Last year on my other blog, bartoncii, I undertook a study of future post-carbon energy sources. I focused primarily on solar and wind generated electricity. My research method was fairly simple, and is based on Karl Popper's theory of knowledge. I set out to test what renewable advocates were saying about solar and wind generated electricity, by looking for information that would tend to disprove their claims. Evidence had to be reasonably construed as factual and had to clearly contradict the claims of wind and solar advocates.
The starting point of my investigation was the intermittent nature of solar and wind power. Could the problem of intermittency be overcome? If not, what would justify a massive implimentation of solar and/or wind power generation sources be justified? Another question had interesting implications. Would some of the objections to nuclear power also apply to "renewable electrical resources.
What I found was that installation costs of renewable electrical generation was not cheaper than nuclear energy if judged by actual electrical output rather than name plate generating capacity. nuclear power plants are simply more reliable producers of electrical energy. In order to make up for the unreliability of renewable electrical sources, generating capacity needs to be replicated from 2.5 to 5 times, and the surplus energy placed in storage, to be retrieved for electrical generation when energy output drops. The cost of such redundancy was such that it costs as much to build solar and wind generation capacity that would actually have the potential to displace current fossil fuel generating systems, as it would to displace those systems with nuclear power, without factoring in the cost of energy storage and retrieval, required to compensate for intermittency.
None of storage options I looked at were inexpensive. In fact, options like battery storage or pump storage were not significantly cheaper than the cost of constructing nuclear power plants, and offered inferior benefits. The rub was that if one took the superior nuclear option - superior in terms of energy returned for the investment dollar, and superior in terms of EROEI - there was little room for solar and wind generated electricity in the post-carbon energy world.
Hence the renewable electrical generation systems I looked at were going to be far more rather than less expensive than nuclear electrical generation systems. Let me look at for a moment comparative costs. The cost of an installed large land based wind generator is now running at around $2.00 per name plate watt. The cost of a sea based wind generator is considerable more expensive. The most regular winds in the United States will produce about 40% of a windmills rated energy. So if you want to average producing 1 billion watts of electrical energy from a wind generating system that produces at 40% of name plate capacity, you will need to build windmills with a name plate capacity of 2.5 billion watts. This would be the best case scenario. At $2.00 a watt, the generating system would cost $5 billion. In addition, if you wanted to produce electricity 24 hours a day with this system, you would need a storage system capable of storing the equivalent of 15 GWh of electricity. A pump storage system would to that, but at todays construction cost, that pump storage system could well cost another 5 billion dollars.
Hence the cost of a carbon free wind generating system under a best case scenario would be around 10 billion dollars per GW of electrical reliably delivered. This would be overnight costs, and does not factor in the cost of interest. Systems using offshore winds would be considerably more expensive.
What would building the equivalent nuclear generating capacity cost? Paul Bowers of the Southern Co., which owns Georgia Power, in early April estimated that building Westinghouse's AP1000 reactors would cost Georgia power between $2,500 and $3,500 per kilowatt. Westinghouse has recently estimated the kit price of AP-1000s to be something more than $1.50 per watt. An estimated 16 million to 20 million man-hours to build an AP-1000 reactor according to Bowers. TVA estimates that the cost of its first two AP-1000's will run between $2.5 and $3.0 billion on partially prepared sites. Thus $3 per watt overnight cost will get you a reactor, and of course interest must be added. Thus reactors are 40% less expensive than equivalent wind generating capacity, and three times less expensive than a system that would deliver an equivalent amount of carbon free wind generated electricity 24 hours a day.
Similar cost problems confront PV electrical generating systems, and many ST - also called CSP - systems.
Not all of my conclusions about renewable energy were negative. Solar systems are cost effective for water heating in much of the United States and for space heating in parts of the United States. The use of solar heating for space and water heating would tend to decrease the difference between day and night electrical demands.
Since I did this assessment, new claims are being made about other renewable electrical generating technologies. In a recent article in Salon, Joseph Romm argued for the economy of CPS. Romm claimed "According to a 2008 Sandia National Laboratory presentation, costs are projected to drop to 8 to 10 cents per kilowatt hour when capacity exceeds 3,000 MW."
The Sandia cost estimates are not backed by published data, and significant questions remain. Construction costs including the price of building materials have doubled during the last five years, and can be expected to continue to rise. It is not clear to what extent Sandia based their cost estimate on building realistic cost estimates. Romm claims, "[t]he technology has no obvious bottlenecks and uses mostly commodity materials -- steel, concrete and glass." but that is exactly the rub, the cost of steel, concrete and probably glass have doubled in the last 5 years and can be expected to inflate rapidly. Research on CSP technology and construction of CSP facilities are regionally concentrated in the American Southwest. In fact the desert climate of the Southwest means that useful sunshine will be available for an average of 8 hours a day or even more. In other parts of the country the availability of sunshine drops. Two thirds of the United States can expect an average of 5.5 hours of sunshine a day or less. Even worse cloud cover is not uniformly distributed in time, so in many locations clouds may obstruct the sun for several days at a time.
