Ambitious Solar Plans in France

The following guest post was written by Tom Standing, a "semi-retired, part-time civil engineer for the City of San Francisco." In Arizona Solar Power Project Calculations, published a few days ago, Tom took a look at a 280 MW solar thermal plant in Arizona. In this post, Tom examines France's ambitious solar plans.

The December 1 issue of the Oil and Gas Journal carried a “Quick Take” article about France’s “national plan for renewable energies” that they unveiled on November 17. Their plan includes all the popular ideas for alternative energy: biomass, wind, hydro, waves and tides, with a major emphasis on solar. For now, France has 13 megawatts of installed capacity in solar, but the energy minister wants solar to be an ambitious 5,400 MW in 2020. He says that France will change its carbon-based energy model to a completely decarbonized model: each home, company, and community will produce its own energy.

Why is France doing this? They already have the least carbon-intensive energy system of any industrialized nation. They generate 75% of their electricity with nuclear, supported by the most extensive technology to reprocess spent nuclear fuel, compared to any other nation in the world. Practically 100% of their rail system is electrified, packed with people, whether on the Paris Metro or speedy intercity trains. France has already developed the working model of a low-carbon energy system for other nations to emulate.

Let’s do some calculations on France’s solar plan, similar to my last email. We can see what surface area of collectors would be needed, and how much electricity would be generated.

Collector Area

As I explained previously, solar panels are rated at their maximum output, when the sun is near its highest altitude for the year under cloudless skies. Under such ideal conditions, insolation is about 1,000 watts per square meter. The most cost-effective panels convert about 10% of insolation into useful electricity, a factor that has remained unchanged for 10 years. Some PVs might convert 15%, but they cost more and are not mass-produced. Thus a typical panel of one square meter is rated at 100 watts.

To estimate the area of PV panels that France wants to install, we simply divide 5,400 MW by 100 W/m2 and we get an incomprehensible 54 million m2! It means that one million homes and businesses would have to be covered with 54 m2 of panels. A typical home can accommodate only 25-30 m2, so more than a million buildings would have to install PV. I do not know the worldwide capacity for manufacturing PV panels, but I would guess that current capacity is a small fraction of 54 million m2/year. Certainly, capacity will grow, but what about PV for Germany, Spain, UK, the Low Countries, and the US? California alone would suck up a major chunk of that capacity.

Solar Electricity Generated

Let’s say that France actually installs 54 million m2 of solar by 2020. (I doubt it is possible, but let’s carry the scenario through.) How much electricity will the fully built-out system generate?

First, we need to estimate the insolation upon the collectors. While I have copious insolation data from NREL for the US, I have no site-specific data for Europe. But I can make a reasonable estimate. Having traveled throughout France at various times for a total of about 3 months (typically in the summer), I can say that France has mild sun conditions. I would compare the French summer sun to that of Cleveland or Minneapolis. However, France is located at somewhat higher latitudes, which tends to reduce midday sun strength and spreads it out over more daylight hours. The northern suburbs of Paris, say around de Gaulle International, is latitude 49o, the northern-most boundary between the US and Canada. The south-most reaches of France, are between latitude 43 and 44, equivalent to Buffalo, NY or Portland, Maine.

I will pick a number on the generous side for annual average insolation in France, equivalent to that of Boston, New York, Chicago, and Minneapolis:

4.6 kWh/ (m2-day).

This level of insolation is for optimum panel orientation: facing due south with no shading, tilted at an angle equal to local latitude. Varying amounts of shading with less than ideal orientation will reduce the insolation on the collectors of most installations.

Now we are ready to calculate the annual energy generated from the fully-built French PV system. As I showed in Section 8 of my previous email, the annual energy generated by a solar installation is the product of four factors:

Insolation, average day during a year = 4.6 kWh/ (m2-day)

10% conversion of insolation into electricity, the industry standard for PV

Area of solar collectors = 54 million m2

365 days/year

Cancelling out units and carefully watching orders of magnitude, we come up with 9 billion kWh of useful electricity generated during the first year of complete build-out. But we need to give this number some perspective.

Energy Generation in Perspective

EIA statistics show that French consumption of electricity grew from 353 billion kWh in 1992 to 415 billion in 2002, or 62 billion kWh in 10 years for an average gain of 6.2 billion kWh/year.

It means then, that this huge solar development would, at best, produce the equivalent of only 1.5 years of gain in France’s electricity consumption. And it would take a 12-year crash program to install that much solar!

