SoCal Stirling dish plant nears approval.

[skids]

"The dish generators are costly beasts -- $250,000 each. But that's because most have been hand-crafted in sporadic lots of one or two units. Building a group of 40 or 50 would trim the cost to $150,000 apiece, Osborn says, and the company has estimated that mass production could slash dish costs to $80,000, or perhaps just $50,000."

...

"When SoCal Edison said it wanted to buy more renewable energy, Osborn's outfit proposed the 500-MW project as the means of solving its chicken-or-egg impasse. Producing that much electricity will require 20,000 Stirling dishes, built in a steady flow over several years. "We're in the process of ramping up production now," says Osborn."

...

"Ultimately, 100 square miles of Stirling dishes could replace all the coal now burned to generate electricity in the entire U.S. -- if some dishes get coupled to systems that can store solar energy for use after sunset, such as massive flywheels and fuel cells. Whether that remains a utopian dream or emerges as a viable plan probably hinges on Stirling Energy's success in delivering on its deal with SoCal Edison. "

(...not sure if the math in the last paragraph was done by a energy professional, or by the journalist.)

http://www.businessweek.com/technology/content/aug2005/tc20050819_0041_tc024.htm

[NNadir]

1. 40 dishes, $150,000 each for 1 watt. 500 watts $3 billion dollars.

If it runs at 30% capacity loading (high for a solar scheme), it will be the equivalent of a $10,000,000,000 plant. Then you have to build another plant to provide power for when that one's not working, night time for instance.

[skids]

2. 20,000 dishes, $20K to $80K per dish.

$400,000,000 to $1,600,000,000 total cost is what the article infers.

This is a peak load plant, generating electricity for when it is needed for the region. Unless they add storage, which they probably wont until such a time as peak eletricity is no longer at a high premium.

[NNadir]

5. 20K to 80k/dish "COULD" be the cost. The current cost is much higher.

It is $250,000 each right now.

When calculating the fixed costs of a plant (ie construction costs, maintenance, insurance, salaries, etc) it is worth noting that irrespective of what a plant does and when it operates, the fixed costs do not stop. For instance, a bank does not stop charging the interest costs on a tire factory when all the workers clock out.

These plants are limited by the laws of physics, including the laws of planetary motion, to operating for a portion of the day.

The addition of flywheels or whatever will represent an additional economic and environmental cost which may be correspondingly huge.

Let us assume (generously) that the plant can be built with the still imagined cost of $20,000/dish as opposed to the current cost which is more than 10X higher. Then the plant costs $400,000,000. Let us also assume (generously) that the plant can produce for 1/3 of the time its peak rating of 500 MW. Then the total energy produced by the plant is 500,000,000*365.25*86400/3 = 5.2 petajoules. This is the equivalent of around 1.5 billion kilowatt hours. Let's say that the plant amortizes its cost over a period of 20 years. It will thus produce 30 billion kilowatt-hours of electricity.

This works out to a cost of $0.013/kw-hour, which is cheaper than coal, cheaper the nuclear, and cheaper than wind power. It is, in this case, a very good deal. (For comparison, the cost of generating nuclear power in the US is $0.018/kw-hour.)

http://www.eia.doe.gov/cneaf/nuclear/page/analysis/nuclearpower.html

If however, the cost can be reduced only to $80,000 (which is still speculative and not a demonstrated capability) a similar calculation shows that the cost is now $0.055/kw-hour or about 3X the cost of nuclear power.

If however the optimistic assumptions about reducing the cost of the Stirling mirrors isn't realized, and the costs are only reduced to say, $150,000/per mirror, the generating cost is now near the current retail price of electricity, $0.10/kw-hr. This is not such a good bargain.

All this of course, ignores maintenance, which will include among other things, damage to the construction materials from heat focused on the generation devices, dust deposited on mirrors, wear on electrical equipment such as inverters, wind damage that may cause misalignments and scratches on the mirror surfaces (sandblasting), so on and so forth. Moreover, there is no assurance that these plants will actually produce their rated power for 1/3 of the year on average. While they may not be as doddering as PV plants (15% capacity loading) they may still not reach more than 25% of capacity. On the plus side however, there is no reason to assume that the plants could not function for 40 years or longer, thus reducing their operating costs even further.

Overall, I think that the outlook for these stirling solar plants is respectable enough that the idea should be tried out and taken very seriously indeed. As you note, solar energy is well suited for meeting peak demand. Pricing of electricity based on grid load would make this idea even more attractive.

