Solar technology roadmap

February 19, 2007 7:42 AM

Cheap solar power poised to undercut oil and gas. The "tipping point" will arrive when the capital cost of solar power falls below $1 per watt, roughly the cost of carbon power.
posted by stbalbach (88 comments total)

We've heard this kind of thing before. And like always, it's a transparent attempt by this "startup" to attract investors with rosey claims about future success.

Anil Sethi, the chief executive of the Swiss start-up company Flisom, says he looks forward to the day - not so far off - when entire cities in America and Europe generate their heating, lighting and air-conditioning needs from solar films on buildings with enough left over to feed a surplus back into the grid.

...except at night.
posted by Steven C. Den Beste at 7:52 AM on February 19, 2007

Need more nuclear plants.
posted by tadellin at 7:59 AM on February 19, 2007

Please say 'hydrocarbon' or 'carbon dioxide' rather then just "carbon dioxide", depending on what you mean.
posted by delmoi at 8:00 AM on February 19, 2007

...except at night.

Unless they can generate what they during the day, store it, and use it at night.
posted by delmoi at 8:00 AM on February 19, 2007

Unless they can generate what they during the day, store it, and use it at night.

That is the other big energy challenge. Electricity doesn't store well.
posted by b1tr0t at 8:09 AM on February 19, 2007

Laugh all you want, but solar has been making great strides for the last ten years. China is heavily invested. And guess what? Solar works. The problems are the cost and the attitude of naysayers that don't understand that solar is not meant to replace central energy generation. Solar is decentralized and can fill a substantial niche. Remember, whatever power solar generates is power that doesn't rely on hydrocarbons.
posted by Benny Andajetz at 8:12 AM on February 19, 2007

Electricity doesn't store well.

Electricity stores better than it travels. Central power generation is a dinosaur.
posted by Benny Andajetz at 8:14 AM on February 19, 2007

That is the other big energy challenge. Electricity doesn't store well.

Nor does it transport well, but we seem to have gotten along with that for a while now. In fact, storing energy is much more efficient than shipping it over long distances, in terms of power loss, it's just more expensive financially. It seems like sufficient battery/capacitor arrays might be largely a startup cost, though, that averages out over time if solar generation really becomes more efficient. Remains to be seen, though, I guess.

Anyway, the lion's share of power is used during the day, so even granting SCdB's comment, it's more of a caveat than a gotcha: solar will obviously not completely replace fossil fuels anytime soon, but with just a modest marginal increase in efficiency, it could seriously reduce their use. Sounds like a victory to me.
posted by rkent at 8:17 AM on February 19, 2007

That is the other big energy challenge. Electricity doesn't store well.

Pump water up behind a dam in the day time. Go hydroelectric at night.
posted by Mr_Zero at 8:17 AM on February 19, 2007 [1 favorite]

Benny Andajetz is right on this one: solar power alone might not be able to address 100% of power needs at the moment, but it could alleviate the environmental burden. It may not be the solution in and of itself, but it doesn't hurt and probably helps, and if cheap solar can be developed, more power to the people developing it.
posted by graymouser at 8:18 AM on February 19, 2007

Remember, whatever power solar generates is power that doesn't rely on hydrocarbons.

Only if it is net energy positive, that is, more energy comes out over the lifetime of the energy cell than was used to produce it.

That having been said, this does sound promising. It could just be a startup scam as SCDB suggests.

If I understand the article correctly, this technology is not a radical change in solar cell efficiency. Instead, it is largely a production/manufacturing change that could potentially make our current low-efficiency cells viable on a per KWH basis.

You still have problems of

(1) electricity storage
(2) finding enough surface area for all these panels
posted by b1tr0t at 8:19 AM on February 19, 2007

Pump water up behind a dam in the day time. Go hydroelectric at night.

Ditto for producing hydogen during the day, store it in tanks, use it at night.
posted by gwint at 8:19 AM on February 19, 2007

Oh, and SCdB: We've heard this kind of thing before. And like always, it's a transparent attempt by this "startup" to attract investors with rosey claims about future success.

Way to ignore the rest of the article, where the CEO of Applied Materials, hardly a VC-craving startup, basically concurs that the industry will reach something like $1 per watt within 2 years.
posted by rkent at 8:21 AM on February 19, 2007

Applied Materials makes the machines that you use to make semiconductors. Asking them if solar cells are going to be competitive in a few years is like asking Cisco if the internet needs more capacity.
posted by b1tr0t at 8:26 AM on February 19, 2007

As far off in the future as it may be, I am fascinated by what the macroeconomic effect of plentiful energy at a low cost of production would be. (Keeping things light, I think of these of problems of “Star Trek” economics) Assume that after the initial capitol expenditure that the cost of maintenance would stabilize and decrease slightly over time as cost saving tech improved and the business model matured. Prior to maturity, might we see artificial market limits imposed by a cartel type system? Later, could the cost of energy decline so far and remain so stable that it was not longer a macroeconomic variable? I love speculating about the possible effects of a nearly global decrease in costs. Could everything get a little bit better for everyone?
posted by BeerGrin at 8:26 AM on February 19, 2007

Ditto for producing hydogen during the day, store it in tanks, use it at night.