Another comparison is with the annual power output of one of France’s 1,000 MW nuclear power reactors. If the reactor operates for a year at 90% capacity (typical for the industry), the three factors to multiply are: 90%, one million kW, and 8,760 hours/year. Multiplying out these factors, we find that a single reactor would produce about 8 billion kWh/year, roughly the equivalent electricity as all the solar panels covering nearly 2 million homes and businesses.

Costs for PV

In the US, homes and businesses that install PV typically receive “rebates,” (another word for “subsidies”) from state or local governments, or the utility, to be paid for by all ratepayers. Rebates usually amount to about half of the total installed cost. The unit cost for solar installations has changed little since 2000, in the range of $600 - $700 per square meter, or in the terms of the industry, $6,000 - $7,000 per rated kW. Thus a homeowner usually qualifies for a rebate of 6 or 7 thousand $ after installing a 2 kW PV system.

I don’t know if French taxpayers and ratepayers will subsidize solar installations, but the unsubsidized cost remains the same and must be paid by somebody. Total installed cost of the solar plan for France, then, would run in the neighborhood of $35 billion. What is the cost of building a single reactor in a nuclear power plant? Considerably less I would say.

Solar Capacity Factors

We can calculate capacity factors for any solar project directly from insolation data. This provides a shortcut to calculating electrical output on an annual or monthly basis when we are given the nameplate capacity. The capacity factor depends only on insolation during the time period in question, and is independent of the conversion factor between insolation and electricity.

The capacity factor is defined as the ratio of actual energy generated, to the energy generated at maximum insolation, i.e. nameplate capacity. Therefore, our ratio is:

actual average isolation/maximum insolation

Maximum insolation, we have seen, is 1,000 W/m2.

Actual insolation is given in kWh/ (m2-day). We have to convert this quantity into units of W/m2. We cancel out hours and days by dividing actual insolation by 24 h/d. For example, if insolation is 5.0 kWh/ (m2-day), the average power output during one day is 5,000/24 = 208 watts. Average power output divided by maximum insolation (1,000 W/m2) gives a capacity factor of 20.8%.

Solar Capacity Factors

Because the ratings that are given to solar installations are only during the most ideal solar conditions (which, for many locales are achieved only during 40 or 50 hours per year) we must include the solar capacity factor for the locale under consideration. Multiplying the capacity factor by the local insolation enables us to determine actual power output and energy generated for a given time span.

From the NREL insolation database, we can calculate solar capacity factors for a wide expanse of locations across the U.S.

Select the link “In alphabetical order by state and city.” Then choose any city to bring up the data as a spreadsheet.

The striking fact gleaned from the database is that the range of solar potential across the entire U.S. is quite narrow. In the Lower 48, capacity factors range from the highest, 27.5% in California’s Mojave Desert, to about 15.4% for the Pacific Northwest. Vast reaches of the Desert Southwest: most of Arizona, at least half of Nevada, southern Utah, the southern half of New Mexico, and the western tip of Texas exhibit capacity factors of 26 to 27%.

Other Regions

  • Southern California basins: Los Angeles and San Diego – 23.3 to 24.2%
  • San Francisco Bay region – 21.7 to 23.0%
  • Denver/Boulder region – 23.0%
  • Mississippi Valley (Minnesota through Louisiana) – 19.2 to 20.8%
  • Great Lakes region – 17.1 to 18.3%
  • Eastern Seaboard (Maine through Virginia) – 19.0 to 21.0%
  • Florida – same as Southern California – 23.3 to 24.2%

From these capacity factors we can calculate that a large rooftop system of 30 square meters of PV (a 3-kilowatt system) costing about $20,000, oriented optimally toward the sun, located in the U.S. Heartland (in or around Kansas City or St. Louis), would actually generate about 600 watts, averaged over night and day for a year. The average daily generation works out to 14.4 kWh. This is roughly the usage of a typical household with modest lighting and moderate use of electrical appliances. For the same location, the capacity factor in July would be 24.6%; in December, 13.3%. The 3-kW system, then, would yield a power output averaging 738 watts in July and 400 watts in December. Average daily electrical generation would be 17.7 kWh in July and 9.6 kWh in December.

I welcome those interested in solar applications to calculate average power output and average daily energy generated, from the capacity factors shown above, for typical household-sized solar arrays.