Thus CSP, even if it turns out to be a cost effective electrical generating source for the Southwest, still might turnout to be not cost effective for the rest of the country. Romm in the best tradition of renewable energy advocates, completely ignored the regional nature of CSP. Romm tells us, Finally, we will need more electric transmission in this country. "The good news is that because it [CPS] matches the load most of the day and has cheap storage, CSP can share power lines with wind farms. When the country gets serious about global warming, we will need to get serious about a building a transmission system for a low-carbon economy. "
In short good old Joe wants to build a lot of extremely expensive and highly vulnerable to terrorist attack electrical transmission lines to export CSP generated power from the Southwest to the rest of the country. How much would Romm's national CSP electrical generation system cost? No one knows, but it is a good assumption that building a thousand or so AP-1000 would be cheaper, and building several thousand LFTR's in a factory and using innovative siting approaches would be cheaper still.
The starting point of my investigation was the intermittent nature of solar and wind power. Could the problem of intermittency be overcome? If not, what would justify a massive implimentation of solar and/or wind power generation sources be justified? Another question had interesting implications. Would some of the objections to nuclear power also apply to "renewable electrical resources.
What I found was that installation costs of renewable electrical generation was not cheaper than nuclear energy if judged by actual electrical output rather than name plate generating capacity. nuclear power plants are simply more reliable producers of electrical energy. In order to make up for the unreliability of renewable electrical sources, generating capacity needs to be replicated from 2.5 to 5 times, and the surplus energy placed in storage, to be retrieved for electrical generation when energy output drops. The cost of such redundancy was such that it costs as much to build solar and wind generation capacity that would actually have the potential to displace current fossil fuel generating systems, as it would to displace those systems with nuclear power, without factoring in the cost of energy storage and retrieval, required to compensate for intermittency.
None of storage options I looked at were inexpensive. In fact, options like battery storage or pump storage were not significantly cheaper than the cost of constructing nuclear power plants, and offered inferior benefits. The rub was that if one took the superior nuclear option - superior in terms of energy returned for the investment dollar, and superior in terms of EROEI - there was little room for solar and wind generated electricity in the post-carbon energy world.
Hence the renewable electrical generation systems I looked at were going to be far more rather than less expensive than nuclear electrical generation systems. Let me look at for a moment comparative costs. The cost of an installed large land based wind generator is now running at around $2.00 per name plate watt. The cost of a sea based wind generator is considerable more expensive. The most regular winds in the United States will produce about 40% of a windmills rated energy. So if you want to average producing 1 billion watts of electrical energy from a wind generating system that produces at 40% of name plate capacity, you will need to build windmills with a name plate capacity of 2.5 billion watts. This would be the best case scenario. At $2.00 a watt, the generating system would cost $5 billion. In addition, if you wanted to produce electricity 24 hours a day with this system, you would need a storage system capable of storing the equivalent of 15 GWh of electricity. A pump storage system would to that, but at todays construction cost, that pump storage system could well cost another 5 billion dollars.
Hence the cost of a carbon free wind generating system under a best case scenario would be around 10 billion dollars per GW of electrical reliably delivered. This would be overnight costs, and does not factor in the cost of interest. Systems using offshore winds would be considerably more expensive.
What would building the equivalent nuclear generating capacity cost? Paul Bowers of the Southern Co., which owns Georgia Power, in early April estimated that building Westinghouse's AP1000 reactors would cost Georgia power between $2,500 and $3,500 per kilowatt. Westinghouse has recently estimated the kit price of AP-1000s to be something more than $1.50 per watt. An estimated 16 million to 20 million man-hours to build an AP-1000 reactor according to Bowers. TVA estimates that the cost of its first two AP-1000's will run between $2.5 and $3.0 billion on partially prepared sites. Thus $3 per watt overnight cost will get you a reactor, and of course interest must be added. Thus reactors are 40% less expensive than equivalent wind generating capacity, and three times less expensive than a system that would deliver an equivalent amount of carbon free wind generated electricity 24 hours a day.
Similar cost problems confront PV electrical generating systems, and many ST - also called CSP - systems.
Not all of my conclusions about renewable energy were negative. Solar systems are cost effective for water heating in much of the United States and for space heating in parts of the United States. The use of solar heating for space and water heating would tend to decrease the difference between day and night electrical demands.
