively safe sodium cooled fast reactor - the outcome of decades of research at Argonne. It is just waiting for somebody to build it. It may very well be the best and safest reactor design ever.

China is commencing construction of another Gen IV Forum design - the gas cooled high temperature reactor after running the test HTR-10 since 2003.

The question is not really whether these will work, but why shouldn't they work.

Molten salt reactors are another matter and are going to take a lot longer to get to this point. A search for claimed perfection with LFTRs should not hold up development of the merely very good.

David B. Benson said...

To set the history straight, the first nulcear weapons exploded in 1945 and it wasn't until several years later for the first demonstration of nuclear generated electic power, enough to light one ligt bulb. The first serious use of nuclear power was for submarine propulsion; the current PWR/BWR designs are all based on that.

Unfortunately as it may seem to some, of the scalable, low carbon power genration technologies available now, only nuclear can meet the requirementss for a reliable, on-demand power supply at utility scale and for reasonable cost..

John Mashey said...

quokka: I don't understand your comment. Did you not see the comment posted a few minutes before yours? I've visited Argonne & Fermilab, etc out there, and I was the first in this thread to mention IFR, and at least one of my links notes that people thought it was one of the most promising. That's why I brought it up, having been asked about MSR.

Steve Kirsch is an old friend, see this. Of course, there were the comments starting here.

I wrote:
"Realistically, conditions vary geographically, whether or not one it more of the Gen IV technologies works is still unknown, although my sense, talking to serious people, is that at least one of them can be made to work within the next decade or two."

Serious people = folks like Richter and Bouchard. Maybe I misinterpreted them ... but they've spent their careers doing advanced R&D, and such people tend to know that not all good-sounding things can be made to work well .. which is why GIF is pursuing multiple routes, exactly as they should. Some will drop out.

Maybe Burt will have more to say in 2 weeks, in his GCEP talk, "Perspectives on Nuclear Energy."

David B. Benson said...

Graham Palmer has a useful essay:

quokka said...

What is entirely missing here is a recognition that LCOE from PV under the very best circumstances and even with ambitious projections for decrease is very much not the be all and end all. A very questionable assertion about optimal LCOE for solar vs nuclear is being used for purposes for which it is unfit. One thing that is entirely missing is sensitivity analysis.

Refer to the Section 6.2.4 Construction costs and lead times of the IEA Projected Costs of Electricity Generation. LCOE of PV is the most sensitive to both construction delays and cost escalation. Substantially more so than nuclear.

Of much more and probably critical importance is Section 6.2.5 - sensitivity to load factor. LCOE of PV and wind are highly sensitive to load factor - far more than nuclear.

At small penetration in an obliging grid where the operators are forced to take intermittent generation and the generators can run at their full load factor, not so much of a problem. But consider the case of PV in Germany.

Germany has about 18 GW of PV capacity. On a good day in mid-summer, output may peak at perhaps 14-15 GW. On a poor day in winter, it is next to useless. PV currently supplies no more than 3% of Germany's electricity.

If PV in Germany were expanded to meet say 10% of Germany's electricity demand, then for some period on the best days in summer, every other generator on the grid - including wind - would have to yield to PV if PV is to achieve it's maximum load factor. The capacity factor of PV in Germany is already poor at around 10-12%. Further growth of PV and the situation grows progressively worse.

Then, there is storage. But storage is a very uncertain matter at this time even to the extent of what is likely to be the dominant technology or technologies. At the moment, it's pretty much pumped hydro or nothing. What is absolutely certain is that it will cost money and must inevitably increase the LCOE.

Attempts to close off the nuclear option based on extremely flimsy arguments resemble anti-nuke activism far more than genuine analysis or even concern about climate change. Shutting down nuclear is a risky business indeed. If the results of the German experiment turn out to be less than hoped for, which is entirely possible, what then? A nuclear industry cannot be turned on and off like a tap and we are in for real trouble and no way out.

Sobering facts: Today nuclear produces 21% of OECD electricity. Wind, solar, geothermal and other about 4%. After how long?

quokka said...

