Food Security and Solar Land Requirements
As solar farms gobble up land, what happens to food production and ecosystems?
As solar farms gobble up land, what happens to food production and ecosystems?
Last week I was reading an article on Oilprice.com by Irina Slav regarding the planned expansion of solar capacity in the US. It appears a report by the Solar Energy Industries Association is expecting a big increase in solar capacity. Here is an except from the article:
According to the SEIA report, the country’s solar market will expand by 40 percent over baseline projections by 2027 thanks to the stipulations in the IRA (Inflation Reduction Act). This is equal to 62 GW of capacity. Thanks to this additional stimulus, utility-scale solar in the U.S. is seen adding some 162 GW over the next five years, with total solar reaching 336 GW by 2027.
Ok. So utility-scale solar capacity will expand by 162 GW over the next 5 years.
That caused me to ask myself, “How many acres of land will that require? And where will that land come from?” Reasonable questions, I believe.
So I decided to do some quick back-of-the-envelope math to get a rough idea of what we would be looking at. The results would be as good as the information I could glean from the web. Here is what I found on the internet:
Five largest solar farms in Texas
Roadrunner Solar Farm, capacity 497 MW on 2,770 acres. 5.57 acres per MW of capacity.
Uptown 2 Solar Farm, capacity 213 MW on 1900 acres. 8.9 acres per MW of capacity.
Roserock Solar Farm, capacity 212 MW on 1300 acres. 6.13 acres per MW of capacity.
Buckthorn Solar Farm, capacity 202 MW on 1250 acres. 6.19 acres per MW of capacity.
Holstein Solar Farm, capacity 200 MW on 1300 acres. 6.5 acres per MW of capacity.
These calculations may not be perfectly accurate, but all 5 solar farms take somewhere between 5 1/2 and 8.9 acres per MW of capacity.
The average of the 5 solar farms was 6.6 acres per MW of capacity.
Acres required to add 162 GW of Solar Capacity in the US
Since 1 GW equals 1,000 MW, to meet this goal would require adding 162,000 MW of utility-scale solar farms. If these new solar farms are similar in acreage requirements to the Texas solar farms, then it will take 1.08 MILLION acres of land to build them. Even if you use Roadrunner Solar Farm’s acreage usage of only 5.57 acres per MW of capacity, you would still need 902,340 acres, which is pretty close to a million.
Where can we find a million acres for new solar farms?
One option is to do as this article written in April proudly proclaims: Texas rice fields in Texas are converting to solar PV farms. Yes, we could convert more rice farmland on the Coastal Bend of Texas into solar farms. Cropland works well because it is relatively flat and doesn’t require the bulldozing of trees and such. The only drawback to taking farmland out of production is that less food will be produced, reducing supply and increasing price pressures. In situations like we see in the world today where there is food inflation and predictions of food shortages around the world, one would think that we should be increasing food production as much as possible. This is not a good option.
A second option is deforestation. Many large cities in our country are located in generally wooded areas where large acreage tracts suitable for a 1,200 acre solar farm are currently covered in vegitation. These plants and trees would need to be removed, along with any wildlife that lived there. This is not a very good option either.
A third option would be desert. I can hear a chorus of people saying, “Yeah! That’s the ticket!” But there are a few problems with this option as well. First of all, it is difficult to transmit electricity over distances greater than 500 miles, assuming you can afford the metals to build 500 miles of transmission lines. Tell me. Is there a desert within 500 miles of where you are reading this right now? The problem is most population centers using the electricity are not near a desert. Also, people forget that deserts are ecosystems, too.
Tortoises Pay the Price for New Solar Farm
Last year the Las Vegas Review-Journal had an article titled Desert Tortoise Deaths Raise Concerns as Solar Farms Solve Energy Need.
Obviously, the need for the solar farm outweighed the tortoises’ need to stay in their habitat.