Your assumed roof areas seem a bit small. Mind you, I am assuming that if you go to the trouble of doing this, you go to the trouble of building the house to match, so that the roof is braced to take the extra load, and you might as well at that point is the classic American saltbox house design, in which the roof peak is at one edge, not the center, so that you can use the entire roof. A few million homes over 20 years appears to refer only to new home construction. Even if you do not use saltboxes, 25-30 m^2 is only 250-300 square feet of receptor, which seems rather small.

Some would suggest that relocating the solar panels to Algeria and laying cables would work better.

We've talked about the Algeria option (which makes much more sense) here before:

With regards to the article, why were figures for northern France rather than southern France used ? Surely PV would go mainly towards the Mediterranean coast, where the hours of sunligh are longer and the periods of cloud far fewer ?

Tom - thanks for your interesting analysis on this second type of solar installation. Here is some more detail on solar panel efficiencies/capacity factors from a consulting engineer I know who works in this field. This data is for Michigan.

"A number of factors come into play when calculating the PV energy production per surface area of the PV panels. The 5.73kwh/sq.ft./year production for the Oakland University system
are based on the system's "amorphous-silicon" PV singles and with some of the PV shingles mounted on east and west facing roof sections, rather than all on the south facing roof sections normally used to maximize production. Not all available energy is captured when the panels are oriented other than south and amorphous-silicon PV cells are only about 7% efficient in converting sunlight to electric power. By comparison, multi-crystalline silicon panels are about 12.8% efficient and the latest technology of combined single crystalline cells with a thin layer of amorphous-silicon is rated at 17.4% efficiency.

So, here are some real numbers:

1) A utility connected PV system using popular multi-crystalline PV panels (12.8% efficiency) yields approximately 15.5 kWh/sq.ft./year into the AC power system.

2) A state-of-the-art PV system using 17.4% PV panels yields approximately 21 kWh/sq.ft./year into the AC power system.

3) An average PV system using single crystalline-silicon PV modules yields approximately 18 kWh/sq.ft./year into the AC power system.

These numbers depend on having a properly designed PV system using modern, high efficiency power inverters and low wiring losses. The PV panels would face south and be tilted at approximately 35 degrees above horizontal. If there were multiple rows of the panels, there would have to be a larger "footprint" in order to prevent shadowing of the rows placed behind the initial row."

Can anyone comment on this? It seems highly dubious:

>> The most cost-effective panels convert about 10% of insolation into useful electricity, a factor that has remained unchanged for 10 years.

Scanning Wikipedia, installations were 2.8 GWpeak in 2007, so a total of 6 GWpeak by 2020 for France alone could be viable. It is only one country -- but that's also a decade away.

Furthermore, 54 million m2 is a huge area, but it's "only" 54 km2. France is 675,000 km2; Paris is 87 km2 (small by North American standards). Only a fraction of urban area is viable for rooftop panels, but it should be possible to find 54 km2 in France.

- - - - - - -

On the other hand, the comment about the relative expense of nuclear vs. solar seems fair. McKinsey published some analyses of solar cost estimates which I'll need to dig up, but Climate Progress cited a study arguing total-cost-of-nuclear is in the 25-30 cent per kWh range.

Nuclear opponents always look at 1970s us plant construction for costing data and neglect that "the nuclear" industry DIDN'T die in the rest of the world. World nuclear power plant construction tends to come in around $2000/kw for single-plant construction (no multiple unit discounts for design and parts), and is extremely well documented.

Beware of anyone with an axe to grind.

Real estimates from investor owned utilities are coming in much higher than $2,000 per kw. See

Tianwan-1 & 2 2 gwe nameplate pricetag US$3.2 billion onliine 2007
Lingao "They are reported to have cost $1800 per kilowatt"

So, long story short, it may just be that america is useless for new projects.

A bullet in the back of the head for those that criticize does wonders for cost control.

One cannot "cherry pick" one data point about the Cinese economy without considering the entirety of the society.


Chinese nuclear plants may look cheap, yet chinese coal plants are much cheaper still, and easier & faster to build with a more mundane workforce. Chinese coal plants aren't exactly the cleanest, however. And I wouldn't trust these cost figures from China too much anyway. The subsidy regime is opaque and sometimes costs aren't counted in the same consistent manner between projects (or in different countries).

France also includes overseas departments in the Caribbean, French Guiana, Reunion in the Indian Ocean plus smaller populated islands.

2,597,318 people lived in the French Overseas Departments and Territories in January 2008 and none of them have nuclear power (also true of Corsica) and few hydroelectric stations I noted that the new tram-train stations on Reunion Island (France in the Indian Ocean) will all have solar panels.