@John Mashey

Sorry, I didn't see your comment. Otherwise I probably wouldn't have posted.

John Mashey said...

quokka: no problem. I've been on Rottnest, so I'm fond of quokkas.

Alex said...

The thing that stands out in that chart is that there is a very obvious inflection point in 1985...

Anonymous said...

The problem is that comparing Solar to Nuclear using kW is not very instructive - as Solar's availability is much lower. Graphs comparing GWhr delivered per year would be much more illuminating.

seamus said...

Cost isn't the right question, costs will reach parity, more or less. Capacity; scale. Those are the things to look at. And when you do, it's obvious that renewables + conservation alone has zero chance of providing the capacity to displace fossil fuels. Intentional or not, anti-nuclear is pro-coal. Period.

" became obvious to us that one could put a total reactor concept together that would at the same time give you safety of a kind that reactors today don't have, that would allow complete recycling of the fuel, and thus extension of the ability to produce energy (very roughly, by a factor of 100), and also a waste product that did not contain the most dangerous elements. So with one concept you attack all of the principal real issues that there are for the use of nuclear energy."

Dr. Charles Till | Nuclear Reaction | FRONTLINE | PBS

John Mashey said...

Till = IFR.

Meanwhile, I heard a good PV solar talk, for which this has more background.

Normally, I would run in fear from GaAs, but Eli Yablonovitch was sharp (besides having been a Bell Labs research Director, not a job filled by dummies.) The video will be up sometime, and meanwhile, there is a link to a paper, although one may want a bit of physics background to go in there.

IF this works, potential = very low material cost (~ First Solar's thin film, which gets 11% efficiency), but efficiency @ 28% now, likely gets above 30%, not too far from SQ limit for single-junction, at much lower cost than Sunpower's 22%.

He showed a sample tested by NREL, in flexible piece of plastic.

Of course, reducing material cost does not itself help BoS costs, but this might be a real step function (as opposed to the slow, relentless cost/volume curves).

EliRabett said...

Not enough Ga (or In) in the world to make GaAs useful for such applications

John Mashey said...

Availability: That's one reason why I would normally run in fear from GaAs ... but Eli says they are at 1 micron thick now, think they can get it down to 10% of that, i.e., 100nm, assuming no surprises in what they are seeing with epitaxial liftoff. hence, tiny amounts of material go a long way.

Over long term, I hope more for Nate Lewis's Caltech, MIT efforts to do nanostructured Si, i.e., to use hugely-available elements, see p.12 of his GCEP talk, which also goes straight to H2 and O2.

Still, as per R2-D2, I would label Nate's that as R1/R2, whereas Alta Devices is more like ~D2, i.e., a few steps further along. I.e., one is in research/applied research, the other is into Advanced Development.

As I was seeing this talk, I was thinking: I wonder what if my NEA friends think of this, thinking I should alert them ... but of course, Yablonovitch would have worked for Arno Penzias at BTL, and this is the kind of thing (an old boss of mine) Forest Baskett looks for ... and by end of the talk, Eli had noted that NEA and Kleiner-Perkins had already invested. (Between them and others there is already $72M.)

Folks like NEA & K-P do not fund research and do not fund things unless they think they can be grown big. (For those not into VC-land, NEA and K-P are 2 of the top Silicon valley VCs. That does not guarantee success, but it's a hint that very serious folks take this quite seriously. Disclosure: I'm an LP (investor) in one of the NEA funds, although not this particular one, so I have no financial interest.)

EliRabett said...

AFAEIK GaAnything is the condensed matter version of fusion energy. . in the next 30 years, etc. It is REALLY rare.

John Mashey said...

I don't think this is a good analogy.
1) People use GaAs, it works.

2) The issue is the level of rarity, and again, see p.12 of Nate Lewis' talk.
His chart may or may not be right, but certainly Ga and As look less rare than, say, Cd and Te (i.e., First Solar). Of course, for really huge deployments, we want the elements at upper left, which is what Nate's arguing for.

3) Some of the same techniques help Si cells, just not as much. But, if somebody can ship $Bs of low-cost cells, that will put mroe pressure on BoS costs.

