February 12, 2009
Solar Photovoltaic Prices Seen Falling
Prices are declining for polysilicon used to make silicon photovoltaic (PV) cells.
In a comprehensive report on the subject this morning, Collins Stewart solar analyst Dan Ries notes that spot market poly prices have fallen from a peak of about $450/kg in mid-2008 to the $130-$150/kg range more recently. That’s a pretty dramatic move - but the decline is far from over.
Ries contends that spot prices by mid-2009 will plunge to the $40-$60/kg level, due to a severe oversupply.
One analyst expects PV prices to drop so far that PV will start to compete with other methods of generating electricity.
The silver lining here is that in the long run, much lower prices for polysilicon are the most direct way to bring down solar electricity production costs low enough to compete with conventional utility scale power generation. With poly in the $40-$60/kg range, he says, module prices would drop to the $1.70-$2/watt range, and utility scale projects could produce power for 11 cents/watt. At that rate, he says, solar would be “reasonably competitive” with combined cycle natural gas facilities and wind turbines.
The world may be running out of oil. But it is not running out of energy. We can shift to solar, wind, and nuclear. We just need great batteries for the cars.
Some HSBC analysts expect a 20% decline in solar system costs in 2009. Big drop.
Analysts at HSBC forecast average selling prices for solar systems will drop by about a fifth in 2009 given oversupply and a tighter credit environment, but prices for cells and modules have so far fallen much faster than those for silicon and wafer.
If you are thinking about putting PV on your house now is not the time to do it. Wait a year and save big.
PV prices have already dropped in the face of contracting economies.
In the U.S. solar industry, the ripple effects of the crisis extend all the way to the panels that homeowners put on their roofs. The price of solar panels has fallen by 25 percent in six months, according to Rhone Resch, president of the Solar Energy Industries Association, who said he expected a further drop of 10 percent by midsummer.
For homeowners, however, the savings will not be as substantial, partly because panels account for only about 60 percent of total installation costs.
PV costs have declined. Yet a new announcement for the biggest set of solar power facilities in history highlights the use of solar thermal concentrators to create steam.
The largest series of solar installations in history, more than 1,300 megawatts, is planned for the desert outside Los Angeles, according to a new deal between the utility Southern California Edison and solar power plant maker, BrightSource.
The technology isn't the familiar photovoltaics — the direct conversion of sunlight into electricity — but solar thermal power, which concentrates the sun's rays to create steam in a boiler and spin a turbine.
The big solar and wind facility builds in California are driven by state government requirements on electric utility companies to get more electric power from renewables. This announcement tells me that the utilities see solar concentrators as still cheaper than PV.
Glad to hear or the big solar thermal installation near Los Angeles. I have contended that PV is best suited for local use and solar thermal is better for utilities creating grid power.
As always facts should trump MO, I'm sure not going to become an expert in solar.
There are two reasons I believe ST will prevail: no DC is involved; and, there is no complex network to carry current from each panel - the mirrors deliver the solar radiation to the generation point wirelessly.
Regards PV versus ST: Can ST's costs fall as much as PV's costs? I hear more about the prospects for declines in PV costs than I do for ST.
I also wonder if PV module costs fall by, say, a factor of 3 does the non-module cost of PV become the intractable part of the equation that keeps ST cheaper? Just how big is ST's cost advantage today? I would like a good source for this.
While I'm at it: Increases in conversion efficiency should reduce the cost of cables since the higher the conversion efficiency the less area that needs covering to generate the same amount of electricity.
So does PV have greater prospects of increased conversion efficiency than ST? I never read conversion efficiency numbers for concentrating solar power. You seen any?
Good to see you are keeping the faith as oil falls to $38/barrel! "The world may be running out of oil."
Still putting off the econ reading?
One analyst expects PV prices to drop so far that PV will start to compete with other methods of generating electricity.
Not any time soon. Since solar energy only shows up when nature says it can, the cost of solar has to include the cost of the 100% non-solar backup system that can carry the entire system load when solar goes randomly offline. It's as if you had a diesel generator that operated only a few hours a day sometimes. You would need a second generator that could take over completely from the first at a moments notice. The cost of the system would be the cost of both generators.
Until we efficiently store solar power so we can use it when and where we need it, it remains an expensive toy source of power for all but a few minor applications.
Actually, electric power companies are already successfully experimenting with cheap building-sized battery farms to store electricity during off-peak hours. Lead-acid batteries are very cheap and very easy to recycle, and the weight and size does not matter when these are just kept in a battery farm.
But what we do have to worry about photovoltaic cells, is the that some of these are made with very toxic chemicals like cadmium, and when the whole world starts using these, the result can be apocalyptic. It is very important to research organic or at least safe versions of these photovoltaic cells. This was mentioned in a previous article posted here.
Randall, Frankly I don't know about the cost curves of Thermal v. PV now, or for the near future. And who can given the variety of PV manufacturing methods and the pace of improvements? The related cost of the devices needed for conversion from DC to AC may also fall.
I don't expect thermal costs to fall greatly because the overall design is simpler and well understood. We probably aren't going to improve mirrors greatly. Or the generators running off the concentrated heat. But wide adoption of a good design would bring economies of scale.
My observation is that simplicity seems to favor thermal in construction and maintenance. Certainly PV is more suited to point-of-use generation such as home rooftops. So until PV costs fall I expect utilities will find thermal best while off-grid will generation constitute much of the PV market.
The wholesale price of electricity already varies enormously with demand with prices swinging well over an order of magnitude from winter night to hot summer late afternoon in the US - whereas Britain peaks in winter, not summer.
The high period overlaps with the PV electric power period. PV doesn't have to compete with baseload prices. It has to compete with peaking generators that run only a small fraction of the time burning natural gas.
Concentrating Solar Power (Solar Thermal concentrators) has a big advantage in this regard: it is easier to store heat for a few hours to use it to generate electricity in the hot summer evenings when everyone comes home from work and turns on the air conditioning. Sure, there's a cost to that heat storage. But that incremental cost has to compete with natural gas peaking generators, not with coal or nuclear.
Yes, I also expect a bigger drop in the price of PV. I am curious to know, however, how much of the cost of a concentrating solar system is the mirrors and how much the apparatus the mirrors point at and how much the generator.
A point about the generator: it is probably not going to drop in cost vis a vis generators in coal or natural gas plants. So any improvements in concentrating solar costs have to come in other areas. Can some nanotech material lower the cost of the mirrors? Can a cheaper way be developed to aim the mirrors dynamically to increase their efficiency? The mirrors all look like they are passively positioned in pictures I see of these sorts of installations.
Randall: you touched on solar thermal, which to me means ST used by utilities, to easily store heat for generation later. The storage medium would probably be water. Frankly I forgot this but hadn't intended to comment about the storage problem anyway.
Detailed design is beyond my time limits, and good design would be beyond my limits period. Some hypotheticals about ST v. PV:
(1) Mirror shadows would slightly cool the land beneath them but so would PV. What will be the net difference and the effect on the ecology?