A team of biologists relocated 139 tortoises from their habitat to make way for the solar panels in the Yellow Pine Solar Project, one of four large solar energy developments initiated in Southern Nevada. The tortoises were moved across the road to Stump Springs in May. In a span of a few weeks, 30 tortoises were killed, possibly by badgers. Conservationists believe relocation stress made the reptiles more vulnerable and drought caused badgers to look for new sources of prey. Wildlife experts are still looking into the exact cause.
Unfortunately, these tortoises are members of a threatened species.
The desert tortoise has been listed as a threatened species under the U.S. Endangered Species Act because of population decline due to predation, collection by humans, off-highway vehicles, and upper respiratory tract disease.
Deserts are not a good option either.
One last option I can think of is putting the solar panels on top of warehouses and other large commercial buildings, none of which are 1,200 acres. But if you put solar panels on enough commercial buildings it could be possible, maybe to add up to 162 GW. Amazon thought that was a good idea, but has had a lot of problems as you can see outlined in this CNBC article: Amazon took all U.S. solar rooftops offline last year after a flurry of fires, electrical explosions. Here is the summary of the article:
•Between April 2020 and June 2021, solar panels atop Amazon fulfillment centers caught fire or experienced electrical explosions at least six different times.
•“The rate of dangerous incidents is unacceptable, and above industry averages,” an Amazon employee wrote in an internal report viewed by CNBC.
•Rooftop solar is part of Amazon’s broad plan to zero out emissions by 2040.
Evidently, until some more bugs are worked out of these systems, it would be unwise to go with this option either.
Conclusion
We will have to give up a million acres of land in order for the five year solar power goals of the US to be met. And nobody believes it will stop there. More and more land will be gobbled up over the next few decades if we don’t slow down and think things through.
________________
Add to the problem of desert or remote solar is the Capacity Factor is only ~25% and the transmission lines must be sized for peak output so 4X oversized compared to a conventional power source. And the power is in oversupply during the solar peak on top of only 25% average loading, so adding more aluminum to the transmission lines to reduce line loss has a low return on investment. ROI determines how much aluminum you put into your transmission lines. So long distance solar transmission has a very poor ROI which means minimum aluminum = max line loss. And that gets worse, much worse with increased solar grid penetration. Expect line loss for transmission of solar of over 25%. Add to oversupply losses, negative pricing, curtailments, seasonal demand mismatch and you are going to be wasting probably 50% of your desert solar farm output.
And there are more grid inefficiencies on top of that. Like having coal boilers fired up building steam waiting for the solar to dwindle in the late afternoon. Germany figured out a nice scam, they disconnect the coal generators from the grid when the coal power plants are idling, so they don't have to declare the emissions they produce. "Let's just ignore all that CO2 produced". So just like VW Germany cooks their electricity sector emissions numbers. How come they don't get big fines for that?
Great piece! Doubtful any of the dreamers are bothering to do even this back of the napkin type math.
Doomberg stated in this post last year that it takes 5,000 tons of polysilicon to manufacture 1,000 MW of solar panels.
https://doomberg.substack.com/p/herbie-spoils-the-party
So doing some of the same back of the napkin math: 162GW over 5 years is 32.4 GW/yr -> 162,000 tons of polysilicon per year.
Doomberg also mentions three polysilicon plants in the US: REC, Hemlock, and Wacker.
REC closed in 2019 but is apparently going to reopen at a hefty cost to taxpayers thanks to the IRA (https://kpq.com/moses-lake-plant-to-benefit-from-inflation-reduction-act/) and supposedly have a production capacity of 22k tons per year. Hemlock and Wacker are 36k and 20k respectively. Total that's 78k tons of processing capacity in the US.
Our current capacity, would only be able to produce about 48% of what's needed each year assuming the American production went only to industrial solar and not to other sources such as rooftop.
Where do we get the rest?
These plants take years, if not a decade, to build from scratch.
Polysilicon is a commodity that raises up and down in price too.