Although nuclear plants do not require the additional cost of expensive devices like batteries and flywheels to allow them to operate at night, they are nonetheless sluggish at meeting peak demands. Nuclear plants run best when they are fully loaded and continuously producing near or at their peak power rating. It may be possible to address this issue by using nuclear power for off-peak demand like aluminum manufacture, but this is at best, a partial solution at best to the fact that nuclear energy is ill suited to addressing short term power surges.

It is worth noting that solar ray focusing technology can be used to provide stored energy via thermochemical means. Although such use may not be generally applicable from an economic standpoint, it may work in niche markets.

[dcfirefighter]

7. Alcoa et al

have already moved to places with hydro

I tend to think that if peak pricing is charged, peaks will flatten -
day time cooling loads will be 'amortized' overnight with thermal mass.

Note also that the stats for the stirling mirrors only apply in the SW, they'd only be half / 2/3d's as productive in the rest of the country.

They demonstrated 29.4% efficiency vs incident light energy, desert southwest averages 7-8 kW-h / m^2 / day with a heliostat, which is equivalent to rated power 29-33% of the time.

I don't think that production costs dropping from $250,000/u to $75,000/u is unreasonable, IF the scale of manufacture is in fact on the order of several million (vs. one-off today).

I also don't think it'd be unreasonable to look for methods that are viable in the $0.12-16/kW-h retail range - no doubt energy prices are going to go up, whether by choice (internalizing externalities) or by burning all the cheap fuel, leaving only the expensive fuel.

[NNadir]

10. We need to differentiate between efficiency and capacity loading.

In solar-speak "megawatt" ALWAYS means "peak power under ideal conditions."

One condition which is not ideal is night-time. Another is when it is raining.

When you look at the energy produced by various solar schemes, it is always far less impressive than advertised.

The 0.15 factor that one uses to show that PV power advocates (and I know this is not another dumb PV scheme) are usually being misleading at best, and full of shit at worst when they crow about various solar installations, is not an effect of how much of the incident solar energy they convert to electricity. It is, instead, a function of the fact that conditions are "ideal" for a very small portion of the day for a portion of the year.

29.4% efficiency for the collection of incident light is great, but what is the capacity loading?

Can you provide links for the data you are using?

I more or less agree with everything else you said, by the way.

[dcfirefighter]

11. capacity loading and energy efficiency

the efficiency figure came from the businessweek article: http://yahoo.businessweek.com/technology/content/aug2005/tc20050819_0041_tc024.htm
SUNNY OPTIMISM. For years, Stirling dishes have been roughly twice as efficient as the best solar cells, even before Stirling Energy took over the technology in 1996. Today, a Stirling dish converts 29.4% of the sun's energy into electricity, according to tests by Sandia National Laboratories at a six-dish operation in Albuquerque.


A couple of other useful quotes:

from http://www.energylan.sandia.gov/sunlab/faqs.htm
Question:
Do concentrating solar plants require a lot of land?


Answer:
Relatively speaking, no. Consider the Hoover Dam. Lake Mead covers nearly 250 square miles. A Concentrating Solar Power system occupying only 10-20 square miles of land could generate as much power on an annual basis as the Hoover Dam did last year. Considering the land required for mining, concentrating solar power plants also use less land than coal power plants.



from http://www.stirlingenergy.com/faq.asp?Type=all
How much does the Stirling solar power cost?
That depends on the scale of the power plant. Based on a power plant producing 1,000 MW, the cost per kWh would be approximately six cents. Compare this with the 22.5 cents per kWh on average that customers in California paid for power is 2001.


How much power does one of these Stirling dishes produce?
One dish on an annual basis can produce 60,000 kWh of electricity. Put another way, a solar dish farm of 10.8 sq. miles could produce 2,100 MW of electricity annually – as much as the Hoover Dam, which uses 247 sq. miles of land (including Lake Mead).

Using this figure the 25kW dish producing 60,000 kWh/y is a 27.4% power loading. I had figured 7 to 8 kWh/m2/day annual average insolation from the DOE's ">solar insolation map with 2 axis tracking Apparently I was a little off, because I assumed the solar unit's power rating was vs. 1kW/m2 insolation, though perhaps they also accounted for non-solar related down time.

[NNadir]

12. Not one link gives the production in units of energy. All give power.

Power is not energy.