(2) I recall environmentalists screaming about the proper disposal of PV cells. I am not sure why. Mirrors would simply be replated or repolished. Maintenance cost must affect calculations. Both cells and mirrors have to be cleaned on some schedule so cleaning may be a wash (ha ha) but perhaps not; it strikes me as a delicate business to clean a cell which can crack.
(3) Both PV and mirrors can produce more by tracking the sun. I'll skip over PV tracking. Mirrors have no innate connection to the generator and can be adjusted by a simple sun tracking device at each mirror. But it would be better to have a central control. So let mirrors receive pointing instructions by simple wi-fi type radio signal. Of course gears, adjustment servos, and signal receptors each have a cost.
I don't see how nano is going to add much efficiency or lower the cost of mirrors, they are pretty efficient anyway. But I am open to a surprise.
Summing up: I don't see falling costs for ST other than from economies of scale - building many. My guess is that rising costs of fossil energy will make ST attractive again and permanently.
And I can't answer about the division of costs between mirrors, pointing devices, and generator costs. That can no doubt be found. But subsidies and interesting accounting methods and advocacy make reported costs tricky in all alternative energy activities.
The little video I have seen of solar thermal showed the collector glowing bright white. That is reflected light which means that a lot of energy is being wasted right there. Some type of arrangement where the light is bounced around in a cavity until it is absorbed might help.
Are you frakin kidding? Solar and wind will save the world?
Gee just one word to debunk all this farkin green stuff
If it don't scale up, its a dead idea.
EESTOR is primarily being hyped as the holy grail of storage for full EVs, but if it works it could jump start the entire home solar market. If two years from now you can buy a cheap solar array plus EESTOR, that's real energy independence.
Solar thermal needn't use water as a storage medium. The Solar One demonstration project in California used sodium salts, which they were able to heat enough to keep generating power on a twenty four hour basis. The cost per KWH in that project was horrendous, but I don't know if the driver was the sodium or not.
The big solar thermal project in the Mojave can better be described as political connections meet utility under the thumb of the politicians. When you have a project that has an investment by the politically connected Democrats of course PG&E is going to buy from them regardless of the cost (which by state law cannot be disclosed) or the value of the project. Follow the money and see how rate payers are going to enrich the political parasites.
"Oakland firm signs solar energy deal with Edison" by Marla Dickerson in the Los Angeles Times on 12-Feb-2009:
Renewable-energy analysts have high hopes for solar-thermal power, which some believe will generate electricity as cheaply as dirty coal within 15 years. But the technology is loathed by some environmentalists. Solar thermal plants require vast tracts of land as well as precious water for generating steam and cooling the turbines. Costly transmission lines are needed to transport the power long distances to where it's used. With dozens of solar, wind and geothermal projects planned for California's deserts, some fear that this unique habitat will be destroyed.
IF we had spent only half the money lost in Iraq on energy research, then already by now solar thermal power would have already become cheaper than coal.
But seriously, even Obama is not making a massive cash injection for energy research. We need $250 billion per year for energy research ASAP.
"Solar can scale. Dr. David Goodstein of CalTech says that with 10% efficient PV we would need an area 300 by 300 miles to get as much energy as we get now from fossil fuels. Of course, get up to 30% efficiency and the area needed shrinks. That 300 by 300 mile area is less than all the rooftop area in the United States. So how does PV not scale?"
And what will the 90,000 Sq miles (300 X 300 just sounds so much smaller and more attainable than 90,000, doesn't it)? Well, if you use $50/sq. ft. that gets us a "mere" $10,000,000,000,000,000 (Quadrillion). With the current world GDP @ $60 trillion... about 165 years. No problem... I'm sure when the economies of scale & higher efficiency panels are available this will lower to "only"...50 years worth.
"We need $250 billion per year for energy research ASAP."
Are you insane? The world does not have that many PhDs in that field to put that money to good use. It would just be a waste.
"IF we had spent only half the money lost in Iraq on energy research"
Left-wing stupidity like welfare costs a lot more than Iraq. Of course, you only oppose Iraq because it is fashionable, not based on actual knowledge of the subject.
Who said I only oppose Iraq? Also, not all the $250 billion per year will go to "PhDs", maybe I should have said Research AND Development: most of it will go to projects that are already viable, which involve construction of cellulose alcohol plants that are already viable, genetic engineering firms that are already successfully inventing new trees that grow very fast, solar-thermal plants that are already improving a lot, and newer improved wind turbines that are cheaper and more efficient, not to mention better photovoltaic cells. All these things require laboratory equipment that is much more expensive than PhDs scientist salaries. Battery research involves more than salaries.
OK, even $100 billion per year would have solved the problem in 5 years,.
PV doesn't have to compete with baseload prices. It has to compete with peaking generators that run only a small fraction of the time burning natural gas.
Absolutely not. You cannot use solar to replace peak generators because you have no means of knowing if the solar power will be there when the peak is. You can't use statistics to predict solar powers availability. Using statistics you can show that since solar power generates peak 6 hours out of the day, then that means you can get 1/4 output from PV at midnight. Even being an hour offline unexpectedly destroys its utility.
Power has to be there when and where you need it. Solar power picks its own when and where. Yes, it will help with peak usage but only sometimes.
Dr. David Goodstein of CalTech says that with 10% efficient PV we would need an area 300 by 300 miles to get as much energy as we get now from fossil fuels.
SOME OF THE TIME! The rest of the time you have to have a complete backup. Back in the early 90's Arizona had a freak tropical storm hit that caused 5 days of torrential rains. Any solar power plants in the area would have been knocked completely offline.
The great stupidity here is that we're jumping through hoops to try to scavenge energy from the ambient environment sheerly out of social and political desires that have nothing whatsoever to do with the technology. We're going to end up a couple of decades down the road unable to keep the lights on much less power heavy industry while the Chinese and other will be powering their economy with great gobs of dense power energy from mass produced nuclear plants.
We're going to end up in a circumstance like that faced by agrarian societies who were powered by biomass faced when they ran into industrial societies powered by fossil fuels.
Please check your math.
300 miles = 5340 * 300 ft = 1,602,000 ft.
300 sq miles = 1,602,000 ft * 1,602,000 ft = 2,566,404,000,000 sq ft
@$50/sf = 128,320,200,000,000 = 128 trillion. I.e. 2 years' worth of world GDP, not 165.
I didn't use statistics to show that PV can produce 1/4th of its output at midnite. I too can construct strawmen to knock down. But that is not productive and you know that.
Predict solar power availability: Here in sunny SoCal clouds in the summer are extremely rare. They do happen. But then nuclear power plants, coal plants, and natural gas plants unexpectedly break down. We've got power lines bringing power in from other states in part for that reason.
Goodstein's point is that the total amount of solar power available is enormous. The biggest problem is cost of PV. Storage is another problem. But most electric demand is during the day. At a low enough price (a price that I think is inevitable) solar can replace almost all peaking natural gas and it will actually make day time electric power cost less than night time electric power.