The Hopewell Project - a solar-hydrogen house in New Jersey.
posted by stbalbach at 8:31 AM on February 19, 2007

Mike Splinter, chief executive of the US semiconductor group Applied Materials, told me his company is two years away from a solar product that reaches the magic level of $1 a watt.
posted by stbalbach at 8:34 AM on February 19, 2007

The above is a quote from the article, in response to b1tr0t's comment.
posted by stbalbach at 8:35 AM on February 19, 2007

I love speculating about the possible effects of a nearly global decrease in costs.

This has happened rather steadily throughout the Industrial to Modern age. Energy efficacy has gone up while cheaper, plentiful hydrocarbon deposits continued to rise. The net effect is that energy became very cheap. As we have seen, the cheaper energy becomes the more uses we have for it. No longer are we burning coal or wood to make nails, but we do so for all sorts of tasks that were previously cheaper to do by other (usually human) means. I do not believe we will ever have enough energy, but rather we wait for inventors and entrepreneurs to catch up with energy advances. If there is one constant throughout history, it is that energy will be used if it is there.
posted by geoff. at 8:35 AM on February 19, 2007

bitrot: First of all, I was specifically responding to SCdB's "silly startups" comment. But beyond that... what? You can never ask technology manufacturers what the state of technology is and how it's advancing? Whose word would you trust? Furthermore, the trend in cost-per-Watt in solar technology has been exponentially decreasing for at least a decade in some markets, is it really so incredible for industry figures to predict a continuation of that trend?
posted by rkent at 8:37 AM on February 19, 2007 [1 favorite]

Electricity stores better than it travels. Central power generation is a dinosaur.

How do you get solar generated power to places that don't get much sun then? Speaking as someone who lives in one of the cloudiest cities in the US.
posted by octothorpe at 8:37 AM on February 19, 2007

Prior to maturity, might we see artificial market limits imposed by a cartel type system? Later, could the cost of energy decline so far and remain so stable that it was not longer a macroeconomic variable?

There already is a cartel: the oil suppliers and energy producers. Imagine where they would be if electricity were generated on a saner neighborhood- or single house-generated scale?
posted by Benny Andajetz at 8:42 AM on February 19, 2007

As far off in the future as it may be, I am fascinated by what the macroeconomic effect of plentiful energy at a low cost of production would be.

At best, you would have an energy network that looks like long haul telecom in the United States. It doesn't cost anything to produce bits, but it does cost money to install and maintain the infrastructure to transmit them. When you write a bandwidth contract, you are buying capacity.

Electricity won't be quite that good because you can't make the electron->photon jump that telecom made. You need big transmission lines to carry massive amounts of current from point to point. You lose a lot of power over large distances. With data, you can squeeze more TB of data onto the same fiber pair by installing more precise lasers. Optical signals don't induce signals in adjacent transmission lines, so you can pack a dozen high bandwidth fiber pairs in a single small conduit.

If we do get to a point where we are only paying for electricity transmission, and not generation, then the next issue would be environmental. If power is free, why not line every sidewalk and every road with heating elements so that they never ice up? What would the environmental impact of so much heat be? In a solar only economy, it shouldn't be much worse than just redistributing the sun's heat. If you add in nuclear power, you could throw a lot of heat directly into the biosphere.
posted by b1tr0t at 8:43 AM on February 19, 2007

bitrot: First of all, I was specifically responding to SCdB's "silly startups" comment. But beyond that... what? You can never ask technology manufacturers what the state of technology is and how it's advancing? Whose word would you trust?

You can certainly ask them, but you should also ask some independent sources. How about a few solid-state physicists, a supply chain manager, and a few engineers that specialize in building production lines similar to what this startup would need?

AMAT wants to sell more semiconductor fab equipment. That is great, but I'd like to have a second opinion.
posted by b1tr0t at 8:47 AM on February 19, 2007

Pump water up behind a dam in the day time. Go hydroelectric at night.

Just don't forget to turn off the pumps when it's full.
posted by zsazsa at 8:48 AM on February 19, 2007 [1 favorite]

can someone point to an up to date review of solar energy which has been published in a peer reviewed scientific journal?
posted by paradroid at 8:49 AM on February 19, 2007

How do you get solar generated power to places that don't get much sun then? Speaking as someone who lives in one of the cloudiest cities in the US.

At least for right now, you're thinking too big. Solar is an answer, not the answer. It should be utilized when and where it will work - which is most places, to some degree.
posted by Benny Andajetz at 8:50 AM on February 19, 2007

Pump water up behind a dam in the day time. Go hydroelectric at night.

You lose 20% of your power and the pumped storage facility has very high upfront capaital cost, which is a (if not the) key reason that renewables aren't competitive.

Hydrogen isn't even worth thinking about at the moment, about 50% efficient, the current evidence is you'd be better off with electrical cars than hydrogen powered cars, and until you have high penetration of renewables then you might as well focus on displacing fossil fuel directly at the time of generation.