Despite the mentality that Paris = France, may I suggest that there is a VERY large potential market for solar PV in the France that does not border Germany, Spain and Italy and largely burns oil for electricity.

Best Hopes for French Solar in Corsica, Reunion, Tahiti, the Caribbean, Guiana, and other areas of France,


#If oil is the primary alternative (some bagasse in Reunion I know, perhaps elsewhere) how cost effective is solar PV ? How much is needed for this population (+ Corsica population 260,196) ?

The most cost-effective panels convert about 10% of insolation into useful electricity, a factor that has remained unchanged for 10 years. Some PVs might convert 15%, but they cost more and are not mass-produced.

I think your data is out of date. Some of the most popular solar panels on sale today are by Kyocera. These are mass produced and easily purchased today. These panels are about 14% efficient. See (The cells are over 16% but if you do the calculation for the panel it works out about 14%.)

Another very popular brand is Sanyo. Their cell efficiency is now 19.7% and panel efficiency 17.2%. See

Another advancement has been in the inverters. Modern inverters achieve higher average efficiency by tracking the maximum power point and using more modern semiconductors. In the past you might easily loose 10%, now it is more like 5%.

You say that efficiency has remained unchanged over that last 10 years, but clearly this is not true.

You say that the cost has little changed since the year 2000. I believe most in the industry see cost continuing to fall, especially over the next couple of years as several new manufacturing plants come on line at the same time as credit is harder to arrange.

The above graph is taken (without permission) from here, where the source is quoted as Deutsche Bank.

Although I am just attacking some points in your article, I'm glad for the discussion. I take your point that solar does have limitations. My view is that we need every technology available to us. Solar will not be the answer to maintaining business as usual, but it does have the potential to be a serious low carbon energy source for homes, farms and most businesses.

France has led the world in maintaining its investment in the nuclear industry. I think we would all be well placed to emulate France in its energy policy, nuclear and solar.

You say that the cost has little changed since the year 2000.

I tend to watch this site for trends in solar PV prices:

Since 2005, prices have climbed slightly.

Actually, module prices have fallen considerably between 2000 and 2005. Since then, because supply could not satisfy the steep rising demand, prices have climbed again. However, the manufacturing costs have continued to fall. As the PV module shortage is finally dissolving, expect this year module prices to fall by 15% to 20%.

Also note that the PV module represents roughly only half of the total cost of a PV installation. Other costs (installation, inverter, ...) did continue their way down and more than compensated for the rise in PV module prices since 2005.

Do you have a link for 2000-2005 data, or a link that includes the total cost of the installation? I don't think I have ever run across anything like that.

I must admit that it isn't easy to find good statistics on the web.

First, take care with the prices from solarbuzz, as they are based solely on prices for a single module purchase, found on the web. Real purchase prices are clearly lower. Moreover, it was quite astonishing that, although the U.K. market in absolute terms is totally negligible in comparison with the rest of Europe, the recent devaluation of the British Pound had such an effect on the European Module Price Index. From (January 2009):

Price movement started throughout the PV chain during the Fourth Quarter, but it has yet to be fully reflected in retail prices through this sales channel. While there has been a dramatic move in the European index this month, this was actually a consequence of exchange rate movements (mainly Sterling/Euro) within Europe rather than a large number of price changes in the survey result.

Nevertheless, the statistics from Solarbuzz are still valuable, especially because they are updated monthly. Me too, I regularly check them.

For 2000-2005 data, IEA-PVPS could be a source:

Figure 10 on page 28 gives a nice graph with 1997-2007 module and system prices for Germany (country #1), USA (country #2) and Japan (country #3).

For trends since 2006, personally, I have more trust in some other statistics, like these from photonconsulting (module and system prices for 2006 and 2007, Table 1):

or from BSW-Solar (updated every quarter, in euro):

To end with, I would like to point to a rather old (2003), but still interesting presentation, saying something more about the importance of BOS Cost (Balance of System Cost = System Cost - Module Cost):

They were also correct expecting price increases because of silicon feedstock problems.

I absolutely disagree with your point about it being unlikely to build 5 400 MW of solar capacity in france by 2020... We are in 2009 and have about 11 years to that date.

That would be 490 MV each year.

Global production of PV cells in 2007 was about 3800 MW

In 2006 production was at 2 521 MW

I don't know the figure for 2008 but I expect it to be over 5000 MW.

