The Bell Tolls for PV

– Guest post by Benjamin Gatti from our Solar Turbine Google Group.

This post features an eye-opening perspective on photovoltaic (PV) panels. The claim is: it is very unlikely that PV will ever be cheap. Personally, I’ve been following PV since 1995. I’ve heard predictions to the tune of “$1/watt PV panels coming out next year” since then. This has not yet materialized. My recent personal experience in self-made solar panels is that it cost us $2/watt in materials just to turn donated solar cells into panels – which is the same cost as we are aiming for in our complete Solar Turbine concentrator solar power (CSP)project. Moreover, my bottom line statement is: if it costs $2/square foot for mirrors, and $40/square foot for silicon PV material – I do not see any chance that solar cells will compete with concentrator solar power, if both yield approximately 10% efficiencies. This thinking stems from our discussions within, and motivation behind, the solar turbine project. Read the following guest post, discussion, and see if you can poke any holes in the arguments presented. We don’t have any final words here – just thoughtful perspective for worshippers of PV.

(PS. I must add that while I presently don’t believe that ‘solar panels will ever be cheap,’ this applies only to the ‘pure capitalist’ mode of production. I do believe that innovative, participatory, open source, community-supported manufacturing scenarios – supported by a mindset that values clean energy over war – can succeed in making affordable PV a reality.)

The Bell Tolls for PV

by Benjamin Gatti

Many governments, Germany Spain, the US etc… are holding out hope, in
the form of a technology-specific-subsidies, that Solar PV panels are
the world’s only best chance of energy independence. Too strongly stated
you argue, but then tell me – why are the subsidies for a watt of PV
electricity three times higher than the subsidies for a watt of wind
energy, or geothermal energy – if it isn’t because the governments have
placed a disproportionate emphasis on their importance?

Have a listen to the clanging dissonance. The price of PV will fall, we
are told, when the Silicon shortage is resolved. If you believe that
then answer this report: At the
current price of Silicon, there are some new plants going forward, and
some planned plants being canceled. This would suggest that the
viability of a new silicon plant – in the current price environment –
rests on a knife edge. The current price is way too high for economic
viability of PV, and yet too low for the viability of new silicon
production. That my friends is a dead end. Like any non-product, PV can
be kept artificially alive as long as it is kept on governmental
life-support, but it will never leave the ICU.

The kindest choice for PV is to let it pass in peace. What you ask – is
wrong with endless government socialism? after all, every other form of
energy is socialized energy – so why not PV? That is true of course,
when it comes to energy, we are all Maoists, we all genuflect at the
grave of Lenin, and worship at the feet of Stalin. The government has
been choosing winners in this field, and imposing Monopolies for
generation and distribution for years, and this is principally why the
energy infrastructure in the US is about as anachronistic and decrepit
as the power grids of Eastern Europe. So why not socialize PV? The
reason is that the benefits of PV are privatized – exclusively for the
richest – while the costs are not merely socialized, which would spread
them fairly, instead, they are regressively-socialized. The extra costs
of supporting net-metering are borne by everyone /other than/ those who
can afford to net-meter. Those costs include all the building and
operating costs of the grid and generator, all the costs of driving
around after storms and fixing broken power lines. Without the income
from net-metering subscribers, those costs must be borne by those who
can least afford to purchase PV panels. The poor have enough challenges,
we cannot morally justify raising the costs of their electricity and
food even if the ends have the appearance of “green energy”.

Surely we can agree that sustainable energy must satisfy a requirement
that its widespread use will not threaten the ability of the least
fortunate to sustain themselves. Corn ethanol does not pass this test,
neither unfortunately does PV. Wind appears to shine in this contest,
and it looks as though Solar thermal energy is an attractive candidate.


  1. Marcin


    I just read the link to the report. It says that production of
    polysilicon is skyrocketing. How does this fit in with your discussion
    that silicon production is ‘on an edge’?

    And what is your response to non-slicon technologies, like Nanosolar?