Solar will be a lot more competitive at $1/W but that's not going to happen very soon, I'd agree with SCDB 's assessment that these guys are trying to pump up interest in their company.
posted by biffa at 8:52 AM on February 19, 2007

A little off topic: We are talking about active solar here. Passive solar gains can be had for nothing, if needs are addressed prior to construction. South facing windowwalls, heat sinks. roof overhangs, and natural windbreaks can all put serious dents in existing energy usage. Why these issues don't get serious consideration blows my mind.
posted by Benny Andajetz at 8:54 AM on February 19, 2007

"There already is a cartel: the oil suppliers and energy producers. Imagine where they would be if electricity were generated on a saner neighborhood- or single house-generated scale?"
posted by Benny Andajetz

I'm aware of OPEC Benny. I was not attacking the solar idea.

As far as imagining where they would be...I'd rather not think about the loss of human life that will proably occur as the easy money the props up Oil-State Governments goes away.
posted by BeerGrin at 8:55 AM on February 19, 2007

Not OPEC - domestic energy providers.
posted by Benny Andajetz at 8:57 AM on February 19, 2007

I'd agree with SCDB 's assessment that these guys are trying to pump up interest in their company.

Notwithstanding the fact that major industry players agree with them and both parties' predictions are consistent with a decade's worth of price-trend data.
posted by rkent at 8:58 AM on February 19, 2007

How do you get solar generated power to places that don't get much sun then?

Mirrors, dude. Big, big fuckin' mirrors.
posted by From Bklyn at 9:08 AM on February 19, 2007

I'll never understand the outright dismissal of solar that consistently comes from some circles. Just think about the billions in government money that was poured into nuclear power during its first 50 years. If just a fraction of that amount was to get directed towards solar solutions, most of the difficulties mentioned by the poo-pooers would be easily overcome. Of course, decentralized power doesn't make centralized companies much money, so...
posted by mediareport at 9:10 AM on February 19, 2007

I'll never understand the outright dismissal of solar that consistently comes from some circles

Usable, economic solar power would be great, but this is a speculative startup. Solar costs may be driven down to a level where the economics make sense for broad use of solar, but they may not.

If just a fraction of that amount was to get directed towards solar solutions, most of the difficulties mentioned by the poo-pooers would be easily overcome.

Billions have been spent on improving semiconductor manufacturing. Applied Material's primary business is supplying semiconductor fab equipment to companies like Intel. It would be interesting to crunch the numbers to find out if more has been spent developing nuclear or semiconductor technology.

You can't just throw more money at a problem and expect a solution to pop out. This proposed technology appears to be in the manufacturing space rather than the semiconductor space. All the money in the world thrown at semiconductor physics would not have stumbled across this solution.
posted by b1tr0t at 9:19 AM on February 19, 2007

I'll never understand the outright dismissal of solar that consistently comes from some circles.

I agree. I think integrating as many energy sources as we can should be the goal. Every home should have passive and active solar, green roofing, sumping systems and bio fuel capacity. Why the hell not? The more we start dong it the cheaper it becomes and the more we can work out the bugs.
posted by tkchrist at 9:22 AM on February 19, 2007

The dismissal comes from people who are afraid of change. It's the only explanation I can muster. What other skin comes off their asses if solar turns out to work well?
posted by five fresh fish at 9:30 AM on February 19, 2007

I envision an earth a billion years in the future, when the great race of Yith sees a dying sun and and speculates that the price of burning those fossilized remains of the human race might be more economical than relying on their immense solar-powered network.

Not really relevant at all; just thought i'd share. :)
posted by mr_book at 9:32 AM on February 19, 2007 [1 favorite]

How mirrors can light up the world using concentrated solar power plants, making deserts sustainable powerhouses (pdf).
posted by adamvasco at 9:44 AM on February 19, 2007

If this shit is so light and cheap, maybe this means I can have my powersats after all. It would be totally awesome if my last job before I retired in 2040 were in an O'Neill colony.
posted by ROU_Xenophobe at 9:46 AM on February 19, 2007

Does this mean we can bury biofuels now?
posted by DenOfSizer at 9:56 AM on February 19, 2007

Interestingly in my own personal research it seems that Solar Panels in their current state are mostly viable already.

Putting in a 5 KW array would cost me about 30K with a 20-30 year lifespan. Over the life of the array, I'd end up about even with power company.

Problem is a 5KW array is pretty big and I'd really like to put in a 10KW array which is even larger and my city home in Seattle just doesn't have enough yard space to do this.

Now the day that roofing and potentially window materials can be made to integrate into a whole house PV system then I'll have plenty of space (my whole roof and lots of south facing windows) to cover and collect solar energy. That's what folks think may be available in the next 2-10 years, more flexibility in where and how you mount your arrays. In windows, roofs even embedded in siding.
posted by aaronscool at 10:06 AM on February 19, 2007

How do you get solar generated power to places that don't get much sun then? Speaking as someone who lives in one of the cloudiest cities in the US.