Since I did this assessment, new claims are being made about other renewable electrical generating technologies. In a recent article in Salon, Joseph Romm argued for the economy of CPS. Romm claimed "According to a 2008 Sandia National Laboratory presentation, costs are projected to drop to 8 to 10 cents per kilowatt hour when capacity exceeds 3,000 MW."
The Sandia cost estimates are not backed by published data, and significant questions remain. Construction costs including the price of building materials have doubled during the last five years, and can be expected to continue to rise. It is not clear to what extent Sandia based their cost estimate on building realistic cost estimates. Romm claims, "[t]he technology has no obvious bottlenecks and uses mostly commodity materials -- steel, concrete and glass." but that is exactly the rub, the cost of steel, concrete and probably glass have doubled in the last 5 years and can be expected to inflate rapidly. Research on CSP technology and construction of CSP facilities are regionally concentrated in the American Southwest. In fact the desert climate of the Southwest means that useful sunshine will be available for an average of 8 hours a day or even more. In other parts of the country the availability of sunshine drops. Two thirds of the United States can expect an average of 5.5 hours of sunshine a day or less. Even worse cloud cover is not uniformly distributed in time, so in many locations clouds may obstruct the sun for several days at a time.
Thus CSP, even if it turns out to be a cost effective electrical generating source for the Southwest, still might turnout to be not cost effective for the rest of the country. Romm in the best tradition of renewable energy advocates, completely ignored the regional nature of CSP. Romm tells us, Finally, we will need more electric transmission in this country. "The good news is that because it [CPS] matches the load most of the day and has cheap storage, CSP can share power lines with wind farms. When the country gets serious about global warming, we will need to get serious about a building a transmission system for a low-carbon economy. "
In short good old Joe wants to build a lot of extremely expensive and highly vulnerable to terrorist attack electrical transmission lines to export CSP generated power from the Southwest to the rest of the country. How much would Romm's national CSP electrical generation system cost? No one knows, but it is a good assumption that building a thousand or so AP-1000 would be cheaper, and building several thousand LFTR's in a factory and using innovative siting approaches would be cheaper still.
Posts
6 comments:
Charles,
you left out that both solar and wind-farms are highly diluted power sources. This means the energy harvested per sqft is very small and has to be concentrated to high power grid connection points. This requires power lines and access-roads to each and every small generator part. Being it a windmill, solar panel or, to a smaller extent, mirror for ST.
The cost in materials, fossil energy and manpower to build and maintain such a grid is often discounted, but can rival in scope and size that of the distribution grid of an industrialized country.
Good comment, I have dealt with this issue in the past, and written about resource, energy and land use issues. My real interest her was the cost comparison between CSP and nuclear power.
CSP. Seems to be the going thing these days. EVERY CSP plant built in the world today or soon to be built is coming in at least $10,000 KW installed.
I do not know if that includes storage.
If it does, it is still very expensive. Has any one build a hot-salt storage system of, say, 300 MWS? What does it cost? If it's using molten-salt, this means it's using a steam turbine to return power to the grid. If it uses steam turbines you will need either vast quanties of water OR very large, power draining air-cooled condensers.
I just visited an air cooled condenser at a local PG&E plant. It's one of the biggest 'things' I've ever seen. Imangine 40 rotating fan blades, each on as big as the blades on a Blackhawk helicopter and you get the idea.
Also, the Enviros are split on CSP. Largly because the people pushing CSP are...the large fossil and nuclear generators, that is, big utilities. Oh well.
David
The story is that the Messiah is suppose to come and bring with him cheap CSP. The Messiah was suppose to come before, and did not show up. So maybe cheap CSP can't be counted on.
Charles,
I recommend you consider relaxing a bit and focusing on one issue at a time. You have a lot of experience, but you try to bring it all forth at once! You have a problem with calculations of EROIE, so explain it clearly and move on. Don't attack credentials of others as an argument in your favor, but rather try to cite valid sources of information. We all want to make the world a better place, but we can't do this with divisive rhetoric. I'm trying hard to educate the concerned public, and I am influencing the thinking in environmental groups. I even used your information about the India thorium strategic plan in my Dartmouth ILEAD course. Please leaf through
http://rethinkingnuclearpower.googlepages.com.
Sorry to be critical, but I'm just as old and crachity as you are :).
Robert no offense taken. The Oil Drum posts were written in Oil Drum debating style. Which means they were not very litterate. I mainly wrote the posts for the benefit of Charles Hall, whop was clearly over his head with the EROEI essays in the Oil Drum. The idea of nuclear EROEI is silly, so I wanted to put it to rest.
Post a Comment