Since most money spent on electric power is spent on higher priced daytime electric power that means that PV will eventually slash electric power costs without replacing all other power sources.
Arizona freak tropical storm: Yes, and it was a rare event, right? How about dynamically priced power with industries that are ready to shut down for a few days once every 20 years or so?
Look, I'm not arguing for solar to totally replace all other power sources. But solar plus wind plus nuclear can replace coal and a very large portion of natural gas. When that happens depends on how fast costs drop. This brings us back to the topic of this post: solar power price declines. These declines are going to keep happening for some years to come. Materials advances will allow roof shingles to serve as PV. That'll remove the cost of an additional roof layer which we see with current PV.
"...utility-scale projects could produce power for 11 cents per watt."
I hope this is some sort of misprint. Around here we pay around 12 cents per kilowatt, retail.
I too can construct strawmen to knock down
It's not a strawman. It is how people argue for solar power while ignoring the very real problem of intermittency. Electricity is not a luxury. It is the lifeblood of our technological infrastructure. When it goes away people die. It's like breathing. It doesn't matter if you breath regularly for decades, you stop breathing for 4 minutes and your dead. It would not matter if you had a PV system that could provide all our electricity needs if that PV system cut out randomly 1% of the time you would have to pay for a 100% redundant backup system to handle that 1% of the time.
Here in sunny SoCal clouds in the summer are extremely rare.
That is rather the problem. People in SoCal have this strange delusion that the rest of the country and the world shares their freakishly stable weather. Well, we don't. Even in SoCal, you would face random massive outages if you relied on solar and wind for the majority of your power.
But then nuclear power plants, coal plants, and natural gas plants unexpectedly break down.
Individual components break at random but entire plants never go completely offline. Even if they do, guess what, you have built in redundancy in other plants that take up the slack. Solar power has no such capacity for redundancy. When the sun goes down or the weather freaks out, you can't use your neighbors PV to run your house. Again, solar does not provide power when and where you need it on demand.
We've got power lines bringing power in from other states in part for that reason.
No, you import power because in the 1980 California's power planners decreed that electricty demand would be flat forever. Environmentalist shut down plants in state and instead forced the state to export the environmental consequences of power generation to other states so Californians wouldn't have to deal with it. Texas has its own power grid and generates all its own power. California used to do the same and still could today if they had not made so many air headed decisions based on political fads.
Goodstein's point is that the total amount of solar power available is enormous.
But its so diffuse and inconstant as to render it useless. People talk about energy as if it was somekind of fluid we harvest an bottle. It's not. Rather it is the ability to move matter and energy when and where we want to. Solar doesn't really allow us to do so.
Storage is another problem.
No, storage is the show stopper. Having ten times the energy you need is as useless as having a tenth of the energy you need. When and where are everything. If we can develop a storage technology that works reliably as fossil and nuclear then we can use very inefficient PV or other technology to scavenge solar energy. Until then, even the most efficient are highly useless.
Arizona freak tropical storm: Yes, and it was a rare event, right?
So, is being smothered for 4 minutes once in your life. Again, electricity not a luxury. The Arizona storms would have caused a solar power blackout lasting 5 days.
How about dynamically priced power with industries that are ready to shut down for a few days once every 20 years or so?
Plants that shutdown on schedule are not a problem. Again, where and when. Plants that shutdown for political or financial reasons are not a technological issue.
But solar plus wind plus nuclear can replace coal and a very large portion of natural gas.
No, the situation you describe requires a nuclear power infrastructure that provides 100% of needed power so that it can take over when solar and wind go offline. In that case, why not just use nuclear all the time if you've got to build 100% capacity anyway?
Overselling solar power does very real harm. I just watched a brutal fight over a new generation coal plant because the opponents claimed they could completely replace the coal plant with solar and wind. A lot of people believed them despite the fact that the sun goes down and the wind stops blowing.
One of the substantial advantages of solar thermal over PV is the availability issue. You can easily add a gas-firing capability to a solar thermal system with very little additional capital cost. Being able to store the solar heat would be even better.
Let me go over the basics:
1) Dynamic pricing will allow load shifting to different times of the day. Given a large enough gradient in costs from lowest to highest priced times many ways exist to shift demand.
Ease of demand shifting will vary by industry and process. It will also vary by technological advance. For example, the trend in recent years is for computers to be a dropping fraction of computing costs with electricity and cooling becoming rising fractions of total computing cost. If electricity costs exceed computing costs (and some observers predict this) then it would make sense to have multiple computing farms around the globe and run each farm far more when the sun is shining. For design work that could mean running sims most heavily on sunny mornings before afternoon HVAC demand starts competing.
Another example is with electric car battery recharging. Once electric cars become widespread the best time to charge them will become around noon time. Also, while today gasoline costs more in the summer and less in the winter that'll change in the electric car era. The cheapest time to take a vacation road trip will come around the longest day of the year. Ditto for powering electric trains.
2) Storage does not always take the form of trying to regenerate electricity at a later hour. That's the expensive part. But there are other ways.
For example, skyscrapers in Manhattan are putting big heat/cold storage tanks in the basement so they can run their heating/cooling systems at night when electricity is cheapest. There the storage is in the form of heat or coolness of a material and it is already today seen as cheaper than the price of afternoon peaking power.
3) Solar power can compete with (expensive) peaking natural gas once solar gets below the incremental cost of burning the natural gas. Most of the cost of peaking plants is from the natural gas, not from the physical plant. Solar will lower the cost of peaking power during the afternoon. The natural gas peaking plant operators will have to amortize the cost of their physical plant over a smaller window of evening time and so evening peaking power costs will probably rise. But the result will be lower overall electric power costs.
4) Suppose PV becomes as cheap as baseload. That'll lower the price of afternoon power below the cost of late night power. The market can handle this. The price of late night power will go up while the price of afternoon power will go below the current night time power.
Paul F. Dietz,
I've also read reports of FPL and other utility operators to marry concentrating solar power with existing coal and natural gas plants. A recent report in Technology Review surveys concentrating solar with natural gas and coal.
Feeding heat from the sun into coal-fired power stations could turn out to be the cheapest way to simultaneously expand the use of solar energy and trim coal plants' oversize carbon footprints.
At least that's what the Electric Power Research Institute (EPRI), a nonprofit organization backed by the electricity industry, is hoping. Last week, the institute launched a nine-month, $640,000 study to pin down the scale of the opportunity and the engineering challenges involved with making these seemingly disparate technologies work together. The study will examine the potential use of solar-thermal technology at a pair of coal-fired power stations, in New Mexico and North Carolina.
Combining solar power with fossil fuels is not a wholly new idea: over half a dozen new and existing natural-gas power stations are being designed or adapted to incorporate solar-thermal technology, which involves capturing heat generated using fields of mirrors and heat-collection tubes.