RTFA: "It'll even work on a cold, grey, cloudy day in England, which still produces 25pc to 30pc of the optimal light level. That is enough, if you cover half the roof," he said.
posted by afx114 at 10:19 AM on February 19, 2007

rkent: Thanks for the link. I've often wondered about it but have been too lazy to look it up!
posted by ZenMasterThis at 10:22 AM on February 19, 2007

Pump water up behind a dam in the day time. Go hydroelectric at night.

Ditto for producing hydogen during the day, store it in tanks, use it at night.

What biffa said. The requirement for energy storage will more than double your $1/W capital cost.

I'll never understand the optimistic ignorance that consistently accompanies discussions of photovoltaic power. The levelized unit energy cost of PV electricity is 2 to 10 times fossil fuelled electricity. (We can debate the source of those numbers if you like, but the range is consistent with other estimates). If you are willing to pay 5 times more for electricity, then by all means, lobby your representative to outlaw fossil fuels. But I doubt you'll convince many of your neighbors to sign on.
posted by Popular Ethics at 10:27 AM on February 19, 2007

I envision an earth a billion years in the future, when the great race of Yith sees a dying sun and and speculates that the price of burning those fossilized remains of the human race might be more economical than relying on their immense solar-powered network.

I doubt either will survive the suns expansion into a red giant before it colapses to a white dwarf.
posted by delmoi at 10:30 AM on February 19, 2007

I'll never understand the optimistic ignorance that consistently accompanies discussions of photovoltaic power. The levelized unit energy cost of PV electricity is 2 to 10 times fossil fuelled electricity.

How about the cost when the cost to remove that CO₂ from the atmosphere is figured in?
posted by delmoi at 10:31 AM on February 19, 2007

How about the cost when the cost to remove that CO2 from the atmosphere is figured in?

A very good question, one that lots of economists are looking at right now.

Like I said, you can make solar power competitive by outlawing fossil fuel emissions, but you would have to accept energy bills 2 to 5x higher than what you currently pay. I would honestly vote for that, but I doubt many others would.
posted by Popular Ethics at 10:44 AM on February 19, 2007

I'll never understand the optimistic ignorance that consistently accompanies discussions of photovoltaic power

PV frees fossil hydrocarbon stocks for other productive uses.

20 years ago 16Mhz CPUs were the shit, now thanks to free-market throwing billions into the pursuit of R&D we've got CPUs 100x faster; is there any reason not to suspect a similar yield curve would apply to PVs, should the free market enterprise work its magics at similar scales of investment?

The reason PV is only-breakeven with PG&E today is that the PV suppliers, thanks to a favorable supply/demand curve, are free to jack up their prices such that the capital investment in PV tracks the monthly cost of a hookup.

We humans are trained from birth that energy is needfully Expensive, that the nth Law of Thermodynamics means there is no free lunch, etc etc. This is BS. We live in a world of megajoules nearly free for the taking -- from the latent heat of the earth, from the 1KW/sqm insolation, from the wind & waves, should we only spend the time & brains to develop the right apparati to harness these energy sources.
posted by Heywood Mogroot at 10:52 AM on February 19, 2007

How are the solar cells described in the article manufactured? What of other solar solutions?

While they may reduce dependance on oil, I wonder if their net environmental impact is really much different. What sort of mining operations are required to obtain the materials necessary for their production? What are the byproducts of their manufacture?

I'm not in a postion to answer these questions, obviously, but I've yet to read a straight, cogent and well researched response to them. This article, like every other I've read, doesn't raise the issue at all.

Will the manufacture of photoelectric cells and massive batteries required for solar power result in a smaller environmental footprint or are we just trading one brand of pollution for another?
posted by aladfar at 10:54 AM on February 19, 2007

Heywood: I admire your optimism, but I don't share it. Renewable energy has never suffered from lack of investment. Wouldn't you want to be the company that comes up with free power? The fact is that electricity is, and always will be, many (many) times easier to extract from fossil fuels than any of the sources you cite.

Meanwhile, that "trust in technology" attitude keeps people from making real choices about the environmental cost of their lifestyle. Why tax carbon when free solar power is just around the corner?
posted by Popular Ethics at 11:07 AM on February 19, 2007

aladfar: I found these articles when I was researching my comments. (I don't know if either was peer-reviewed). They suggest that the lifecycle CO2 emissions of PV power is comparable to nuclear power, which is to say, not nothing, but not much.
posted by Popular Ethics at 11:13 AM on February 19, 2007

I went and checked my figures and it seems likely I was wrong to challenge the $1/W figure. Sloppy thinking on my part. Where I should have quibbled is with what this will translate to cheap enough energy to achieve some 'tipping point' in comparison with other generating technologies. Even at these prices for capacity, the capacity factor of solar in many locations is likely to mean very high unit costs per energy, which will mean they will have high payback periods and remain unattractive investments except in niche applications. The scale of cost-appropriate niches will increase over time but this is going to be longer than the next 2 years.

What biffa said. The requirement for energy storage will more than double your $1/W capital cost.

