Because that square meter represents capital infrastructure which must be built, maintained, integrated into the grid (inverters, transmission lines), and eventually replaced and updated.
Again, don't get me wrong: I think there's absolutely no question that our future will rely primarily on renewable and sustainable energy sources. The question to me isn't whether this will be the case, but if it will be a high-tech renewable energy infrastructure (with electricity, grids, etc.) or low-tech: farms, grains, and draught animals.
The costs of renewable energy strike me as vastly higher than most of the general public seems to think. You'll get fossil-fuel fans arguing this as a reason that renewables aren't viable. I think that's a misplaced argument of a legitimate concern: it's whether or not a technological civilization is viable.
I don't know one way or the other. I think we're skating really, really close to the edge.
Erik Lindberg's recent Resiliance / Transition Milwaukee piece, "Six Myths About Climate Change that Liberals Rarely Question", lays out most of these factors:
If I'm reading EIA's capacity data correctly, there are about 313 GW of installed coal generating capacity in the US. Let's double that on a basis of capacity factor, plant maintenance, and other factors.
GRIST, an environmental site, allows for one square mile (640 acres) for a nominal 1 GW power plant. So we're talking about a maximum of around 600 square miles dedicated to coal plant capital itself. That's a region 24.5 miles on a side.
And note that I'm not talking about total land-use footprint: coal storage, transport, mining, and disposal all have footprints, and they're substantial. But most of them are also pretty non-technical -- a mine, after all, is for the most part a hole in the ground, not a machine.
But of that square mile, the bulk of it isn't actually technical infrastructure. An Indian engineering report puts the main plant at 4.6% of total land area for a 5 x 800 MW plant, 92 acres. That's 43.4 MW/acre.
Which means that of the 600 square miles devoted to coal power plant siting, only 11.25 mi^2 of that is actually coal plant itself -- the rest is (mostly) much less technical land use.
By comparison, with a nominal 1 kW/m^2 of incident sunlight, you're starting with 4 MW/acre available energy. By the time you apply PV efficiency (15%), spacing factor (55%), capacity factor (30%), and inverter efficiency, you're down to 0.09 MW/acre delivered energy. You'll need 477 acres of solar infrastructure to provide the same output as a single acre of coal plant.
GNU units is handy for calculating various factors, and we can compute how much area needs to be dedicated to an equivalent amount of solar generating capacity:
That is: we'd need 5,425 square miles, or a region 73 miles on a side.
(Note I've omitted considerations such as storage or other factors -- this analysis is favorable to solar power.)
There are a lot of other factors you can consider. Coal plants generally don't do well sited immediately adjacent to other land uses: agriculture, residential, commercial, or even industrial. You can park solar panels on top of pretty much anything -- they're nonrivalrous for much human activity (though they compete with ag for access to sunlight). That's not what I'm considering.
But even with that, you've got to provide:
⚫ Mounting / foundation structures.
⚫ Panels themselves.
⚫ Rectifiers.
⚫ Grid interconnects.
All of those represent installation costs.
Once installed, there is some maintenance required as well. I don't have good cost data on this, or labor requirements. But if nothing else, someone's got to move through the installation, and area means distance means time.
If you want to go further with solar power and allow for excess capacity to provide for storage, fuel synthesis (I suspect that's going to be part of the equation), and substitution for other energy uses (e.g., not just present electrical consumption), then you're talking land-use that's hundreds of miles on a side for the US alone. This would be a substantial portion of a large state (say: Nevada, Arizona, New Mexico, Utah, etc.). Yes, the land use can be broken up and distributed across other states, but you simply cannot get away from the fact that it's a lot of area, and that you're parking a highly technical infrastructure on it.
And if you think this is a pessimistic analysis, try calculating land-use requirements for biofuels given per-acre productions of 30-300 gallons/acre-year for typical oil crops (corn, canola, hemp), and 6.75 billion barrels of present US annual oil consumption. There are about 409 million acres of arable land in the U.S., and 2,379 million acres total. Try it yourself with GNU units or Wolfram+Alpha.
Hmmm, I still don't think that is a valid metric to measure both by. There are too many variables and analogies to make it a good comparison. There is a huuuuugggeee variable you didn't even talk about which is the electric grid. Huge amounts of land is taken up for this and it is very expensive to maintain. One of the reasons I want solar is because where I live (midwest US) falling trees/branches during wind storms and snow storms causes lots of down time.
I could go on about other things, but I still think comparing the land use of both of them is fairly ridiculous.
You've got long-distance transmission. The issue here isn't the power losses (actually fairly low and constant at about 6%), but the per-mile costs: $1m - $2m. Even on linear distance, that's not too bad, but if you're creating a situation where you've got to have high-capacity transmission all over the place (areal, not lineal density), it starts adding up. That's one of the costs the Solar Fucking Idiot Roadways people failed to take into account.