John said...

Brian, your figures purport to show that the cost of photovoltaic solar power will be comparable to nuclear power in a few years.
Burton Richter (Nobel-prize winning physicists at SLAC) has a book, Beyond Smoke and Mirrors, in which he gives figures for the cost. Richter says that nuclear power is comparable to fossil fuels, while photovoltaic is about 4 times more expensive.

These are not precise numbers, because for fossil fuels, the cost of fuel is a large fraction of the total cost, while for nuclear power, it's the up-front capital cost (not fuel costs) that dominates the total costs. (What is the interest rate on the loan? How long does it take to build the plant? etc.)

The cost of photovoltaic includes not just the solar panels, but "balance of system" (everything else) which is sometimes overlooked.

Don't forget: at night, solar photovoltaic produces zero power. You can store the energy during the day, and get some of it back at night, but this process is about 50% efficient. So solar energy, stored during the day and released at night, is now 8 times more expensive then nuclear.

IMHO, the present big three power sources (coal, oil, natural gas) will be replaced by solar, wind, and nuclear power. Nuclear provides baseline power (available 24 hours/day) which wind and solar can't always provide.

EliRabett said...

GaAs works, but there is a thinning limit which is not much below 1 micron, because as it gets thinner, you lose absorption of the sun light at the longer wavelengths. At that point you are decreasing efficiency while decreasing costs.

and Ga is rare.

John said...


You mentioned in passing what Germany is doing. Germany is phasing out nuclear power in Germany. So where is Germany going to get electric power? They're buying it from France, which gets electricity from their own nuclear power plants!

TCO said...

coal is the shiznet. And I say that as someone with way more technical and econimic ersonal rationale for supporint either nukes or solar.

I'm just a fucking honest man.

Brian said...

John: if both trend lines are correct - nukes getting more expensive, solar getting cheaper - then the absolute values are somewhat less important. Eventually they cross. If OTOH the absolute values are off by 400% as you suggest, then the analysis is so bad that there's not much point wasting time to figure out if the trend lines are accurate.

Anonymous said...

You can see disparities of the magnitude of 400% can arise if you look at the max output of solar plants and then their annual output.

The Japanese Power Utility, TEPCO, of Fukushima Dai-ichi fame has recently inaugurated a Solar Plant at Ukishima. It has a max output of 7MW, and an expected annual output of 7,400MWh. To convert this annual output to the average power production level you just divide by 8160 - the number of hours in a year (365 days). You get:

0.85MW - around an eighth of the maximum output.

Now the caveat is that this is Japan, and other sites around the world will probably get better ratings - but given the effect of night, morning and evening I doubt they'd more than double that value. Anyone got a table of global atmospheric transparencies - something to confirm or refute my suspicions?*

And as an aside I was in Tokyo Station last week, early afternoon. They have a display of the outputs of their solar panels in the Shinkansen waiting area - they were getting .23kW per M3, and the power output was displayed in terms of households powered. The output varied from 4 to 6 households - in a display of 26 households, so at noon in late summer they were getting around 25% of the total capacity. Weather was fair, not too cloudy. Power was fluctuating 50% from minimum over a period of minutes. I doesn't appear that solar is the panacea it is made out to be - though it obviously has a role in the future power sector.

Anonymous said...

Forget I mentioned something as silly as "atmospheric tranparencies" how's about a table of average solar insolation for a wide range of places around the world.

Edward Robins said...

The perfect solution of energy problem is only using solar energy.

solar power

Solar power for homes said...

Solar energy is the biggest source of energy. It can save lot money for us.

KAP said...

If you go back to the original paper from which Romm drew his chart ( , figure 13) take a close look at figure 3, which shows what's really driving the rise in French NPP capital costs: it's construction time, which roughly doubled between 1983 and 1998, as capital costs roughly doubled during the same period. Then there's that one outlier in 1999.

Richard C. Lambert said...

The Guardian view on solar power: put in the shade just when it needs the sun ... Solar power to light the way for Africa as low-carbon campaign