That ability to build hybrids is a big advantage that concentrating solar has over PV.
1) Dynamic pricing will allow load shifting to different times of the day.
Which won't help if the power simply isn't there. Again, you have to have a 100% redundant backup for the solar system which means a dual generation system no matter how you do the accounting.
2) Storage does not always take the form of trying to regenerate electricity at a later hour.
Yes it does when your talking about the entirety of the electric grid. Again, you talk about systems that depend on predictable steady electricity. Solar does not give you that. If you don't know how much power the solar will generate from day to day or hour to hour, how can you price it for such systems?
3) Solar power can compete with (expensive) peaking natural gas once solar gets below the incremental cost of burning the natural gas.
Absolutely not. Natural gas plants are used primarily to compensate for variation in the grid. They can be spun up quickly to handle unanticipated demand. In reality, solar and wind power will require more natural gas plants or something similar, not fewer. Due to the unreliability of solar, the natural gas plants will have to stand by to take up the slack. If you have a city that has 20%+ of its peak power coming from solar power what happens when when a rain storm sweeps through and knocks out the solar? That's when you need the redundant backup which right now means natural gas.
4) Suppose PV becomes as cheap as baseload.
It never really will due to the redundancy problem. The cost of solar will always be the cost of solar power system plus the cost of its 100% non-solar backup. You will always have to plan your grid to have two or more sources of power that can carry the entire load.
The only benefit solar brings is a reduction in carbon emissions. If were going non-carbon that means we have to have solar plus nuclear. Since you have to have 100% nuclear anyway, there is no point investing in significant solar power.
300 miles = 5340 * 300 ft = 1,602,000 ft....WHAT?????
A square mile last I checked is 5280' X 5280' = 27,878,400 square feet.
X $50 sq./ft = $1.4 billion sq. mi.
X 92,000 sq. mi = $10 QUADRILLION (+/- a quadrillion but why quibble)
I suggest..YOU...check your understanding of what a square mile is.
And even if we take your (incorrect) numbers do you understand WHAT world GDP means???? Are you suggesting that EVERY SINGLE DOLLAR spent for 2 years is a reasonable expenditure for a part time source of electricity?
What do you suggest the 6 billion people on earth do for those 2 years that ALL the production of goods & services are spent building this boondoggle?
Randall, Goodstein was estimating for the whole world, and forgot to adjust for thermal generation inefficiency (I asked him when I noticed the math was wrong, and that's what he said).
100W/SF insolation x 10% efficiency = 10W/SF. 10W x 5,250^2 = 250MW/sq mile. 450GW average US electrical consumption / 20% capacity factor = 2.25TW capacity needed.
2.25TW / 250MW/Sq Mile = 9,000 sq miles, or a square about 95 miles on edge.
As far as costs go, I would note that no one is suggesting that solar provide 100% of our needs, just some of the peak daytime needs. Wind is much cheaper and more scalable, and will be for a while: solar's getting cheaper, but wind has a big head start, and is also getting cheaper as well.
Wind could replace coal and power all light vehicles for only $2T, which is the long run will be cheaper and low emissions.
PV is best for installation at end-users, because it doesn't need much economy of scale: a 100KW installation on a commercial roof can be done quite cheaply, probably enough so to achieve grid-parity without subsidies in S CA.
100% nuclear won't work because it is too costly as a source of peaking power. Solar can at least satisfy some of the peak demand. But I do not see 100% nuclear or 100% solar as cost effective. I do not see why any of these power sources have to be judged on their ability to scale to 100% of all electric demand.
The price of electric power at the time it is generated is an important consideration even if it can't be generated 24x7.
You see the need for back-up peaking natural gas as a fatal flaw for solar power. I do not see why that is fatal. I've read estimates (which vary depending on the current price of natural gas) of cost for peaking natural gas electric as about 70% of the cost coming from the natural gas. The natural gas is the bigger cost, not the plant that sits there not running most of the time.
Given cheap enough solar the cost of solar plus back-up peaking natural gas electric is cheaper than using peaking natural gas alone. Do you dispute this?
Also, concentrating solar can reduce consumption of coal and natural gas at more efficient electric power plants and so can cut fossil fuels usage in a system that is 24x7. It just won't contribute 24 hours per day.
Nobody is arguing for a shift to 100% solar at the current prices for photovoltaics.
Part time source of electricity: Solar is there when America most needs it: Hot summer afternoons. Though not all countries have the same peak electric demand profile (e.g. the UK peaks in winter).
$50/square feet: We'll have higher conversion efficiency (so fewer square feet required) and lower cost per square foot by the time solar scales up to double digit percentage of total electric power generated.
I wasn't suggesting a shift to 100% solar at the current prices for photovoltaics
...Goodstein thru out the 300 X 300 figure. All I was attempting to do was show how ridiculous & unrealistic a 100% solar world would be. What percentage do you think is reasonable? 10% would cost $1 Quadrillion, 1% is still $100 Trillion not including infrastructure & storage. Your belief in lowered cost/improved efficiencies hasn't been realized in the last 30 years...what "breakthoughs" are you not telling us about? The higher conversion efficiency panels I've read about are multiples in cost of current panels...so you'll be starting from a higher threshold.
This discussion reminds me of what I believe to be the greatest technological embarassement of the 20th century... the meager improvement of battery technology relative to the money spent. Solar panels are following the same path.
Supposedly First Solar's current production cost for PV is around $1.14 per watt which is about a third of the market price of a year ago. I expect to see prices drop and First Solar to continue to cut their production cost. Thin film rivals claim they can get below $1 per watt.
If Nick G is correct then we only need 95x95 rather than 300x300. That lowers costs almost an order of magnitude and thin films can lower costs by a factor of 3. Amorphous silicon on glass might get costs to $0.70 per watt:
Another thin-film technology, called amorphous silicon on glass, is already making an impact on the solar market. It has efficiencies of around 7 percent, and because it uses only tiny quantities of silicon, it has been largely unscathed by the silicon shortage. Also, because manufacturing equipment is more readily available than for CdTe technology, there is a lower barrier to entry for would-be manufacturers. “The customer can get everything from us,” says Juerg Steinmann, head of marketing communications at Oerlikon Solar, in Truebbach, Switzerland. Its services are proving to be popular, and customers are currently inquiring about production systems for the manufacture of 40 MW or 60 MW per year. “Within a relatively short time we will have gigawatt factories,” says Steinmann.
Oerlikon’s tools produce 85-W solar panels covering 1.4 square meters, using a 0.3-μm layer of amorphous silicon that strongly absorbs visible light. However, output can be increased by nearly half with the addition of a 1.5-μm microcrystalline layer of silicon that absorbs infrared radiation too. Late last year Oerlikon introduced modified process equipment for microcrystalline growth. Even with this additional growth step, manufacturing throughput is fast, and the Swiss company contends that a manufacturer wielding its tools could make a solar module about as quickly as First Solar can. Nevertheless, the all-important cost-per-watt ratio is slightly inferior. “Today we’re looking at $1.50 per watt, and by 2010 our goal is going to be $0.70 per watt,” says Steinmann.
