Land Area Requirement for Concentrating Solar Plant (CSP) in Promising Areas
I recently read a blog post titled The world’s first molten salt concentrating plant that introduced me to Archimede Solar, a subsidiary of the Angelantoni Industrial Group S.p.A., with Siemens as a participating partner.
The first of a kind plant has a peak rating of 5 MWe and cost $60 million. That is a pretty steep price for a power plant on a per kilowatt of capacity basis, but this is, after all, a FOAK plant that must absorb first of a kind costs. I cannot hold that against the technology, yet.
However, while looking at the technical sales information available on Archimede’s site, I found an interesting section that needs to be better understood by those who tout the wonderful advances that are being made in concentrating solar plant (CSP) technology. It is a description of potential areas where CSP might be installed and contains numbers that do not immediately jump out as items of concern. However, they should contribute to a suspension of enthusiasm.
The southwestern part of the United States is in fact one of the world’s best areas for insolation and the Mojave Desert receives the sunlight up to twice respecting other regions of the Country.
There are several solar power plants in this area which supply power to the electricity grid. These plants have a combined capacity of 354 megawatts (MW) making them the largest solar power installation. Typically, individual CSP plants are between 50 and 280MW in size.
Other promising areas of the world to apply CSP technology include Southern Europe, Northern Africa and the Middle East, parts of India, China, and Australia. These regions have peculiar territorial features as large amounts of atmospheric humidity, dust and fumes so that 1 sq km of land is enough to generate 100-130 gigawatt hours of solar electricity a year, using solar thermal technology.
I am not quite sure how several solar plants can have a combined capacity of 354 megawatts and have individual capacities of between 50 and 280 MW; that would yield a maximum of three plants – one with 280 MW, one with 50 MW and one with just 24 MW.
The other set of numbers worth deconstructing is the 1 square kilometer of land to generate 100-130 gigawatt-hours of electricity. If the CSP operates like a baseload plant at 90% capacity factor, that means it would be rated at about 16 MWe. If it is run as a load following plant with an average annual capacity factor of 30%, it would have a rating of about 50 MWe. If the CF is much lower than that, one would have to question the value of having any storage at all; a typical solar system without storage can achieve a CF of 20% in a decent solar area.
The peak solar insolation for a 1 square kilometer piece of land near the equator should be approximately 1 GW (assuming 1000 watts per square meter). That indicates that the molten salt CSP in the target areas will only be collecting about 5% of the available solar energy. This factor should be included when evaluating claims like those found in Natural Gas as Panacea: Dubious Path to a Green Future a recent blog post by a university professor making the following statement:
Currently available wind and solar energy technologies, on the other hand, are up to the job right now. There just aren’t enough wind and solar installations, so today they provide less than 1 percent of the nation’s energy. We will need to rapidly scale up, so that by 2050 we can receive the majority of our energy from wind and solar power. That’s an enormous task: The U.S. Census Bureau forecasts that our population will reach 440 million by 2050 — nearly a 50 percent increase from today. That’s 150 million more people, each hoping to live at the standard of living we have grown accustomed to. When that happens, the amount of fossil fuels we use today, and which provide 86 percent of America’s energy, would provide those 440 million with less than two-thirds the energy they would need, if per-capita energy use remains the same as today.
Contrary to standard criticisms of solar and wind, providing this much energy in the future would not use up a lot of land. Based on current installations, less than 1 percent of U.S. land area would be required. Right now, 22 percent of U.S. land is in agriculture, not counting grassland pasture and range used by grazing animals.
One good idea has come out of concentrating solar plants-the molten salt storage of heat. This needs to be applied to all the electric power plants to cover variation between between generation and consumption. I would particularly recommend it for coal and nuclear plants. As far as India is concerned, it is quite densely populated, with fierce opposition to any change in land usage. The recent examples are abandonment of a nearly complete car plant (which could have brought in jobs!) and the richest uranium deposit. Solar energy in India can be only local and diffused.
“These plants have a combined capacity of 354 megawatts (MW) making them the largest solar power installation.”
This refers to the Solar Energy Generating Systems, built in the last solar boom.
http://en.wikipedia.org/wiki/SEGS
Typically, individual CSP plants are between 50 and 280MW in size.
This refers to the new generation of solar thermal plants. 50 MW is a very popular size in Spain; as I understand it, larger individual plants get subsidized at a lower rate.
http://en.wikipedia.org/wiki/List_of_solar_thermal_power_stations
Do I do the math correctly when I calculate this as the equivalent of spending $19,200,000,000.00 ($19.2 billion) for nameplate capacity equivalent to that of an AREVA 1.6 GWe EPR (1600/5 x 60,000,000), and, if we impute the projected 42% “plant load factor” that Mr. Ombrello mentions in his response comment to you as the “capacity factor” of this plant, and presume a 90% capacity factor on the AREVA plant (fair, unless it is operated in a load following fashion) then (90/42 x $19.2 Billion) the equivalent cost of capacity (ignoring the 25 year life of the CSP plant versus the 80 year expected life of the AREVA EPR) for an increment of this solar FOAK equivalent to the AREVA FOAK is over $41 billion in capital cost? And this for a plant that lasts 25 years as opposed to the AREVA plant projection of 80 years?