A bolt of lightening produces about one billion watts of power, 1000 MW, about as much as a typical nuclear power plant. However it does this over a few thousandths of a second. Thus it is useless to discuss powering a city by the use of lightening.

http://www.cltskywarn.org/Specsky.htm

I am increasingly suspicious of technologies that do not advertise themselves in units of energy, choosing power instead. It's become a latter day form of doublespeak and a means of avoiding the measurement of real costs, environmental and otherwise.

[dcfirefighter]

13. One dish on an annual basis can produce 60,000 kWh of electricity

I had included that link specifically for you, knowing your preference for units of energy vs. units of power.

[NNadir]

15. I'm sorry, but I don't see that link.

However, taking your word for the 60,000 kW-hr figure, we see that this is the equivalent of 216 billion joules. Multiplying this by the 40 dishes that are said to produce 1 "megawatt" we see that the actual energy output of 40 dishes is 8.6 trillion joules. The amount of power produced by 1 (real) megawatt of power operating at full capacity is 31.6 trillion joules.

Now we have the loading capacity of the plant, which is simply 8.6/31.6 = 0.27 to two significant figures. This, 27%, is very high for a solar plant, higher than any of the 50 largest solar plants in the world but extraordinarily low for any other kind of plants. There is one major exception, which is oil and gas fired plants, which operate at 27% of capacity loading, probably because most of these plants are, in fact, designed for peak loads.

http://www.eia.doe.gov/cneaf/nuclear/page/analysis/nuclearpower.html

This suggests that the optimal use for these solar plants is roughly comparable, as a peak demand load meeting plant, as you and others have suggested.

It is also relatively easy to calculate the cost of the dish by itself, neglecting all other infrastructure costs, the cost of maintenance, the cost of inverters, the cost of land, transmission lines and - this is the big one - energy storage devices that may be employed. If the dishes cost $20,000 each, ie, their cost is reduced from current cost by 92%, the dishes cost (amortized over 20 years) $0.017/kW-hr, which is quite reasonable. If the cost is reduced to $150,000/dish, the cost of the dish alone is 0.125/kw-hr, which is very, very expensive by current standards, though still less than 1/3 the cost of magical residential PV power in a sunny climate, and slightly less than 7 times cheaper than magical residential solar PV power in a cloudy climate.

http://www.solarbuzz.com/SolarIndices.htm

[jpak]

20. No, nuclear plants need expensive "devices" like

uranium mines and mills

uranium conversion plants

uranium enrichment plants

and spent fuel and depleted UF6 repositories

most of which are subsidized by the taxpayers...

[htuttle]

22. We need to stop thinking about single source solutions

Nothing can take petroleum's place in our energy appetite by itself. Not solar, not wind, not biomass, not nuclear. We have to start thinking about producing electricity by sustainable means, everywhere possible (and practical). Conservation HAS to be a very large part of this, or there's no hope at all. And conserving our environment obviously has to be another big consideration, or the energy will end up being used for disaster relief efforts instead of running television sets and cooking stoves...

As far as Stirling dishes goes, they've always been one of my favorite means of using solar energy directly. The operating costs are very low, and the technology itself is very basic compared to many other energy systems. They are fairly scalable, too. Stirling engines come in almost every size imaginable, from hand-held novelties to huge installations like the one mentioned in the article.

[Oerdin]

9. I have to agree with you.

Solar is a rich man's toy which is economicly hopeless. We get greens and other people who haven't bothered to seriously look at the numbers and compare it to every other option saying how wonderful solar is but the reality is after decades of subsidies & investment very little real progress has been made and it remains horrendously inefficent and no where even close to cost effective on a per megawatt basis.

Even the co-founder of green peace has given up on solar as a serious way to fill out energy needs and is now backing nuclear power as the only way we can continue to meet our energy needs while not producing green house gases.

[dcfirefighter]

14. Solar isn't economically hopeless

it's just unsuitable for the majority of applications. It is however very suitable for remote power applications, as well as peak load generation in some climates.

Many alternatives become more economically viable as energy prices rise - more so if externalities are considered or better, internalized.

Personally, I see an increase in gas, nuclear, and even partial solar cogeneration, as well as an increase in wind generation. I don't think we'll see many new oil plants (though many existing ones will be converted to gas). I don't think any of the 'clean' coal options are viable without subsidy and a severe blind eye.

[jpak]

21. LOL!!!

I guess that's why Japan and Germany are investing heavily in PV module manufacturing facilities.

...and I guess that's why Portugal is building 64 MW and 116 MW PV farms...