That 1.4 square meters might put out 125 watts if their microcrystalline growth method works out. They'd charge about $100 for that 1.4 square meters. Your $50/square foot isn't directly comparable though because I'm guessing it is based on complete system cost for a house. I found another quote for a house system that worked out to a total cost of $54 per square foot. So that's why I suspect you are quoting a home roof installation total cost.
My expectation is that PV costs will drop a lot more when PV shingles become feasible. Then the PV will be designed in and the extra costs associated with tacking on PV to an existing structure will be avoided.
If average power is 20% of peak output then 1 watt of power becomes 4.8 watt-hours per day. So $1000 worth at $1/watt is 4.8 kwh per day for $1000 worth of an installation. That 4.8 kwh is probably about 50 cents worth of electricity per day or about $180 per year. But the $1000 for the modules is only part of the total system cost in the home.
If a pure solar United States would need 2.25TW capacity then let us chop some numbers based on this idea. At $1 per watt we are talking 2.25 times 10 to the 12th power. Or $2.25 trillion. But, how much would the non-module cost of this massive solar farm run us? What percentage of total system cost comes from module cost?
If we do not put solar panels on houses does that substantially lower total system cost?
Randall, a few thoughts.
As you noted, it makes no sense to try to solve our supply problems with just one source. A pure solar US doesn't make much sense for a lot of reasons: 1) 2.25TW of PV would produce a peak of production about 5x as high as noon-time consumption. PV doesn't have built in storage, so that would be a big problem. For CSP, even if molten salt storage is pretty cheap that's still a lot of unnecessary infrastructure. 2) winter is also a big problem. In particular, wind is better at night and during winter. 3) a massive PV farm would be on the utility side, and would have to compete with wholesale prices, which are half those of retail.
"how much would the non-module cost of this massive solar farm run us?"
It makes more sense for PV to be on I/C rooftops, which is where 80% of CA installations (by KW) are happening, because of economies of scale and flat roofs. New construction is best because of integration with the roof, but these days new I/C construction won't provide a lot of square feet. The ideal size for consumer-side installation is about 90% of the consumer's noon demand - that eliminates the need to deal with selling power back to the utility, and maximizes savings.
Residential is 80% of installations, but they're much, much smaller than I/C on average, and much less economic. Residential only make sense currently because of the non-economic value to the home-owner, which can be substantial. Residential new construction might make sense in the future if the industry achieves very large integrated roof-module manufacturing economies of scale (right now PV roof tiles are surprisingly expensive), and if installation becomes very efficient through integration with the construction process.
The Balance of System includes controls, inverters, mounting structures, and wiring. These are a very fast moving target. Their prices have fallen very quickly in the last few years, and some innovative approaches are in competition. I would hope that overall system prices would fall to $2/W in 10 years, which would give $.10/KWH. That's the average US price currently, so I would think that it would beat peak prices in most of the country at that time.
As I discuss at my blog, I think that we're now at grid parity without subsidies in ideal locations. That means that PV will continue to grow very, very quickly due to demand. Heck, even now in the worst bank panic since 1929, PV is still growing, albeit slowly enough that supply can finally catch up with demand and prices can begin to fall as quickly as costs.
"No sun? No problem: These parabolic mirrors gather heat energy for a 150-megawatt hybrid solar/natural-gas power plant under construction south of Cairo. During the day, solar heat will displace a fraction of the natural gas required to drive the plant's turbines. At night, natural gas alone will assure continued power generation."
Since its first commercial application in california almost 30 years ago, solar thermal has increased its solar efficiency from 14% to 20%.
This must be why Florida Power, with an impressive debt load itself, decided to waste its money on a bunch of mirrors that amounts to a total of .02% of this plants' output. Results like these, I can't wait for the article on the next innovative, high-tech, all-clean, all-out failure, what this country really needs is a good $10 gallon gas tax, right?
Good to see you are keeping the faith as oil falls to $38/barrel! "The world may be running out of oil."
Simple supply and demand. Have you looked at what is happening to production now... and what did not happen to production as prices spiked to $147/bbl? World oil demand ran into peak production capability (which is declining in all 7 out of 7 supergiant fields in the world) and that was the last straw for the financial system. Now demand is collapsing faster than supply is falling, so prices are falling. This won't last.
Actually, electric power companies are already successfully experimenting with cheap building-sized battery farms to store electricity during off-peak hours.
It's even cheaper to store some of the products of electricity. Since so much of the daytime SoCal demand goes to A/C, it would be a simple matter to buffer demand with a bit of ice storage. When a cloud goes by, shut off the compressors and coast on ice for a while; make up the difference overnight.
Randall: If I were Energy Czar of California, I'd ask the makers of the Stirling-cycle dish systems why they couldn't use these units to shade buildings and parking lots; the waste heat might even be usable for absorption A/C. This would eliminate transmission losses and reduce demand in two different ways.
Dr. David Goodstein of CalTech says that with 10% efficient PV we would need an area 300 by 300 miles to get as much energy as we get now from fossil fuels.
SOME OF THE TIME! The rest of the time you have to have a complete backup.
I'm sorry, but that's just not true, and it can be shown rather trivially: if the sun isn't out, the A/C load is much reduced. Since A/C load is directly correlated to PV output, you don't need anything like complete backup for PV to deal with A/C peaks (and ice storage can time-shift cheaply).
You're right that nuclear plants are a damn good idea... but when PV can be added in very small incremental pieces at the point of use, and the supply correlates so well with the local sources of demand, you're going to be able to use plenty of that too.
Quit with the strawman about depending on PV for the majority of your power. Nobody's advocating that, so it's dishonest of you.
Nick G: Peak consumption is what, 800+ GW? Ice-storage A/C could time-shift demand so that the consumption peak corresponds to production, and electric vehicles could increase that further. All of this is either free (a byproduct of other things) or relatively cheap.
EP: Peak demand may correlate with sunshine in the south. But, up north, I have seen brownouts on really cold days. If we are going to run out of oil and other fossil fuels, we will have to heat the cold parts of the country electrically. How are we going to do that?
I'm suggesting that we build heaps and heaps of molten-salt reactors in sizes small enough that the vessels can be constructed in factories and transported to their final locations on trucks. They would then be installed deep underground, making them utterly invulnerable to 9/11-type attacks and most other terrorism scenarios. The waste heat from the units would be converted to low-pressure steam, and used for space heat and absorption A/C.
We should tell the NIMBYs to fuck off and do this beneath our cities. I am calling this the GUMBY position: Go Under My Back Yard.
"Peak consumption is what, 800+ GW?"
I'd love to see data on that. The data seems to be scattered among various ISO's, but my observation of a few suggests to me that the peak is only about 60% higher than the average - that suggests about 750GW.