    I think that the real reason for failure of PV is that the entire
    process is too complicated. We just built our PV panels – and it cost
    us $2/watt to frame them from solar cells into panels. We are hoping
    that cost would be sufficient for the solar turbine option!

    Perhaps another point to mention is the bottom line cost of mirror ($2
    sq ft) is much less than PV ($40/sq ft) – for a product that ends up
    having similar efficiency, around 10%. This to me is the clincher –
    given that mirrors themselves, in our open source mass production
    scenario, are the bulk of the cost.

    I;d like to iron out these issues before I post your note.


  2. Marcin

    Response from Ben:

    If the report had said that new silicon production was at “Full Speed Ahead” we could expect silicon prices to keep coming down; instead the report cites some forward movement, and some braking movement as planned facilities are scuttled. It costs a lot of money to pan a production plant, these companies have made the calculation to throw out that investment rather than to go forward – meaning they no longer see the plant as profitable. This means that the business case for more Silicon production has reached “neutral” at the current price – which makes it very hard to argue that at future lower prices, there will continue to be enthusiasm for investing in more silicon production. The pull back suggest a point-of-diminishing returns.

    The PV market is hugely dependent on State Subsidies, and unlike natural demand, artificial demand can disappear at the stroke of a pen. This has investors worried of course, and it’s hard to imagine that governments can budget increased subsidy payments at the rate that would be required in order for PV to reduce our dependency on oil for example.

    On Nanosolar,

    Simple really, Silicon is only about 25% of the cost of PV, so the maximum benefit of completely free PV cells is a 25% gain; Nanosolar might cut this in half, so a 10% improvement might be expected. Unfortunatly, the panels are less effecienct, so one needs more backing material, more structure, more labor, more wiring, and more washing to generate the same amount of energy. I’m afraid in the end this will look like one step forward, and two steps backward.

  3. Marcin

    Agreed, but silicon is 25% of cost of SILICON PV. Is it foreseeable
    that the Nanosolar or some other version may also print the wiring and
    print the encapsulant, to make it really cheap. Do you see this as a
    possibility, or a pipe dream?


  4. Marcin

    Response from Ben:

    Unfortunately, every outdoor technology has to deal with the question, not merely of wind, but of worst-case winds. That means some structure with a reasonable chance of surviving a 10 year storm. It’s doubtful that these structures can be made for some very low cost. The current price of PV is nowhere close to where it needs to be to address climate and independence goals. Were it at 2x, the odds of success would be much higher, unfortunately, its not that close, and the cost of incremental improvement is very high, and very slow.
    There is more VC activity in Solar Thermal than there is in Thin Film lately, if I understand the news reports correctly, and in particular, the ability of CSP to store energy and deliver electricty into the shifted peak demand.

  5. Marcin

    From Vinay:

    Ben, with all due respect, you appear not to have any idea what you’re talking about.

    Right now, Nanosolar’s cost of manufacturing is 30 cents per watt, with retail of finished panels suitable for long term outdoor use in municipal power applications at $1 per watt. They’re producing something like 1 gigawatt of panels per year right now from two factories, one in the US, one in Germany.

    So, yeah, PV is far from dead. Old school PV is almost certainly dead, but everything points to thin film PV replacing it, at a quarter or less of the cost.


  6. Marcin

    From Ben:

    Every two weeks there is a “breakthrough” technology for PV. MIT released one fairly desperate solar technology last month in the form of dyed plastic. With respect to Nanosolar, it’s been fairly difficult to pin down the facts; what we know is they are treating their panels like McCain is treating Palin – by hiding the product from the prying questions of reporters. You claim Nano has a product for 30 cents a watt or a dollar a watt. If that were even remotely true (and that product were stable for at least 7 years) , one wouldn’t be blamed for thinking it would stop all investments in any competing technology be it wind, CPV, NG, Clean Coal, etc…

    Let’s take the facts we have and propose the simplist explaination that fits:
    1. Nanosolar is hiding their technology – with big claims on cost – any claims on longevity?
    2. Germany has huge subsidies for Solar – which may not require rigorous proof of longevity.
    3. Nanosolar is in a position to profit from the longevity loophole in Germany.
    4. Nanosolar is keeping performance on the down-low is this because they doesn’t want Germany to realize they are funding a boondoggle?