...and I guess that's why global PV module production is increasing at ~25-40% per year.

It's "economicly" hopeless!!!!!

:rofl:

[meow mix]

23. heh heh heh

"rich mans toy" "idiot greens" blah blah blah.

someone has been training parrots im not mentioning any names tho.
LOL!!!!

[dcfirefighter]

3. 29.4% efficient, according to Sandia NL

25kW (peak) @ $250,000 = $10 / W (peak)
25kW (peak) @ $150,000 = $6 / W (peak) (groups of 40)
25kW (peak) @ $80,000 = $3.2 / W (peak) (mass production)
25kW (peak) @ $50,000 = $2 / W (peak) (optimistic mass production)

Mid range mass production costs = $2.6 / W (peak)
500 MW (peak) at midrange production costs = $1.3 Billion

According to the manufacturer's website, units are spaced at about 10 units per acre.

Also according to website, a 100 mile by 100 mile (10,000 square miles) site could supply ALL of the US's electrical needs.

10,000 sq. miles is 6,400,000 acres, and would hold 64,000,000 25 kW (peak) units, producing 1.6 TW (peak).

According to CIA US electricity consumption is 3.66 Trillion kW-h, or 13 x 10^18 J, which would require the above solar farm to have a capacity loading of 26%, not impossible in the desert southwest with a heliostat.

At this capacity factor and using the midrange mass production costs, a 1000 MW (average) plant would cost $10 Billion in construction plus 15,400 acres.

[phantom power]

4. 3.66 Trillion kilowatt-hours per year?

[dcfirefighter]

6. per CIA via google

http://www.google.com/search?hl=en&q=us+electricity+consumption

United States — Electricity - Consumption (KWH): 3,660,000,000,000
http://www.cia.gov/cia/publications/factbook/rankorder/2042rank.html - More sources »

[phantom power]

8. Those numbers work, if we also have storage for night time, etc.

I think I estimated something like 4 terawatt-hours of storage we would need to deploy, for covering the hours of the day not conducive to solar production. And that didn't really take into account the issue of handling a long string of cloudy days, where energy production would be sub-optimal for an extended period.

[IDemo]

16. The thing people invariably fail to factor

about these units is that solar insolation is not the only potential source of heat energy for the Stirling genset. This system was tested at the Pima-Maricopa Indian reservation near Tempe, Arizona a few years ago, and relied on landfill methane and/or natural gas at nightime or on cloudy days.

[skids]

17. It's astounding, isn't it?

...how you can buy a hefty gasoline generator for under a grand, but try to find any type of stirling anywhere near that price... and... well, keep dreaming.

[Massacure]

18. Probably mass production?

There are millions of gasoline generators, but how many stirlings?

[skids]

19. Second contract signed.

Update:

PHOENIX, ARIZONA, USA – Stirling Energy Systems (SES) recently announced a contract with San Diego Gas & Electric (SDG&E) to provide between 300 and 900 megawatts (MW) of solar power -- approximately 30 times more solar power than is now being generated in the San Diego region. This contract represents the second record-breaking solar project signed by the company in the past month, which may surpass the earlier contract to become the world's largest solar installation.

...

SES General Manager Robert Liden points out that “the output from these solar dishes is a great fit for SDG&E's customers because a large percentage of the power is produced when it's needed the most – during the peak demand hours in the middle of the day.�

http://pesn.com/2005/10/12/9600186_Stirling_300MW/

[skids]

24. Approval granted.

Update2: approval has been granted.



A solar project of unprecedented size moved an important step closer this week to being built. This week the California Public Utilities Commission (PUC) approved a contract for the initial deployment of 500 MW of concentrating solar power (CSP), with expansion options up to 850 MW.



Expected "capacity factor" on this technology/site will be 29.3%

http://renewableenergyaccess.com/rea/news/story?id=38672

[skids]

25. UPDATE: Yet more approval granted.

I guess the contracts themselves also needed approval. Looks like the next hurdle is an upgrade
to transmission facilities -- it's not quite explicit how large the plant can get before the
upgrade and how much capacity hinges on the new transmission line.

An unprecedented deployment of concentrated solar thermal in California took one step closer to becoming a real project this week as state regulators gave their approval for the power purchase agreement. Completion, however, will hinge on other factors, including construction of a new proposed 100-mile long, high capacity transmission line that could cost as much as 1.4 billion dollars.

http://www.renewableenergyaccess.com/rea/news/story?id=40914