"Ice-storage A/C could time-shift demand"
Sure, but there is a cost to that, both in terms of capital and efficiency.
"electric vehicles could increase that further."
Sure, though night time charging is a bit more natural for PHEV/EV's.
Wind is at least 50% cheaper than solar: I can't see a need for more than about 500GW of solar, which would give us a market share of very roughly 25% of KWH's.
"heaps and heaps of molten-salt reactors "
That would be great. Thorium would eliminate weapons and proliferation concerns, which I think is really at the heart of most people's concerns. The problem is, that will take a while: we need lots of low-CO2 power sources very, very fast.
"we will have to heat the cold parts of the country electrically. How are we going to do that?"
With wind, mostly. Wind tends to be a bit stronger in the winter. Nuclear would work well also, but we can build wind much faster. We can keep the FF plants around for the 5% of the time that we have prolonged, widespread calm periods.
Quoth Nick G:
night time charging is a bit more natural for PHEV/EV's.
Night's only special because the available generation exceeds the typical load during those hours. Add lots of PV, and noon also fits this description.
"Night's only special (for EV's) because the available generation exceeds the typical load during those hours."
Well, the existing infrastructure for charging PHEV/EV's is in...home garages. Commuter cars won't be in those garages during the day. So, night time charging is a bit more natural fit. That may change, but the fact that vehicles are mostly used during the day will always make night time charging a bit more important.
"Add lots of PV, and noon also fits this description."
Sure, but why would we? PV is more expensive (currently at least $15/W vs $6/W for wind) - why would we want to install so much of it that we began to exceed the typical peak load? I would expect that wind and nuclear would provide base-load (augmented 5% of the time by FF plants), and that solar would provide daytime peaking power where needed.
In the far future all of this may change, and it's useful to be aware that solar could provide all of our needs if necessary. But, for the foreseeable future I should think that solar won't provide more than, say, 500GW in the US, or roughly 25% of our KWH's.
Just to expand a bit: consumer-side PV only makes sense as a provider of more than peak load in the presence of net-metering. Most net-metering programs are limited by statute to a small % of KWH's, perhaps 1%. Those caps can get lifted, as they did in CA lately, but it won't make sense economically to raise them much above 10% or 15%.
Plugging (PH)EVs in during the day requires some conduit, wire, control boxes and flexible cabling. If we're going to slap large amounts of PV on buildings it wouldn't be much of a trick to put plugs in the associated parking lots; even easier if the PV is covering the carports, no?
PV is more expensive (currently at least $15/W vs $6/W for wind
You're way overstating the current cost (even installed), let alone the future cost.
why would we want to install so much of it that we began to exceed the typical peak load?
How about "because the production is peaky and we have good ways to use the excess power right next to where it's made"? Seriously, just between vehicles and ice-storage A/C, there are plenty of ways to productively use the generation peaks from PV. I understand that wind in CA peaks at night, so there's a reason to plug in at night... but the same generation-following system that works for wind also works for PV.
Net-metering works along with time-of-day metering. Generation during demand peaks is worth more than off-peak, and PV is more suited to meeting those in the South and Southwest (esp. because ice-storage would be filled up immediately before it was most needed).
Daytime recharge typically means recharging while you are away from home - at least on most days. If your car is at work then will companies cover their parking lots with PV roofs and then sell electricity to cars parked underneath those roofs?
As for shifting electric power demand to around noon time when PV will put out the most power: Given sufficiently cheap PV electric I can imagine a market for the electric in driving chemical processes to create synthetic liquid fuel and even fertilizer. What I would like to know: What is the non-energy cost of, say, a carbon-hydrogen fixation chemical plant? The problem I see with such a plant is that it would have a low rate of capital utilization. On average it would be operating less than half each day and even less on cloudy days.
What PV needs: industrial processes with high energy costs but low capital costs. Then the excess capital plant needed to compensate for low utilization rate would not be an obstacle for driving such a plant only off of PV and only when the sun is bright enough.
If your car is at work then will companies cover their parking lots with PV roofs and then sell electricity to cars parked underneath those roofs?
I expect that will happen; meanwhile, the cars provide (and sell) grid-regulation and spinning-reserve services.
"You're way overstating the current cost (even installed), let alone the future cost."
My mistake - I should have clarified that I meant the cost per average production, not per nameplate watt. PV is, at the lowest, about $3/W nameplate: for 20% capacity factor, that's $15/average watt. Similarly, wind is about $2 and 30%, for $6.7/avgW. Do you really see PV becoming cheaper than wind per avg watt anytime in the foreseeable future?
If solar is as (or more) expensive as wind, why would we overbuild solar relative to wind? Wind would require less time shifting than solar: chances are that geographical diversity (with a modest amount of long-distance transmission) and Demand Side Management (mostly PHEV/EV dynamic charge scheduling) would eliminate most need to actually store power, with it's attendant inefficiencies. V2G might be needed - that would be the cheapest form of storage, by far, and it would be slightly better suited to night time, when vehicles aren't in use.
Net metering won't happen above very roughly 10% solar market penetration: why would society pay retail prices for solar electricity when it can pay wholesale prices for wind and nuclear?
I'm puzzled: I think we really agree on most of this stuff: the potential for wind and solar, and for DSM etc. for both. Isn't 25% of KWH's from solar perfectly good? Why argue for more?
To clarify further, with an extremely simplified preliminary model: let's assume today's 450GW consumption, plus 75 for PHEV/EV's for a total of 525GW. Current consumption is probably 250GW for 7PM to 7AM, and roughly 650GW for 7AM to 7PM. PHEV/EV charging might raise night time demand to 350GW, and daytime to 700GW. Wind and nuclear might provide the baseload of roughly 350GW and solar could provide the daytime an extra 350GW. That would give an average from solar of 175GW, or 33%.
I should have clarified that I meant the cost per average production, not per nameplate watt.
That does clarify matters.
However, you're only looking at the capital cost of base-load power, not the full cost of the peaking power that PV would largely displace. Peaking is dominated by fuel costs, and even capital cost of $15/average watt is pretty cheap for quite a few of those plants. As PV heads down to the $1/W(peak) mark, you're closer to $5/average watt and a supply curve that offsets far more expensive power.
If solar is as (or more) expensive as wind, why would we overbuild solar relative to wind?
Because availability correlates extremely well with A/C requirements, and it can eliminate the capital cost of transmission & distribution. It also eliminates siting studies and delays for wind farms and lines, and who's going to object if the neighborhood strip mall or office park gets a slightly different-looking roof?
I live in the Great White North (not so white after last week's thaw, but still) where winter heat is far more a factor than summer cooling. My part of the country can benefit a lot from wind farms feeding heat pumps, and EVs to buffer power demand on a time scale of hours. The Southwest has rather different conditions, and different technologies are appropriate.