There are interconnects -- basically inverters and busses which feed back into the grid. I don't have a lot of details on this, but you're basically allocating some of this per specific facility of installed solar.
And then there's grid management. The issue with solar is that it's variable. That doesn't mean "unpredictable", as you actually can predict with very high accuracy 24-72 hours in advance (long enough to take actions) what your supply and demand will be. I've been commenting recently on G+ about the German Fraunhoffer Institute's solar energy who discusses high and low cost points for per MWh electrical costs. There were several of these in 2014, each tied to specific forecasting misses. Generally, either renewable power sources were over or under estimated, or demand was.
Storage, demand-side management, and other options might impose various costs of their own though.
Again, don't get me wrong: I think there's absolutely no question that our future will rely primarily on renewable and sustainable energy sources. The question to me isn't whether this will be the case, but if it will be a high-tech renewable energy infrastructure (with electricity, grids, etc.) or low-tech: farms, grains, and draught animals.
The costs of renewable energy strike me as vastly higher than most of the general public seems to think. You'll get fossil-fuel fans arguing this as a reason that renewables aren't viable. I think that's a misplaced argument of a legitimate concern: it's whether or not a technological civilization is viable.
I don't know one way or the other. I think we're skating really, really close to the edge.
Erik Lindberg's recent Resiliance / Transition Milwaukee piece, "Six Myths About Climate Change that Liberals Rarely Question", lays out most of these factors:
http://www.resilience.org/stories/2014-11-26/six-myths-about...
Back to that output area.
If I'm reading EIA's capacity data correctly, there are about 313 GW of installed coal generating capacity in the US. Let's double that on a basis of capacity factor, plant maintenance, and other factors.
http://www.eia.gov/electricity/capacity/
(Stealth edit: and another source confirms 557 coal plants as of 2012. http://www.eia.gov/electricity/annual/html/epa_04_01.html )
GRIST, an environmental site, allows for one square mile (640 acres) for a nominal 1 GW power plant. So we're talking about a maximum of around 600 square miles dedicated to coal plant capital itself. That's a region 24.5 miles on a side.
http://grist.org/article/2010-11-17-which-has-bigger-footpri...
And note that I'm not talking about total land-use footprint: coal storage, transport, mining, and disposal all have footprints, and they're substantial. But most of them are also pretty non-technical -- a mine, after all, is for the most part a hole in the ground, not a machine.
But of that square mile, the bulk of it isn't actually technical infrastructure. An Indian engineering report puts the main plant at 4.6% of total land area for a 5 x 800 MW plant, 92 acres. That's 43.4 MW/acre.
Which means that of the 600 square miles devoted to coal power plant siting, only 11.25 mi^2 of that is actually coal plant itself -- the rest is (mostly) much less technical land use.
http://www.cea.nic.in/reports/land_review_report.pdf
By comparison, with a nominal 1 kW/m^2 of incident sunlight, you're starting with 4 MW/acre available energy. By the time you apply PV efficiency (15%), spacing factor (55%), capacity factor (30%), and inverter efficiency, you're down to 0.09 MW/acre delivered energy. You'll need 477 acres of solar infrastructure to provide the same output as a single acre of coal plant.
GNU units is handy for calculating various factors, and we can compute how much area needs to be dedicated to an equivalent amount of solar generating capacity:
That is: we'd need 5,425 square miles, or a region 73 miles on a side.(Note I've omitted considerations such as storage or other factors -- this analysis is favorable to solar power.)
There are a lot of other factors you can consider. Coal plants generally don't do well sited immediately adjacent to other land uses: agriculture, residential, commercial, or even industrial. You can park solar panels on top of pretty much anything -- they're nonrivalrous for much human activity (though they compete with ag for access to sunlight). That's not what I'm considering.
But even with that, you've got to provide:
⚫ Mounting / foundation structures.
⚫ Panels themselves.
⚫ Rectifiers.
⚫ Grid interconnects.
All of those represent installation costs.
Once installed, there is some maintenance required as well. I don't have good cost data on this, or labor requirements. But if nothing else, someone's got to move through the installation, and area means distance means time.
If you want to go further with solar power and allow for excess capacity to provide for storage, fuel synthesis (I suspect that's going to be part of the equation), and substitution for other energy uses (e.g., not just present electrical consumption), then you're talking land-use that's hundreds of miles on a side for the US alone. This would be a substantial portion of a large state (say: Nevada, Arizona, New Mexico, Utah, etc.). Yes, the land use can be broken up and distributed across other states, but you simply cannot get away from the fact that it's a lot of area, and that you're parking a highly technical infrastructure on it.
And if you think this is a pessimistic analysis, try calculating land-use requirements for biofuels given per-acre productions of 30-300 gallons/acre-year for typical oil crops (corn, canola, hemp), and 6.75 billion barrels of present US annual oil consumption. There are about 409 million acres of arable land in the U.S., and 2,379 million acres total. Try it yourself with GNU units or Wolfram+Alpha.