There is one norwegian solar cell company that alone was producing silicon wafers in 2008 of in the range of 600 MW each year.

They have extensive plans to build more production capacity and estimate that they will produce close to 1800 MW/year of silicon wafers by 2010.

They currently sit on the most cost efficient production technology, but estimates that they in 2010 will cut production costs per watt by half of what was average industry cost in 2005.

They project their production costs for solar cells to be 1 euro per watt peak in 2012. You talk about the 6-7 dollar/watt market price of current solar systems.

So their estimated production costs will be much much lower than the current market price, wich will be a good thing since they expect there to be a supply glut and dramatic fall in prices of solar panels around that time wich will lead to failure of those solar cell manufacturers that can not produce as efficiently as they can.

Also the REC solar panels have higher conversion efficiency (12.4 - 14 %) than many other solar panels so you will not have to cover as much area. By using their best modules you only need to cover about 38 million sqare meters.Using their cheepest modules you would need to cover 43-44 million sqare meters. They are also working hard on improving conversion efficiency.

You cannot seriously claim that REC is not mass producing in cost efficient ways. You could argue that those who create polisilicon modules based with conversion efficiency closer to 20% are not as cost efficient.

By 2020 a lot of exiting new solar PV tecnology is expected to mature and we can hope to se panels with higher than 20% conversion efficiency produced at lower costs than traditional wafer based silicon solar cells.

You will get flexible thin film cells in all sorts of colours and levels of transparency for good building integration not only on roofs but on walls and windows. And you will get high efficiency concentrating (parabolic dishes) solar PV-systems using multijunction solar cells for use in desert PV-plants. (Currently multijunction cells have achieved over 40 % conversion efficency at high concentration and 31% efficiency under regular sunlight).

My conclusion:

France will easilly achieve a total of 5 400 MV of installed PV by 2020. It will not have to cover 54 million m2. If you go for desert like areas you can use consentrating PV wich has much higher conversion efficiencies. (30-40%) thus needing only a third or a fourth of the area suggested by Robert Rapier.

I would urge France do multiply their goal by a factor of ten and go for 54 000 MW instead of the puny 5 400 suggested in the plan.

As a comparison, have a look at Germany. At the end of 2007, the total installed PV capacity in Germany was 3800 MW, 1100 MW of which was installed in 2007 alone.

The Official numbers of installed capacities in Germany for the different renewable technologies for each year from 1990 till 2007, you find on slide 15:

There aren't numbers available yet for 2008, but I don't think it's improbable that 1600 MW or more could have been installed last year. So, already today, Germany might have achieved the number of 5400 MW of installed PV!

Though France has a smaller population (64 versus 82 million), it receives more insolation, making PV more attractive with grid parity expected before 2015.

I can't believe France would have close to 5400 MW installed by 2020. It will be a LOT more...

I'm surprised the French are even doing this. Areva has just acquired a large uranium mine in Niger and several of their EPR reactors are being built albeit with delays and cost overruns.

It would be great if each house roof supported say 2kw of panels coupled to a 20 kwh UPS with grid tie. That would avoid the need for new transmission especially sabotage prone HVDC cables from the Sahara desert. Unfortunately the cost would be even more astronomical. I note they are not talking about German style feed-in tariffs up to say 45c per kwh. A tax on the poor I say. Also in Europe I note the Spanish floods a few days ago. I wonder how well solar performed under those conditions.

Therefore I'm inclined to think the French solar program is an un-serious feelgood exercise knowing all along that nuclear will bail them out.

There are almost 3 million French citizens without access to ANY nuclear power. And almost all of these are in relatively low latitude areas of France.

Best Hopes for less oil fired electricity in Corsica, Reunion, Guiana, Martinique, etc.,


New nuclear cost goes up with time
New solar cost goes down with time

You might call a french solar programme a feel good exercise; I call it strategic policy. Porter theory effect thing.

You are also overestimating the extinction of solar radiation by the atmosphere. The official figue of 1000watts/M**2 is for what is called an airmass of 1.5 atmospheres. If theta is the angle of the sun above the horizon, this implies sin(theta) = 2/3 (for a sea level site), or an angle of 42 degrees. This implies an extinction factor of .2 per atmosphere.
The proper values for sealevel, as a function of solar angle follow:
degrees solar intensity watts/M**2
90 1105
75 1098
60 1072
45 1017
30 905
15 623
So even at relativly high lattitudes, the solar intensity isn't too bad. Also a surface tilted at the optimal angle towards the equator will do considerably better than a horizontal surface. Most panels will be tilted southward, and will gain considrably from that.