My point was that renewable technologies aren't very economic if they require storage, to displace applications that require transportable energy other than electricity they will require storage which is currently hopelessly ineffiicient. As a result, hydrogen conversion would be ridiculous until such a point as renewables have achieved such deployment that they have displaced a very high fraction of electrical generation from fossil fuels. This isn't to say I don't welcome renewables: considerable fuss is made over intermittency, when its now pretty clear that fluctuations in consumer electrical demand far outstrip fluctuations in availability of, for example, wind, and that penetration of wind to at least 10% of all generation and maybe as high as 20% can be done with minimal additional cost on a large scale distribution/transmission grid. (It's also quite possible that more distributed generation will lead to reductions in losses in transmission and transformation.) Storage isn't a necessity until we have a lot more renewable energy.

Incidentally, to address a point made earlier in the thread transport of electricity and transport of hydrogen are not hugely different in terms of energy losses over distance.

Now the day that roofing ...materials can be made to integrate into a whole house PV system

These have been on the market for ages, they're not cheap but they are commercially available. Solarcentury's c21e for example.

fff:The dismissal comes from people who are afraid of change. It's the only explanation I can muster. What other skin comes off their asses if solar turns out to work well?

I don't like the overly optimistic figures because when they don't deliver its to the future detriment of the technology. Microgeneration is currently taking off in the UK but I have a real fear that poor quality equipment is going to give the technology such a poor reputation that it will damage efforts to increase its use for decades into the future.
posted by biffa at 11:15 AM on February 19, 2007

aladfar: a search for life cycle energy use PV on google scholar gives a few peer reviewed articles on PV cell manufacture.
posted by biffa at 11:17 AM on February 19, 2007

The levelized unit energy cost of PV electricity is 2 to 10 times fossil fuelled electricity.

Use of such figures as predictions (rather than as history) necessarily involves some assumptions about the price of fossil fuel. Which unfortunately appears to be getting more volatile and harder to accurately predict.
posted by Western Infidels at 11:23 AM on February 19, 2007

Random thoughts:

Some of the corporations manufacturing solar cells: BP, Shell, Kyocera, Seiko, Matsushita, Canon. And these are just the big boys. Do you know something they don't?

I think most of you are still thinking too big. Local power generation is the future. The efficiency of running neighborhood-sized power plant vs. a city-sized power plant is substantial.

As far as cost: PV is getting cheaper. Oil, more than likely, will only get more expensive- making PV rapidly more attractive.

If anyone is interested in a saner, more holistic, approach to energy and energy conservation, the web site for Amory Lovins and the Rocky Mountain Institute is a good place to start.
posted by Benny Andajetz at 11:37 AM on February 19, 2007

...except at night.

Ladies and (preponderantly) gents, I give you Steven C. Den Beste: pro-technology only when it conduces to a world of anime nymphets gamboling nud3 on his beloved holodeck, and otherwise a gimlet-eyed skeptic to out-Rifkin them all.
posted by adamgreenfield at 11:38 AM on February 19, 2007 [4 favorites]

Also, Benny: mustn't overlook the real, deep problems with the Lovins' "supercar" approach.

I'm not nearly as wild and woolly as Jim Kunstler, but the car per se is part of the problem, even if it's impelled by God's Own Magic Fart Power.
posted by adamgreenfield at 11:41 AM on February 19, 2007

I'll never understand the optimistic ignorance that consistently accompanies discussions of photovoltaic power

Sorry, but it's not optimistic ignorance. It's a clear-eyed look at the problems - most of which *are* solvable, especially given the same kind of incentives and breaks nuclear has gotten - and opportunities in various kinds of solar research.
posted by mediareport at 11:44 AM on February 19, 2007

Also, Benny: mustn't overlook the real, deep problems with the Lovins' "supercar" approach.

Agreed. Lovins is obsessed, and like most obsessives, he goes off the deep end from time to time. But most of the time, he knows his shit, and very often he has real world data to back it up.
posted by Benny Andajetz at 11:45 AM on February 19, 2007

GENI
Even though two billion people currently do not have access to electricity, the world’s power plants have more than enough capacity to supply the entire planet’s needs. The problem, according to Fuller, is that they’re just not used efficiently. A given generator will produce a relatively constant amount of electricity, whether that power is used or not. If demand dips, the extra power goes to waste—and the power company loses potential profits. This effect is especially apparent at different times of the day—when it’s daytime in one area and electrical demand is high, it’s night somewhere else and demand is low. The solution to this imbalance is well known and has been used successfully for decades: interconnecting power grids from different areas. Thanks to technology that allows electricity to be transmitted over distances as far as 4,300 miles (7,000km), electricity can be shared across time zones or even seasons. So all across North America and Europe, local and regional suppliers have made arrangements with each other to balance the load. This results in more consistent availability of power, lower and more uniform costs, and a reduction in the number of generating plants needed.
posted by hortense at 11:49 AM on February 19, 2007

Thanks to technology that allows electricity to be transmitted over distances as far as 4,300 miles (7,000km), electricity can be shared across time zones or even seasons.