Excluding the transmission upgrades/construction likely required to move the kWh from the busbar to the point-of-use?
If my math is correct, then: wow. If not, someone please correct me, if you are willing.
I note that the AREVA Olkiluoto plant has been roundly criticized by anti-nuclear activists as demonstrating how nuclear is “just too expensive.” It is now projected to come fully online in 2013, and is also a FOAK.
Does anyone know where Mr. Ombrello derives the cost reduction analysis necessary to arrive at his projected final cost – – identified in his responsive comment to Rod – – and what the time frame is in which they are contemplated to be accomplished? NREL predicted that the cost of CSP would be below 8
I find it interesting that it’s a ‘university professor’ who makes these statements and brushes aside the impracticality of wind and solar installations at industrial-scale. He seems to follow the well-worn path of other Malthusians who predict the worst and are rarely, if ever, held to account for their patently incorrect pronouncements.
Some examples and food-for-thought:
http://rayharvey.org/index.php/2010/01/peak-oil/
http://www.21stcenturysciencetech.com/Articles%202008/Energy_cost.pdf
http://www.21stcenturysciencetech.com/Articles%202005/Nuclear2050.pdf
Seldom discussed is transmission technology or from the “Busbar to the Point-Of-Use. When talking Solar or Wind energy in the form of KiloWatt Hours at point of use something is missing. I live near the Horse Heaven Hills wind farm operated by Puget Sound Energy near Ellensburg Washington. Seattle is the point of use location ( 100 miles distant ) for the small 2 MW capacity of this facility. If I could put those generators on 24/7 in tandem with combined cycle Nuclear Fusion Turbines at or near the point of use from mobile rail cars I could make the investors in the Aussy Investment Group that owns Puget Sound Energy very happy.
I did notice an “auxiliary heater” on the diagram of the plant. It would be interesting to see how often – and when – this device is proposed to be used.
You do the math correctly, but your input numbers are wrong. Rod incorrectly reported the cost as $60 million, when (e.g. in the article) it is
So, I guess: wow, holy moly, yikes, etc. Why is it that we have to constantly (I gave a talk the other day, and this was a BIG QUESTION) respond to the “well-known fact” that “nuclear is just too expensive” ?
@Frank Jablonski – I think part of the reason that amateur advocates for atomic energy have to keep responding to the “well known fact” that “nuclear is just too expensive” is because our opposition has no compunction about lying. Unfortunately, another part of the reason is that the CPAs, lawyers and MBAs that are often at the top of established firms that have some nuclear component keep investing their money into lobbying Congress for more subsidies rather than spending it to tell Americans that nuclear energy offers a path towards massive quantities of cheap energy.
Sure, it is eminently possible that the potential for cheap energy will never be developed. After all, unless you have the mindset of the people who brought us ever less expensive and ever higher quality computing and communications experiences, it is perfectly possible to surround the potentially disruptive raw material with massive layers of expense. The revolutionaries who saw what semiconductor transistors COULD possibly do, inspired by people like Gordon Moore, aggressively worked to drive down a higher capacity, lower cost curve that put some dinosaurs out of business. A few of the very large companies – like IBM – have survived and even prospered, but most could never figure out how to profit in a market where prices kept decreasing.
I believe a lot of the companies currently dabbling in nuclear technology are dinosaurs who are really in the established energy industry where increasing prices is the only way they know to increase their profits. However, since their product is everyone else’s raw material, that mode is doomed and those of us who know that even with all of its processing, commercial nuclear fuel costs about 1/10th as much as natural gas need to keep telling people that fact. When it comes right down to the thermodynamics and the material inputs required, there is no reason at all why a nuclear heated machine should NECESSARILY cost more than a natural gas heated machine.
That is a disruptive thought, however, that threatens a very powerful industry. The neat thing is that it offers HOPE to billions of people who are not part of that industry.
The purpose of solar is to provide glossy pictures. Judging by the amount of electricity produced, solar is is all environmental impact and no electricity. Solar is mickey mouse and it should be pointed out not a real rodent but a fantasy. We all love Mickey but having solar panels on your rood is more expensive that having rodent infestation.
There is one constant. Descriptions of solar projects do not really expect an engineer to ask how much electricity is produced. The public relations values has nothing to do with generating electricity.
Very few utility scale projects (like Springerville) provide actual data on generation. Fewer still come close to producing electricity near the design values.
How corrosive/toxic is the salt used in these facilities, especially at high temperatures? Power generation with steam gets inefficient rather rapidly below 550 F (300 C). At these temperatures the corrosive properties could be a problem. Isn’t this one of the reasons liquid sodium was abandoned?