On the issue of worldwide panel production. The French goal of 5.4GWatts, is roughly the same as the current world annual production. Production has been doubling even two years (2009 is expected to fall behind due to the financial crisis, but is still supposed to be 25% greater than 2008). If this exponential continues, by 2020 production capacity will be roughly fifty times greater than today. And this says nothing about solar thermal!

Factors to consider about solar cells.

First: Watts of installed capacity needed to get one watt of continuous power.

Suppose 360 days/year. Suppose solar cell work 6 hours/day, 3 hours each before and after noon other hours of day having sun too high or too low. Suppose 1/6 of the working hours are rainy, clouded, too cold etc for it to work. Solar cells on average work 5 hours/day, 1800 hours/year.

One watt continuous power = 360 * 24 = 8640 hours/year.

8640/1800 = 4.8.

We need 4.8 watts of installed capacity to have one watt of continuous power.

Since no power plant work at 100% capacity and most work at 90% capacity therefore 90% capacity would be agreeable.

4.8 * 0.9 = 4.32.

One watt 90% continuous power = 4.32 watt solar cells installed.

Second Factor: Storage.

Energy produced by solar cells in sunny day time need to be stored somewhere (charging) and got back (de-charging) at non-sunny day times and night times. Suppose 90% efficiency for each of charging and de-charging, we get a combined efficiency of 81% as follows:

0.9 * 0.9 = 0.81

This increase the solar cells needed by the factor of reciprocal of 0.81:

4.32 * 1/0.81 = 5.33

Third Factor: Cost

For the solar cells: $3/watt (Least I can find on net)

For related equipments (batteries, wires etc): $0.75/watt (My guess, 20% of total cost)

Total Cost per watt of installed capacity = $3.75

Total Cost per 90% continuous watt = $3.75 * 5.33 = $20.

Fourth Factor: Return

Lets only look at the return in total gdp. 15 trillion world continuous watts means $48 trillion world gdp.

15 trillion continuous watts = 15 * 1/0.9 = 16.67 trillion 90% continuous watts.

$/watt = 48/16.67 = 2.88

Takes 6.94 years to get the money back in gdp.

Fifth Factor: Left Overs

Return for the energy producer. Very variable I think as it depend on fossil fuel prices.

Maintenance costs.

DC-to-AC conversion energy loss.

Industrial capacity to actually produce the desired number of solar cells.

Wisdom, that seems a very harsh way to run the numbers, for balance you could look at providing the electricity from the grid during peak demand in the middle of a sunny day when peaking plants have to be fired up due to a lack of cooling water at the nuclear plants. We are never going to supply all our power from solar, and PV will be a very useful wedge producing high value electricity with no moving parts. Enough solar 'anti load' could be added to a grid to cover the demands for refrigeration and air conditioning and reduce the need to build new power plants and the associated fuel supply and infrastructure.

France currently sells Switzerland nuclear power at 3 AM at "give away" prices and then buys Swiss hydropower at peak demand at, say, 5x what they sold it to them a half day before (saw some numbers from a financial report for a Swiss utility bragging about this TOD premium shift).

I do not have enough information on time of day wholesale pricing of power in the EU, but I am sure that a MWh of average solar will bring considerably more euros than an MWh of average nuke.

EDF may prefer to write checks to it's customers than to the Swiss utilities.


In view of this, the plan to add solar in France makes perfect sense..

More demand managment also makes sense for France. Valley filling and peak shaving. End use low temperature thermal storage for responsive loads (phase change H2O salt solution refrigeration, hot water for SH and DHW), accellerated adoption of plugin hybrids for G2V (and perhaps later on V2G as well) etc. France is doing next to nothing in these areas, while it makes perfect sense for such large amounts of continuous power sources like nuclear, as well as non-dispatchable sources with large diurnal variability (solar, wind).

Second Factor: Storage

Would batteries be needed for solar? Why would power generated during peak times be stored? I there is excess, sell it on the open market. If nothing else, send excess to Switzerland to backfeed hydro for $$$.

Yes WFP, you did a very well and concentrated basestudy on PVs here. Thanks. It falls into my own way of seeng those "things". PV's is a very hi-tech and difficult way of getting hold of KWhrs and Me, myself, I have serious troubles seeing them in vast quantities soome way down the Hubbert slope.
Till date I never saw a well founded EROEI analysis of PVs eighter ..

