Although it’s been operating since June 2007, today Acciona dedicated their 64 megawatt solar thermal plant in Boulder City, Nevada. According to the Chairman of Acciona, José Manuel Entrecanales, the plant cost $260 million and will operate for several decades. The plant is rated to produce 64 megawatts in full sun and is expected to produce about 130,000 megawatt-hours per year.
 |
| A parabolic trough at Nevada Solar One. (Photo: Acciona) |
While there have been solar thermal plants installed in recent years, particularly in Spain where Acciona has their headquarters, this is the first utility-scale solar thermal plant built in the USA since 1991.
It is interesting to compare this plant to some of the others being proposed. Ausra, for example, expects to generate 177 megawatts on a 640 acre installation - using their modified solar trough technology. Nevada Solar One occupies 400 acres.Â
This means Ausra claims their plant footprint will require 3.62 acres per megawatt output, whereas Acciona’s plant footprint requires 6.25 acres per megawatt - nearly twice as much. The comparison is misleading, however, since Acciona’s plant has a much greater ratio of balance-of-plant to solar field compared with Ausra’s. Also, of course, Ausra’s plant is still on the drawing boards, and until power is being generated to the grid, it isn’t fair to compare a design to a reality.
Another way to look at the footprint of solar electricity is to compare solar thermal to photovoltaic - and in this comparison, photovoltaic displays far more than the 2 to 1 range of efficiencies we’re possibly seeing with solar thermal. Thin film photovoltaics, which display efficiencies as low as 5%, require 4.6 acres per megawatt. Top-end crystalline photovoltaics, on the other hand, are available off the shelf at efficiencies of over 20%, which means they would require only 1.2 acres per megawatt.
If one assumes Californians draw about 30 gigawatt-years of electricity per year (and that’s on the low side if we start charging millions of electric cars every night, read “Gigawatt-hours per electric commuters“), then here’s how much land in California’s southern deserts would have to be given over to solar installations:
Solar thermal - best case:Â 731 square miles.
Solar thermal - conservative case:Â 1,278 square miles.
Photovoltaic - thin film low efficiency:Â 939 square miles.
Photovoltaic - crystalline high efficiency:Â 235 square miles.
If this sounds like a lot of square miles, it isn’t. California’s area is 158,000 square miles. Put another way, since California has 36 million inhabitants, 1,000 square miles, or 640,000 acres, equate to only 774 square feet per person. Much of California’s solar future will be via rooftop installations, which on a per capital basis almost certainly exceeds 774 feet per person.
Not addressed here are the storage issues facing solar power, and all intermittant power. It is no coincidence Nevada Solar One is located in Boulder City, in the shadow of Hoover Dam. If and when solar thermal power scales up into the multiple gigawatt output range, load balancing as the sunlight fades can be accomodated simply by reactivating one more turbine in the powerhouse. As a supplement to hydroelectric power, solar thermal even without storage solutions is very interesting, because the solar field can provide power during the day, allowing the hydroelectric turbines to remain idle until night-time. This can literally double the annual power-output capacity of deep water reservoirs, since the supply of water is often the primary constraint on their output.
February 22nd, 2008 at 7:36 pm
The amount of land for photovoltaics cannot be calculated as claimed. Nor can photovoltaic produce electricity to meet peak demand. Thus it’s of little value as the demand for power grows, which will be true for the next 20 years, at least. Photovoltaic is also more expensive , in addition to having far less value. I’m sick and tired of hearing “experts” discuss various alternative power generation schemes who obviously know virtually nothing of the grid’s requirements. Attempting to cost out generators such as photovoltaic without taking into consideration that duplicative generators will always be needed (thus essentially doubling their cost) is just plain incompetent.
February 22nd, 2008 at 7:39 pm
The amount of land required is of practically no concern. There is enough land in just 10% of the Federal land in Nevada to produce all the power the entire country needs using an Ausra type of solar thermal generation scheme. It is also FAR, FAR cheaper than solar photovoltaic, which has no ability whatsoever to meet peak demand.
February 22nd, 2008 at 10:14 pm
Kent: It’s great that you keep us honest - but the point of the post was that in the case of a solar thermal generating plant existing in the shadow of Hoover Dam, load balancing is quite easy. You have 1.5 gigawatts of power from Hoover Dam that can be attenuated simply by taking one turbine offline. To the extent solar electricity is being fed into the grid during the day, you take hydroelectric turbines off line. At night you reactivate them. True?
March 23rd, 2008 at 10:37 pm
Solar thermal plants are not only far cheaper than photovoltaic cells, but they can easily store the heat generated for nighttime power generation. You can’t do that with other solar options and obviously a bank of mirrors is far cheaper than solar cells that use exotic minerals and cost-intensive manufacturing.
June 13th, 2008 at 6:03 pm
Kent is right, Ed. Your math is “voodoo math” when it comes to determining comparisons between various types of generating options. Facilities are designed based upon peak demands, which don’t correspond well with PV. You do a disservice to the solar industry putting out these simplistic calculations without the knowledge to understand what they mean. This is the way dis-information gets started and spread, confusing everyone who is trying to understand the problem. Shame on you.
June 13th, 2008 at 6:19 pm
Gene: If your specific concern is storage, that is addressed in the last paragraph of the article. Obviously we aren’t going to do well to rely on solar energy if the solar peak is noon and the demand peak is 7 p.m. Everybody knows that. With wind, it’s even worse. The point of this post was simply to debunk the notion that we need to carpet the whole country with solar fields in order to provide meaningful amounts of energy. Finding space for solar isn’t the challenge. The challenge with solar power is load balancing, storage and transmission - and cost, of course.