NB: Some time ago I calculated that it would take ~120 GW to replace the fuel used by gasoline-powered vehicles, assuming no efficiencies from electrification; replacing diesel would take roughly another 60 GW. Much of this use is during daylight hours. If you spread this over an 8-hour day, that's as much as 540 GW that could be productively used just to run transport (serving A/C load would justify quite a bit more). The reason you might want to add 5 GW of PV in the state every year is that it takes many years to plan and build a new base-load plant and lines to get its power to local markets, so local and/or short-term differences between daily demand and supply are easiest to make up with something like PV.
" even capital cost of $15/average watt is pretty cheap for quite a few of those plants."
Could you elaborate on this? I'm not sure what you're referring to. Natural Gas plants can have very low capacity factors, but they compensate with revenue from capacity payments. Have any ISO's done capacity credit calculations for central PV plants?
"As PV heads down to the $1/W(peak) mark, you're closer to $5/average watt"
I'm assuming very roughly $1.50/W (nameplate) for PV panels (First Solar is selling them at $2.50/W), and $1.50/W for all other BOS and installation costs, which seems pretty aggressive to me, at least for a while. Do you think we can get close to $1/W in the next 10 years, and if so, how?
PV may get cheaper, but so will wind. It's hard for me to imagine wind not having a significant cost advantage for a long time. What do you think?
"Because availability correlates extremely well with A/C requirements..."
Well, this relates to peak demand. It seems to me that more than very roughly 800-900GW of solar would take care of peak demand, and start to eat into baseload, inverting the normal relationship between day and night prices. If wind is cheaper than solar, does this make sense?
"it would take ~120 GW to replace the fuel used by gasoline-powered vehicles, assuming no efficiencies from electrification"
Seems like the straightforward calculation automatically takes into account efficiencies: .25KWH/mile x 200M vehicles x 12k miles/year = 75GW.
"replacing diesel would take roughly another 60 GW."
I should think that would be much lower, perhaps 10-15GW as a wild guess: IIRC diesel is only about 1/3 of light vehicle gasoline and most long-haul trucking would go to rail, which is 3x as efficient.
"Much of this use is during daylight hours."
Well, I included that in my mini-model above. If you have a battery of any size, like the Volt's, it's awfully easy to do most of that charging at night, to take advantage of cheaper wind.
oops - that should be ".25KWH/mile x 220M vehicles x 12k miles/year = 75GW"
And, let me offer this revised paragraph:
Well, this relates to peak demand. It seems to me that more than very roughly 800-900GW of solar would take care of peak demand, and start to eat into baseload, inverting the normal relationship between day and night prices. If wind is cheaper than solar (after taking into account marginal overhead for transmission and distribution), does this make sense?
I would think that solar is not a good fit for powering electric rail because freight pretty much moves 24x7. Am I wrong on this point?
Solar vs wind: Solar has advantages where the wind doesn't blow. So, for example, the US southeast is lousy for wind. Can transport losses decline enough for that not to matter?
Wind cost declines vs solar cost declines: I expect greater solar cost declines just because much of a wind tower is not wind tech per se. We've been building upward to support stuff for a very long time. I don't see how that's going to become radically cheaper short of nanotech. Ditto for the electric power generators. Whereas there are lots of ways to make PV material that are getting investigated.
Can concentrating solar decline price as much as PV? Concentrating solar seems far more amenable to storage so that generation can get shifted into the summer early evening peak demand hours.
" even capital cost of $15/average watt is pretty cheap for quite a few of those plants."
Could you elaborate on this? I'm not sure what you're referring to.
An NG plant may only cost $500/kW, but if it operates less than 584 hours a year you're up to $15 per average watt. At this point, PV is competitive based on investment alone... and has no fuel cost. If it's effectively guaranteed to be cranking on the hot, sunny days that would otherwise be served by the gas turbine, it's probably a better deal.
PV may get cheaper, but so will wind. It's hard for me to imagine wind not having a significant cost advantage for a long time. What do you think?
I'm inclined to agree, but as the cost of both wind and PV falls, transmission becomes a bigger and bigger factor for wind but not local PV. According to the Skeptical Inquirer article written by Primary Energy, the capital investment for new T&D can be as much as the generation it handles. PV at point of use eliminates all that.
My fuel vs. electric calculations are some 4.5 years old and didn't even bother to try to estimate the savings from vehicle improvements and mode shifts. Who knows, if we throw Bladerunner trucks with Zebra batteries onto electrified rails we may wind up increasing the amount of energy going to trucks! It's not something I'd project, but I wouldn't bet against it either.
"solar is not a good fit for powering electric rail because freight pretty much moves 24x7."
That's straightforward, though I think freight does move a bit more during the day.
"the US southeast is lousy for wind"
It's not that far from Texas wind. IIRC, % transmission losses arent' that high (7%?), even with A/C lines.
" I expect greater solar cost declines "
Wind gets more efficient/lower cost pretty quickly with increasing size, as we've seen over the last 30 years. Power is square of turbine size while cost is roughly linear to turbine size, and greater heights reach higher speed winds. I there's pretty good potential from kite-based and floating off-shore systems, as well.
"An NG plant may only cost $500/kW, but if it operates less than 584 hours a year you're up to $15 per average watt"
Yes, Natural Gas plants can have very low capacity factors, but they compensate with revenue from capacity payments. In any case, we're dealing with peak capacity here, right? We're roughly agreed that solar is likely to be mostly peaking power for the foreseeable future?
" If it's effectively guaranteed to be cranking on the hot, sunny days "
That sounds reasonable, though I have noticed some objections. Have any ISO's done capacity credit calculations for central PV plants?
"the Skeptical Inquirer article written by Primary Energy"
You wouldn't happen to have the url for that, would you?
"the capital investment for new T&D can be as much as the generation it handles"
Well, we have enough T&D to handle the peak of roughly 750GW. If we build enough PV to entirely lop off the peak, we shouldn't need much more than the existing T&D, pretty much forever.
The problem with solar for peaking comes in the evening when people come home from work. We'll still need natural gas for evening peaking. So the cost of the natural gas peaking plants will get amortized over an even lower number of operating hours, making evening electricity more expensive.
Of course there are two ways to handle that:
1) Use daytime solar electric to make ice to cool houses in the evening.
2) Use concentrating solar power to heat up sodium or other material to use to make steam to generate power in the evenings.
How much will each of these contribute?
Late morning in early June will probably be the cheapest time to buy electricity once solar falls far enough to supply 10-20% of total electricity.
"The problem with solar for peaking comes in the evening when people come home from work. We'll still need natural gas for evening peaking."
That's the traditional utility point of view, which is incentivized by a regulatory framework that's based on capital investment ROI. When faced with a peak demand, they think first of new generation, then of expensive central storage.
Demand Side Management is far better, faster, easier and cheaper. Charge based on time of day, and sell cheap timers that lower the thermostat on the A/C (as well as the fridge) at noon, instead of middle evening. Overall KWH consumption rises slightly (due to a larger differential between inside and outside temperatures), but this would be far more efficient than ice storage.