Any TODers around who know of any advanced EROEI-analysis for this tech ?

Do we know for certain that solar panals are the way they are going to go? Wouldn't thermal solar be better?, like they are doing in Spain.

More maps at SoDa.

This is the kind of discussion that keeps me coming back to TOD! :-)

Some points that came to mind as I read, in no particular order:

1. You guys always continue to impress with your knowledge and ability to find stats on the particulars of energy production and consumption all over the world, and in the EU in particular. Given that many folks do this here as a labor of love, one can only be impressed by the excellent study and research you do. I know I am envious and need to redouble my efforts (how many hours are there in a day, excluding time to work and sleep? :-)

2. While we in the U.S. study, the Germans and French seem to DO IT. This board proves that the problem is not one of lack of talent, but a problem of creating a support network for our talant to actually plan and build the projects. Where is the U.S. NREL, ConEdison, and the TVA on these discussions? We worry about getting the message to Obama, when we may need to be working the executive class of the major energy producers to get the message out.

3. There still seems to be a belief that solar panels, PV and thermal systems are about as good as they are going to get. Some of the statistics seem out of date on cost and conversion efficiency. The difference between 10% conversion and 14% may sould small to those not familiar, but it is HUGE in deciding the viability of many large scale solar programs, and if we assume that 18% to 20% is realistically possible in the next decade (at reasonable cost) it is believable that we are looking at the greatest growth industry in the world since the birthing years of petroleum and autos.

4. It is only correct that we recognize the steller efforts and acheivements of the French in reduction of carbon intensity in their economy. We should be sending students from MIT and CalTech on exchange programs to study what they are doing and what they are planning. For those who do not know, the French have always been on the cutting edge of technology since the days of the French Revolution, among the first to develop modern textile mills, automobiles and aviation. It seems they intend not to get left behind in the coming revolution.

5. And a revolution in energy is coming.


I'm French but living in the USA and I would like to provide a few more informations about this French solar program. I have followed theFrench news about and pushed my parents to install some solar panel (about 3 kW) since it is a good investment based on all the subsidies:
- There is a tax rebate covering 50% of the cost.
- There are additional rebates from cities/regional goverments (around 5-10% of the cost).
- The electricty generated is bought at 0.55 Euro per kW/h for individual and 0.45 Euro per kW/h for industrial ( and up to 0.57 is the panels are integarted in the roof structure). That price is warranted for three years and is tax exempt for installation under 3 kW.
Finally, the governement has promissed to simplify the paperwork to get these rebates since it was a major problem before.
The plan calls for a lot of large scale installation on roof of supermarkets, parking lots, factories, administrative buildings, farm outstructure...
Here is an recent example in Montpellier (southern France with about 2800 hours of sun per year)"
The text is in French but numbers are easy to read: Solar panel Efficiency 14.2%, 1,1 MWc of installed power and expected production 1,42 GWh/year.
The French administration received 22 application for large scale production between September 2007 and July 2008 for a total of 210 Mw of installed power. See
The French secretary of state Jean-Louis Borloo recently argued that now is a good time for France to invest in solar energy since the economical crisis has resulted in a glut of solar panels due to cancellation in other countries.
More recently, the French Riviera region has decided to invest further in local production of electricity (including solar) after a loss of power when the unique 400 kVolt line they rely on broke during a storm.
I think there is a good chance that France will reach its goal in solar power even though the cost might be higher than other energy source.
It is certainly a good idea to diversify you energy source and the extra safety bring by solar power might be worth the higher cost. France lacks any good sources of energy and relies on importation for almost all of it:
France import 95% of its oil and natural gas, has almost no coal left (last coal mine closed in 2004). Even uranium might be a problem since French uranium mines also closed in 2004 for economical reason (there might be some uranium left but I could not find data about these reserves). Most of French uranium comes from Tchad and nearby Niger...which explain why France army maintain over 8,000 soldiers there since the 80's (plus support troup and figther planes in nearby countries). I wonder if the energy cost of the French army should be counted in the calculation of the efficiency of nuclear power plants.