    5. While we know little about Nano, there are other companies such as Daystar using the same technologies.
    6. Every other PV maker makes their product available to the highest bidder.

    My occam’s razer conclusion is that Nano-solar has a niche market in German subsidies, they are doing large scale plants because the only customer is the German tax payer, and the Germans haven’t bothered to see if the product is the real deal.

    The narrative that Nanosolar has a product which the market would demand at $3 a watt but which they agree to sell for $1 to one buyer because they like that buyer stretches credulity.

    I admit, I haven’t done as much digging on Nano lately as when they first announced, maybe you have more information? But generally, I’m not buying the Camelot narrative – and I don’t think the market is either.

    Here’s something to think about – we have been working on sustainable and strategically secure energies since 1970, we are more reliant, and less sustainable today than we were 10, 20, or 30 years ago. During that time, PV has played the center role in clean energy for most people. It turns out to be a bit of a goosehunt. I respect the group of people who continue to trust and believe in the second coming of PV, but I guess in this sense, I’m just not a true believer, and I don’t see any holdable evidence to the contrary.

    For one thing we have Solar-Mirror Technology. As Marcin said Mirrors are pennies per watt compared to solar, and what is more mirror systems can include storage for a much lower price. When you factor in the advantages of Solar-Mirror technology over scarce element processes, It seems not worth the effort of trying to solve the very challenging limitation of PV.

    Here is a recent article:

    Note the end game, also note the insight that all of these panels will be fungible on day one. That is a very challenging business prospect to have zero product loyalty right out of the gate. It is my view that PV doesn’t look very healthy, the bell is tolling, it’s time to bury the body, because it stinketh, and the smell is permeating more healthy technologies.

    But I appreciate other well-defended points of view – especially if those points of view include links to authorative sources.


  7. Marcin

    From Vinay:

    Nanosolar claims 25 years.

    Your points are valid, and we’ll see what actually happens.


  8. Marcin

    From Josef Davies-Coates:

    Re: (Unfortunately, every outdoor technology has to deal with the question, not merely of wind, but of worst-case winds. That means some structure with a reasonable chance of surviving a 10 year storm. It’s doubtful that these structures can be made for some very low cost.)

    I may well be being very silly but what about velcro?

    (velcro straps are strong enough to hold two speed boats going in different directions together – I saw it on TV! 😛 ).


  9. Brent

    Hi there, I’ve been following along with the project for a while (yes a bit of a lurker) and please allow me to put my two cents in. Traditional PV solar’s achilles heel besides the cost of silicon is production time. The process of having to grow amorphous silicon wafers is slow and very finnicky to environmental variables, by all reports. This involves using a batch processing method, and having high failure rates per unit. The differences that allow nanosolar to so dramatically reduce the price per installed watt on their cells is more than just the removal of the silicon from the cell, because most importantly the removal allows for other production methods to be used. The difference in price comes mainly through the efficiencies of scale that are achievable by using a continuous production method. In nanosolars case they use a ink and simply print out their cells. So yes government subsidy plays it’s part in lowering the market cost of PV cells, but with silicon becoming decreasingly relevant to the prodcution of said PV cells, different materials are allowing for far more efficient production per watt installed. Keep up the good work!

  10. Abe

    Here’s a possible steam engine. Seems simple to build, I don’t know about cpoying the design:

    Abe- I was really excited about the Green Steam engine two years ago. I got the kit and I was going to build one. Then I did further diligence on peoples’ experience with the engine. I talked to the inventor several times, and asked for references of people who have built the engine. He declined, for reasons of privacy. So I was frustrated, and thought that if the inventor cannot provide a solid reference to others who have produced the engine, then the engine performance is suspect. I never did follow up on the Green Steam engine, and lost interest in it. Tom Kimmel of Kimmel Steam Power ( told me that he finds so many faults in the engine that he does not know where to start. Now this is hearsay evidence. If anyone in the audience has built the engine, or knows people who have – and was satisfied with the results, I’m interested in talking to them. – Marcin