Also, PHEV/EV charging will avoid peak times - GM is working very closely with EPRI and a large array of utilities and DSM companies to make it work.
People will do simple things like reducing lighting in the evening, which will reduce lighting KWH as well as A/C. People will be creative.
Dynamic pricing could cut peak demand, certainly. The regulators just have to allow it. When's that gonna happen?
But even with dynamic pricing some people are going to come home and say "Hey, I've worked hard all day and I want to cool down" and they'll turn on the air conditioner. There'll be higher demand during some parts of the day.
Reduce lighting in the evening: This really undermines the utility of the light bulb. When do you need light? When it is dark and you are still awake. They aren't going to stay up later so they can turn on the lights at midnight and do more reading then.
"Dynamic pricing could cut peak demand, certainly. The regulators just have to allow it. When's that gonna happen?"
Actually, it's not only allowed, it's mandated by the energy bill of 2005 - all utilities have to be offering it now. I believe PG&E is aggressively rolling it out over the next several years. The stimulus bill throws some money at this as well.
The utilities aren't all that excited by it, because the capex ROI regulatory model is still in place for most utilities. That means that most are sticking at the pilot program point, where customers can have it on request. Here's an example: www.thewattspot.com .
"even with dynamic pricing some people are going to...turn on the air conditioner."
Sure, but 1) I wasn't talking about less A/C, I was talking about earlier A/C (and refrigeration) and 2) you only have to shift part of demand to move the curve earlier, to where solar shines (pun intended).
"Reduce lighting in the evening: This really undermines the utility of the light bulb. "
Yes, that's not the best example. My point was that much lighting isn't needed, and that if people pay a little more attention to turning out the lights when it's most expensive, that will make a difference.
I'd also note that if we move to residential PV, the capex needed for energy will become much more transparent to residential customers, and they're likely to realize that there are much higher ROI opportunities than PV, like better lighting, appliances and A/C. A/C in particular could be much more efficient with relatively small marginal investments at routine replacement points.
Rational responses to dynamic pricing: Sure, there will be some rational responses to dynamic pricing. Put solar power aside for the moment. Dynamic pricing will reduce the late afternoon and evening demand peak. How much? Hard to say.
But there's the twist: However much dynamic pricing reduces the demand peak it reduces the ROI for solar power. Solar's ability to compete is based in part on its ability to be sold at the time of day when prices are higher. It is not a perfect fit in that regard though. Solar peaks at noon. Demand peaks at late noon and early evening in the summer.0
As for turning on air conditioning sooner to reduce A/C demand later: I'm guessing the benefit is minimal without ice storage. But I'd like to know whether my guess is accurate.
"Dynamic pricing will reduce the late afternoon and evening demand peak. How much? Hard to say."
Yeah, we're going to have to experiment. OTOH, we haven't really tried residential time-sensitive pricing yet.
"However much dynamic pricing reduces the demand peak it reduces the ROI for solar power. "
Sure. That's my thinking, too.
" Solar peaks at noon. Demand peaks at late noon and early evening in the summer."
Yes, and proper pricing will help move demand to where it needs to be.
"turning on air conditioning sooner to reduce A/C demand later: I'm guessing the benefit is minimal without ice storage"
Many people turn off their A/C during the day, or turn up the thermostat, and then turn it on/down when they get home. That is an important part of the evening peak in demand. Well, the rise in temperature they're dealing with has been developing all day: the house has been accumulating heat.
That's a sensible strategy under flat pricing: allowing temps to rise in the home reduces the temperature difference between inside and outside, and reduces overall heat gain. A simple calculation shows us the effect: if it's 90 degrees outside, allowing the inside temp to rise from 70 to 75 (for an average difference of 2.5 degrees over the day) reduces heat gain by 2.5/20, or 12.5%. Well, ice storage is likely to be less than 70% efficient, giving a 30% loss which is 2.5x greater than earlier A/C. Also, ice storage has capital and operating costs.
Yes, ice storage has capital costs. Whether it will pay off will depend on the pricing curve and also how much air conditioning one uses. Harder to justify in Santa Barbara than in Palmdale.
Also, does ice production cost more for another reason? I'm thinking about heat pumps that are less efficient the higher the temperature gradient. If an air conditioner puts out 60F air does it use as much total energy as first cooling below 32F and then blowing air over it to produce 60F air later?
Also, the efficacy of earlier A/C activation will be driven in part by quality of insulation. A poorly insulated SoCal house will leak in the heat faster than a well insulated New England house with identical outside temperatures. Why bother cooling in advance if the leak rate is high?
IIRC the air conditioner usually cools much lower than 60°F just to guarantee dehumidification.
I can't see ice storage having such low efficiency. A short ton of ice is a cube roughly 1 meter on a side; let's call it 3.6 feet on a side to include residual liquid volume. Total stored cooling is 288 kBTU.
Total surface area of the cube is just shy of 78 ft². Figuring 2 inch foam at R-5/inch and average 100°F outside temp (Phoenix summer conditions), the ΔT is 68°F, the R-value is 10, and the heat leakage is 6.8 BTU/ft²/hr or 12.7 kBTU/day for the cube. That's a loss of around 5%.
Bigger ice volumes will have lower losses, and insulation can easily be increased to 6 inches. I cannot see leakage as a serious problem for ice storage A/C.
"I can't see ice storage having such low efficiency."
I'm not thinking about heat leakage, I'm thinking about system losses from creating the ice. As Randall notes, it's less efficient to cool to a lower temperature, plus you have a more complex system with greater operating losses.
I've seen ice storage in operation: it's a good idea, but it's not a walk in the park: more maintenance and complexity for the operators can make it fail where theoretically it's a good idea.
Of course, that's the advantage of leaving it to the central utility - simplicity for users. I like DSM a lot, but it has the same problem: more work for users. We're going to have to find ways to make it as easy as possible. KISS.
The system losses are offset by the ability to make use of the marginal watt from un-schedulable sources like PV; it lets you over-size the PV system compared to average use (to push more peaking supplies out of the market) and not have to worry about oversupply. You can run the ice-storage system without using the storage, and you can use it day or night (daytime if you have excess PV supply, nighttime if you have evening wind or for load-levelling during conditions of low RE production).
DSM doesn't have to put any burdens on the user. Ice storage is a perfect example; once it's installed, the utility can figure out when and how much ice needs to be made (and when to kick the dishwasher on), and the user doesn't even have to pay attention.
"The system losses are offset by the ability to make use of the marginal watt..."
Sure. I'm just pointing out that non-storage DSM (where useable) is cheaper, and that DSM in the form of moving A/C earlier in the day is probably the best way to deal with the fact that A/C peak demand is later in the day than solar peak insolation.
I saw an ice-storage system in a large office building - the operating engineers didn't especially like it, and eventually it was removed. This was partly cultural, of course, but part of the problem was additional operating and maintenance complexity.
"the utility can figure out when and how"
Yeah, that would probably be best for most consumers.