Just to add some data from Bavaria in Southern Germany:
I am operating a electrical solar (PV) system since 11 years now. It has a Peak Power of 3.2 kW and generates approx. 2650 kWh per year. It is connected directly to the grid so the full amount of power generated flows into the network (and is refunded with 0,52 Euro/kWh). Our annual electricity consumption (parents with 3 kids) is 1600 kWh so in average we are more than energy neutral. With regard to the peak demands of course we are not but since we are cooking with natural gas and using the hot water generated from our thermal solar system also for the washing machine and dishwasher (and the water bed ;-) ) we are coming closer to "own generation matches current electricity demand". Further more our local electricity supplier will soon install a smart meter for us so that we can see the demand-supply curve electronically and adapt our demand.
So on the one hand the french approach looks ambitious however if you cut down household demand and behaviour this is a step into the right direction. Of course I do not share the "nuclear" philosophy of France since all of us do not know where and how to keep the waste from the plants which are already existing. And this for the next 50 generations of mankind just for the waste one generation generated - this can't be right.

If you installed the same system in Tucson, you'll get 2.5X as much electricity. Generating solar power in metropolitan France is not going to be cheap. If they're willing to pay half a Euro for a kilowatt hour of electricity fine.

Even more than the 0,52 Euro per kWh: the smart meter will enable peak tariffs of 1 to 1,10 Euro (around noon time). Solar Electricity is ideal for demand-supply match. And this is honoured by appropriate tariffs.

You are right, the talent is here but we need feed-in-tariff legislation similar to what is in place in Germany in order to make these distibuted renewable projects work financially. A number of states are working on this but so far no feed in tariff programs have been put in place.

Low cost concentrating PV is on its way. By 2020, 5400MW in France would be easily achieveable. Forget about the flat panel, 10% conversion factor. It's old technology that no one will be using in 2020. With technology such as the Suncube with sun tracking and a 40% conversion factor. This uses Fresnel lenses to concentrate sunlight on to about 9 square centimeters of chip surface.


There is another reason for the diversification of power supply and that is that the French Government may have learnt some lessons from the drought and heatwave of 2003. This eliminated the hydro capacity of France and almost shut down the nuclear reactors on the Rhone, Garonne and other inland sites as the cooling water almost dissappeared. Exemptions were granted allowing water at 30C to be dumped back in the river. Because the drought & heatwave affected the rest of Europe they had difficulty getting alternative supplies. Solar is a perfect alternative for such conditions.

EDF (France's state electricity company) has stated that there will be no new nukes on rivers due to the overheating crisis. This means, de facto, only on the Atlantic coast. A VERY long term issue/problem perhaps.


I've been wondering what would be cheaper/more practical: to build all thermal plants adjacent to the Atlantic, and go for a bit more expensive (=longer) powerlines inland, or to have inland, slightly more expensive dry cooling but with lower transmission infrastructure requirements (shorter ties).

I was thinking dry cooling would be more useful, because of less transmission siting issues overall, and more potential for certain CHP uses inland (CHP would also reduce levelised cost of the dry cooling system).

Perhaps another reason to build Solar PV, in the south of France, Corsica and metropolitan France. The last two w/o nukes and some solar PV to help the grid of Southern France once most of the current nukes are retired.


One of the most important lessons was not to go cheap on cooling systems. Dry cooling, cooling ponds, once-through ocean cooling with diffusor pipe systems etc. Lots of good engineering choices. Just don't get penny wise pound foolish.

A few points to add to the excellent (as usual) debate and data provided:

(1) Nuclear power is a baseload power generation, France still has to deal with nationwide and regional peaks. They use a mix of buying/selling from other countries, etc plus hydroelectric power, otherwise they have to build such an overcapacity in NP that it becomes uneconomic. Being French they are thinking long term, global warming and gas production shortages may (quite possibly will) reduce those capacities, both within and external to France.

Solar will give them a peak capacity during the day, that will help them manage these constraints pretty economically. And who knows what can be achieved with power storage over the next decades.

(2) If they place the orders the plants will ramp up production to match. Plus, I suspect that the French will build their own industries as well, positioning themselves for a much larger future world market.

(3) There may be a mix of solar thermal in there as well. This is also a technology that will experience rapidly dropping costs as mass production ramps up.

(4) Be very careful with costings based on today's values, when you think in 20 year time horizens, the numbers change quite a bit. When France went all the way with nuclear they were similarly criticised ... not looking so stupid now compared to just about everyone else.

E.G Both PV and solar thermal are not even close to the costs that can be achieved in really large scale production. Halving of both IMHO should be readily attainable.

(5) I wish Australia had the same sort of sense. But our coal lobbies have killed solar (in any form) for decades at least, possibly forever.