That's what "the grid" is. It's still not very efficient. According to the US Dept of Energy, in 2001 almost 10% of high voltage electricity was lost in transport. Low voltage electricity and electricity making 4300 mile trips would suffer more.
posted by Benny Andajetz at 12:06 PM on February 19, 2007

It's likely that this figure would be higher for transmision over such a large distance. Losses would have to be around 0.15%/100km to keep loses down to 10% over 7000km, the actual rate of loss is likely to be at least 3 times that high.
posted by biffa at 12:20 PM on February 19, 2007

20 years ago 16Mhz CPUs were the shit, now thanks to free-market throwing billions into the pursuit of R&D we've got CPUs 100x faster; is there any reason not to suspect a similar yield curve would apply to PVs, should the free market enterprise work its magics at similar scales of investment?

I don't know, has that yield curve ever applied to anything else? Batteries, cars, anything? They'll never be able to produce more energy then the light they take in, so even the wildest theories would never have them produce more then 50% more energy.

Heywood: I admire your optimism, but I don't share it. Renewable energy has never suffered from lack of investment. Wouldn't you want to be the company that comes up with free power? The fact is that electricity is, and always will be, many (many) times easier to extract from fossil fuels than any of the sources you cite.

Always? Even when the all currently known reserves are exhausted?
posted by delmoi at 12:27 PM on February 19, 2007

Some of the corporations manufacturing solar cells: BP, Shell, Kyocera, Seiko, Matsushita, Canon. And these are just the big boys. Do you know something they don't?

It's quite possible to make some money from solar without solar being economically competitive for most applications. The environmental benefits of PV means it attracts some subsidies, this reflects both its comparatively benign status environmentally and the fact that some countries want to see competitive high tech industries established on their territory. The wind turbine sector is now worth billions to Denmark and countries like Germany and Japan are happy to subsidise PV now to have their companies dominant in the market later.
posted by biffa at 12:36 PM on February 19, 2007

They'll never be able to produce more energy then the light they take in, so even the wildest theories would never have them produce more then 50% more energy.

$/watt has two terms in it.
posted by Heywood Mogroot at 12:38 PM on February 19, 2007

Every time one of these threads pops up, someone comes in spouting econometrics like a spreadsheet fountain to drown all the hopeful possibility in a sea of actuarial ooze. (Or something like that. Is my metaphor mixed, or just stirred?)

Anyway, I've been looking for awhile for a good default response, and in Walt Patterson's "Energy 21" I've finally found one:

Commentators scrutinize the prices of fuels and electricity, and analyze their movements minutely. However, in our modern interconnected society the prices of fuels and electricity by the unit have long been essentially artificial, shaped by preferential tax regimes, subsidies and cross-subsidies, cartels and outright monopolies, as in the case of electricity networks. With this in mind the highly respected chairman of Ireland’s Electricity Supply Board, Patrick Moriarty, once remarked succinctly, "The price of electricity is what the government wants it to be." Much the same can be said of fuels. Except for short-term advantage, price is not a good enough criterion.

Thin-film solar might or might not be the killer app, but it's worth a shot - and smart people are betting some serious cash that it is. And even if it isn't the Netscape of renewables in the end, still if you care about the future of humanity, you should be cheering for something to find that sweet spot. Not because it'll be a universal fix any more than Netscape was the one and only application that enabled the digital revolution, but because it'll hopefully provide sustainability with the critical mass of giddiness and dumb money it needs to conquer the world.
posted by gompa at 12:45 PM on February 19, 2007

Renewable energy has never suffered from lack of investment.

Oh and this? Spectacularly ignorant.

By one credible estimate, the US has spent something like $1 trillion in public funds to get the nuclear industry to the point where it generates less total energy than wood does. Compare that, please, to the pittance spent to bring wind power to the point where it's unit-cost competitive with coal and gas on a seriously tilted playing field.

Just to name one example.
posted by gompa at 12:49 PM on February 19, 2007

I don't know, has that yield curve ever applied to anything else? Yes - LEDs. The efficiency has been going up at a nice curve while the cost has been going down.

I was reading up about putting in solar panels where I live and found a few interesting articles about solar cost effectiveness. One article made the claim that PV cells have the capability of generating more than enough energy over their lifespan to manufacture their replacements and then some. Intersting article, although a little on the "drank the Kool Aid" side of things.
posted by plinth at 1:01 PM on February 19, 2007

They'll never be able to produce more energy then the light they take in, so even the wildest theories would never have them produce more then 50% more energy.

Err, I meant 50x. Not 50%
posted by delmoi at 1:30 PM on February 19, 2007

My fear with the $1/watt figure is that there is a large capital outlay at the onset. This is fine if you're building a new home; an extra thirty grand won't matter too much at that point. If you already have a home, well ... do I really have thirty grand to shell out right now? Economically, I would be better off investing that thirty grand in something else and just running the meter.

The real number should factor in rising costs in natural gas and electricity, but also pay attention to inflation, potential profits (invested in some low-risk portfolio), maintenance costs for the array (and battery, power conversion), installation, and disposal.

My guess? Try 50cents/watt. Higher if you can build a new subdivision with its own mini-utility.
posted by adipocere at 3:17 PM on February 19, 2007

Having dealt with a large 3MW distributed solar array, I will say that the future for solar is bright. Our biggest problem is a state law that says after 30,000kWh/yr we go from being paid in a net metering manner (we sell for the same price that we buy at) to wholesale prices (about 40% of the price we buy at). So our first goal is to get the law changed. Until then we're stuck adjusting our operations to make the most of the electricity we use before we have to put the rest back on the grid.