  11. Kiashu

    The way I see it, fossil fuel and general resource depletion are going to ensure that, whatever the technology, PV will remain relatively expensive. If some technology makes it much easier to produce in great volumes, then the demand for it will go up – since currently demand is limited by waiting times – and that’ll push the price up. As the final consumer, whether it’s (say) $25,000 for a 3kW system because the PV itself cost $10,000 to make, or because it cost $1,000 to make but everyone wants it, it makes no difference. It still cost me twenty-five grand.

    However, the same things which push the price of PV up will push the price of other power sources up. Fossil fuel and general resource depletion, and any attempt to reduce greenhouse gas emissions, these will all act to push up the price of energy from other sources. So while the price of PV won’t drop, the price of other power sources are likely to go up – PV then becomes the best of a bad lot!

    However, I fully support Factor E Farm’s choice of solar thermal over PV. As I see it, many of the solutions to resource depletion and climate change being offered are solutions for the prosperous West only. A $100,000 electric car seating at most 5 but usually 1.5 is a transport solution for at most 1% of the world’s population. A $100,000 bus seating at most 50 but usually 25 is a transport solution for around 50% of the world’s population.

    With the view to providing solutions which the most people can get to, solar thermal looks better than PV, since it requires less resources and technology, less infrastructure and skills. Solar reflectors can be made from a large variety of metals; PV can be made only from silicon, and good PV requires doping with rarer elements like germanium. So that solar thermal is a more democratic technology, it can be produced and used by a wider section of the world’s population than PV.

  12. Chriswaterguy

    Re Vinay’s comments: I had a chance to talk (at a barbeque) with a solar researcher about to start her PhD at MIT. She said that thin film solar is not yet as cheap as silicon.

    Figures like 30c per watt are wonderful and I hope they’re right, but do you have a concrete source that says they are available for that right now?

    But I agree with Marcin’s questioning of the maximum-25%-improvement idea. The greatest potential in thin film solar comes from not needing the expensive backing, and possibly being laid down straight onto the surfaces of products or buildings.

  13. Benjamin Gatti

    Chriswater suggests laying solar stuff on buildings.

    Let’s give that a smell test.

    There is today a great substance to lay down on buildings in sunny climates – BWP aka “Bright White Paint”. BWP is a special substance that saves Cooling energy during peak energy hours, BWP requires no wires, no inverter, BWP can be pre-applied to building materials, increases albedo, decreases urban heat islanding, and BWP costs mere pennies per kilowatthour saved and requires no maintenance.

    So, instead of BWP, you’re suggesting TOS (Thin organic Solar). TOS Solar is black and absorbs a huge amount of Solar energy (say 90%), converts 10% of that to electricity while sending 80% into either the immediate atmosphere or the building. Since the building is actively cooled, and the outside temperatures are higher, a majority of the surface heat may be absorbed by the building. Being fair, if only half of the energy is absorbed into the building, that’s 40%. We have 10% as electricity, so we run an air-conditioner, say 75% efficient. So net thermal enthalpy is 40% – (10% * 75%) or 33% of Solar incidental energy, BWP, on the other hand may reflect 90%, net enthalpy ~10%, cost about .01% the cost of Solar.

    Yeah, sign me the f*@& Up.

    Hint: If you don’t have an air-gap under your solar panel, you’re plain crazy. That pretty much rules out any role for BIPV in my opinion. Your opinion may vary.

  14. Chriswaterguy

    I wrote a long answer, then FF3 crashed again.

    BIPV = Building integrated photovoltaics – it’s helpful to spell out the acronym the first time.

    A lot of good points, and my sanguine picture is obviously inadequate.

    Still, these kind of solar cells could potentially be a lot cheaper – support would be simpler, and I’m hopeful an imaginative soul might find a way to use *some* existing structures. Maybe even find a way to combine it with solar thermal. I’d like to see what people are planning to do.