Energy costs sure as hell arent getting any cheaper. And fusion is always one of those "15 years away" technologies.

One of the keys is to bring down the price of installation. The materials can be as cheap as dirt, but as long as a qualified electrician costs $50+/hr for installation the initial capital outlay is going to be a problem. One of the properties of our distributed solar array is that it is ground mounted, so its easy to install and maintain. This is where plug-and-play panels would be the most useful. If you can use less expensive labor and only use the electrician for the inverter and grid tie-in, your acquisition costs go down.

The key is to get solar cell efficiency up, from about 20% now to something closer to 40%. Thats when you get all sorts of really neat opportunities. Solar cells on the roof of your vehicle generating energy while your car is parked and you're inside at work. 3-4 sq meters would generate about 5000Wh over the course of 8 hours. The average hybrid car uses 250-500Wh per mile (this also includes energy recaptured through regenerative braking). If leaving your car in the sun while you work could get you 10+ miles gas-free on your drive home would you take it? My commute is now 11 miles each way (and will be 15 soon thanks to our offices moving), but I'd still take it, and my own math says I'd use about 3 gallons of gas per week, plus the electricty cost of plugging in at night.
posted by SirOmega at 4:27 PM on February 19, 2007

SirOmega: i think I'd rather just have my garage recharge my car inductively rather than muss with mobile panels.
posted by Heywood Mogroot at 5:28 PM on February 19, 2007

Its not full panels as one typically thinks of panels - its the individual cells, about 2" by 2" laid out on top of the vehicle. It would look just about the same as the top of a normal car from the profile view.

The real issue is do you recoup the investment in the cells and circuitry and such in the 5 years or however long you keep your car. I'm thinking you might, if you figure that 5kWh/day is 50c/day if you were to plug it in or $1/day in saved gas, at 200 sunny workdays/yr.
posted by SirOmega at 6:17 PM on February 19, 2007

as long as a qualified electrician costs $50+/hr for installation

Bingo! This is the real problem with residential solar. Only when $/W costs are utility-competitive and you can buy a plug-and-play solar-power-in-a-box kit at Home Depot, will residential solar ever take off. Renewable energy contractors and local regulators will be a major obstacle to any real progress.
posted by FreedomTickler at 8:33 PM on February 19, 2007

With this in mind the highly respected chairman of Ireland’s Electricity Supply Board, Patrick Moriarty, once remarked succinctly, "The price of electricity is what the government wants it to be." Much the same can be said of fuels. Except for short-term advantage, price is not a good enough criterion.

I'd agree with this to some extent, but it's not the whole story. There do tend to be large parts of the energy supply system geared to supporting traditional models of energy delivery, i.e., centralised power generation, with passive network control to get to consumers. The regulation supports this and tends to lead to barriers, (effectively meaning higher cost or more extremely exclusion from the market), for distributed generators like renewables. Work is needed to not only reduce subsidies to fossil fuel generation but also to allow the full benefits of distributed generation to the grid to be effectively monetised. These changes require government willingness to embrace change rather than bowing to the needs of established utilities.
Now government can also subsidise renewables and I'm sure theoretically they could step in to the extent of doing enough to make them cheaper than fossil fuels under practically all circumstances but I'm not convinced that many developed countries will be willing to interfere with the market to such an extent.

as long as a qualified electrician costs $50+/hr for installation

A domestic flat plate sollector costs £2,000-£3,000. an evacuated tube collector costs £3,500-£4,500. Another £100-200 for installation is not going to be the decisive factor in installation economics.
posted by biffa at 4:14 AM on February 20, 2007

My evacuated tube collector cost me a great deal less than £3,500-£4,500. Regardless, we are talking about PV systems here, not water heating. And a working PV system costs considerably more than £100-200 for installation.

The monetary expense is not the only aspect of professional installation that depresses demand. Until a solution appears that does not revolve around negotiating with untrustworthy middlemen to tear up your house on an unpredictable schedule, then most people won't bother.
posted by FreedomTickler at 10:00 AM on February 20, 2007

Until a solution appears that does not revolve around negotiating with untrustworthy middlemen to tear up your house on an unpredictable schedule, then most people won't bother.

early last year I was actually investigating going into the solar market for just this reason; the current installers are pretty sketchy & the technology AT THE MOMENT is basically break-even, but if & when PV becomes a free-money sort of deal for my 'hood (Fresno) it should be the mother of all legal business opportunities (next to MtG dealer I suppose).
posted by Heywood Mogroot at 10:52 AM on February 20, 2007

One of my graduates is just starting a business up where he'll consult on microgen for domestic use, then subcontract out installation. The bank is throwing money at him. He plans to go down the legit line, he's already turned customers down where he doesn't think a system is a good idea, so hopefully he'll stay on the side of the angels.

The UK support mechanism for domestic microgen is massively oversubscribed. The three year programme ran out in 6 months. The replacement gives out funds on a monthly basis, this month it ran out in 10 hours. The majority of installations will still have lousy payback but there's a huge demand for the stuff anyway.
posted by biffa at 11:12 AM on February 20, 2007

Furthermore, the trend in cost-per-Watt in solar technology has been exponentially decreasing for at least a decade in some markets,

Err, I can't help but note that the chart you linked to only has data up to 2002. And the one on the front page of the same site shows that the trend did not continue, with costs reaching their low point somewhere around that time and increasing since then. But hey, if this new stuff works out as planned by 2009, that'll be great news.
posted by sfenders at 12:04 PM on February 20, 2007

"I don't know, has that yield curve ever applied to anything else?

Yes - LEDs. The efficiency has been going up at a nice curve while the cost has been going down."

I thought it prudent to point out that LEDs are also semiconductors, just like PV cells and processor chips, so of course they're on a similar (but not Moore's Law-style) curve.

If internal combustion engines could improve on the same curve, we'd have 100cc engines that could turn out 4000hp by now.

If nuclear fission followed the same curve, we'd have Asimovian "walnut-sized" reactors cranking out a megawatt.

And frankly, if PV cells were really on the same curve, we wouldn't be having this discussion. We'd have enough solar power to keep expanding our civilization at an exponential rate for millenia.

Moore's Law does not apply to energy transducers, and never will.
posted by zoogleplex at 12:33 PM on February 20, 2007

Moore's Law does not apply to energy transducers, and never will.

I call bullshit until more than x% of the monies going into semiconductor fab R&D is going into solar fab R&D.

(x being a suitably impressive number)
posted by Heywood Mogroot at 12:55 PM on February 20, 2007

Sorry, I couldn't stay away. It's not just the solar collectors that are getting better and cheaper. The "back-end" storage systems, power inverters, etc. are improving rapidly also.
posted by Benny Andajetz at 12:55 PM on February 20, 2007

And the one on the front page of the same site shows that the trend did not continue, with costs reaching their low point somewhere around that time and increasing since then

believe it or not, we're seeing Micro Econ 101 at work here. Demand (for the good stuff) exceeds supply, and government subsidies are serving as retail price (profit) inflators.
posted by Heywood Mogroot at 12:58 PM on February 20, 2007

also, given the supply/demand curve, PV manufacturers just have to remain marginally below the cost of TEPCO, PG&E, etc to maximize their profits.

PV is an industry/topic where free market fundamentalists really should check their ideologies.
posted by Heywood Mogroot at 1:00 PM on February 20, 2007

I call bullshit until more than x% of the monies going into semiconductor fab R&D is going into solar fab R&D.

Moore's law works because a linear increase in optical technology results in a quadratic increase in semiconductor density. You only need to make a small advance in optical technology in order to produce a chip with twice as many discrete components.

There is no reason to believe that solar technology will ever scale like semiconductor circuit density does.
posted by b1tr0t at 1:09 PM on February 20, 2007

"I call bullshit until more than x% of the monies going into semiconductor fab R&D is going into solar fab R&D."

Call bullshit all you want, but no matter how much cash you dump into it, the power output per unit size of PV cells is not going to double every 18 months. We're not talking about data processing, making circuits smaller and faster and running them with less power, we're talking about converting one form of energy to another. It's a completely different ballgame and there are definite physical limits built into it, otherwise we'd already have cells that could put out a megawatt per square meter per hour.

I suppose it's possible that the cost of the cells could be halved every 18 months, but I think the incredibly smart people who design this stuff would probably already know if that's possible. I don't see that semiconductor fabs are getting cheaper every year, it's just that the processing power yielded goes up per dollar invested.

I'm no free-market fundamentalist, either. There's a whole lot of people a hell of a lot smarter and more experienced than I am who apparently are unable to develop this stuff at a rate which defies the laws of physics. That should tell you something.

To use my example above, there are billions poured into piston-engine R&D and design every year. The engineers are ecstatic if they can squeeze one percent more efficiency out of a new engine that takes years to design and build. If Moore's Law applied to such processes (and PV R&D is pretty much the same thing), we'd have cars that got 1500 miles to the gallon while turning out 1000hp and 950 ft-lb of torque. Which is not physically possible.

There are physical limits to IC processor design, btw. We just haven't hit them quite yet, but there's an absolute maximum number of electron paths you can etch into a square millimeter of silicon. At that point, Moore's Law will utterly cease to apply to silicon-based semiconductor chips, and we'll need to work on some other basis for our data-processing.

All this said, I'm not pessimistic about solar power being an important part of our future, I think it's great that more people are getting more money to work on it, and I think good things will come out of it that will help us all out a lot.

It's just not going to be magic.
posted by zoogleplex at 1:16 PM on February 20, 2007

Benny Andajetz writes "South facing windowwalls, heat sinks. roof overhangs, and natural windbreaks can all put serious dents in existing energy usage. Why these issues don't get serious consideration blows my mind."

Because your average house buyer is interested in cheap square footage and "curb appeal" and your average developer is interested in houses they can plop anywhere in a development regardless of prevailing winds or solar facing.
posted by Mitheral at 4:25 PM on February 21, 2007