Clearly, this is nonsense. A complete waste of time. It’s not our intention to suggest the US will pay for its Mexico border wall by putting solar panels along the south side. That’s definitely probably not going to happen. Though comic book supervillains might draft secret plans to harvest daylight from their enemies, the evidence that POTUS is a comic book supervillian is at best circumstantial. What we have here is a hypothetical.Would it work, though?Stage I: building a Wall“What a great idea!” says Doug McKenzie, a solar analyst and consultant, who goes on to explain why it’s a terrible idea. (Doug blogs at lightsonsolar.com: the principles and calculations that follow are mostly referencing off his work. Thanks also to Energy Futurist / Analyst / Writer / Speaker Chris Nelder for
Bryce Elder considers the following as important: Uncategorised
This could be interesting, too:
FT Alphaville writes Snap AV: Your Trump Trade reversals, charted
Siona Jenkins writes FT Opening Quote: Whitbread brews up sales but misses profits
David Keohane writes Further reading
Guest writer writes Guest post: Why shrinking the Fed balance sheet may have an easing effect
Clearly, this is nonsense. A complete waste of time. It’s not our intention to suggest the US will pay for its Mexico border wall by putting solar panels along the south side. That’s definitely probably not going to happen. Though comic book supervillains might draft secret plans to harvest daylight from their enemies, the evidence that POTUS is a comic book supervillian is at best circumstantial. What we have here is a hypothetical.
Would it work, though?
Stage I: building a Wall
“What a great idea!” says Doug McKenzie, a solar analyst and consultant, who goes on to explain why it’s a terrible idea. (Doug blogs at lightsonsolar.com: the principles and calculations that follow are mostly referencing off his work. Thanks also to Energy Futurist / Analyst / Writer / Speaker Chris Nelder for trying to guide us through the basics.)
The first thing to note is that the US-Mexico international border is quite long. As in, it runs for 3,200 kilometres. America has a treaty with Mexico to protect the environment for 100km on either side of the border, which might limit the ability to build straight lines across the length of the frontier. And the borderline snakes quite a bit, across all sorts of terrain, in a way that solar farms usually don’t. Such complexity demands ballpark estimates where small changes make big differences. Links are peppered through the text for you to run your own numbers, or to backtest ours.
We’re assuming the Wall needs to be wall-shaped. Whereas solar panels work best when pitched at an angle, there would be obvious downsides to building a ramp. Some solutions have already been proposed involving moats and suchlike, but evaluation of their practicality and cost-benefit is beyond the scope of this analysis.
Here’s the design concept we’ll be using:
How high, though?
The Great Wall of China tops out at six metres high. Trump’s Wall will be Greater, which means it’ll be seven metres high; QED. That gives 22 square kilometres of wallspace. Allowing for 2 square kilometres to be left fallow (maintenance access, border checkpoints etc) leaves a usable area equivalent to a microWales.
As well as the area we need an azimuth, the horizontal angle at which the sun’s rays hit the Wall.
Panel alignment is a subject not without controversy, as Dr Joshua Rhodes explains in Scientific American, because south-facing will maximise output while west is preferable for catching peak demand. Fortunately, the Mexico side of the US border faces both south and west in parts, so should offer a convenient balance between efficiency and timeliness.
Plotting a line though a US Geological Survey geospacial shapefile, as per below, gives a median azimuth of 210 degrees.
Photovoltaic solar panels, vertically fixed and benchmarked to Arizona, have an efficiency loss of about 50 per cent. Siting the panels at a 210 degree azimuth loses another 10 per cent. Such a rig setup is far from optimal, but neither is it totally impractical.
Stage II: pricing
Your average panel is two metres square and has a theoretical capacity of 300 watts. To cover a microWales, we’ll need 10m of them.
At pixel time the cheapest quote on a utility-grade fixed tilt array is $1 per watt — though it’s not entirely clear how much hitting that mark depended on support from the Department of Energy’s SunShot Initiative. Here’s where politics injects another uncertainty to the calculations.
Trump’s energy ….. uh, let’s call it a policy ….. favours carbon over silicon. The “America First Energy Plan” aims to deliver energy independence by burning through domestic reserves of shale oil, coal and natural gas. Renewables don’t get a mention. Green-industry types have therefore braced for lean times in anticipation of Congress adding import tariffs and unwinding incentive schemes, potentially including Sunshot and the federal Investment Tax Credit for solar power.
No matter. Solar price per watt has been collapsing for two decades and even an order for 10m panels might not do much to change that trajectory. Utility-scale arrays halved in cost between 2010 and 2015, a Berkeley Lab study showed, while EnergyTrend has estimated that industry supply this year is exceeding demand by between 18 per cent and 35 per cent. Then there’s Swanson’s Law, the observation that the price of a solar module drops by a fifth for every doubling of cumulative shipped volume.
Given all that, $1 per watt seems a reasonable target. It also keeps the maths pretty. So: an order of 10m panels rated 300 watts each gives a 3 megawatt rig at an entry-level cost of $3bn.
Stage III: powering up
Solar panels make imperfect barriers. Any panel spray-painted “vete al carajo El Naranja” will exhibit a degradation in cell efficiency. Likewise modules that have been hit by bricks, or torn off the Wall to be sold for scrap. Plenty of companies are working to ruggedise panels, however, and the performance loss for high security isn’t overly problematic. For example, Solar Roadways does a tempered glass panel that can hold the weight of a semi-trailer truck at a performance handicap of around 10 per cent.
A generic solar panel operates at about 20 per cent efficiency. Vertical alignment halves that, while the 210 degree azimuth and the tempered high-security glass will cost another percentage point each. That gives us at 8 per cent efficiency. Even at that paltry yield, our 3 megawatt Wall will push out 3.6m kilowatt-hours a year.
But wait. Conversion into grid-compatible alternating current will lose some of that power, albeit not much more than 3 per cent, while there will be some drag from glass cleaning, shading from propped ladders, etc. Taking all that into account, it’s reasonable to believe each kilometre of solarised Wall would deliver in the region of 1 kilowatt-hour.
The US consumes about 4 trillion kilowatt-hours of electricity a year, EIA data show. A coast-to-coast solar Wall would generate about 0.1 per cent of this requirement. That’s enough juice for the US to shut down quite a few of its coal, gas, nuclear, and petrol-burning power plants and rely a bit more on solar, wind and hydro. Big win for the environmentalists!
If the US were somehow able to scale up the project to strand its entire carbon power industry, greenhouse gas emissions would be reduced by about a third. Worldwide that’s a 5 per cent overall reduction, which might or might not be sufficient for the US to unilaterally keep at bay the global warming Trump may or may not believe in.
Stage IV: domestic sourcing
Problem. Just six companies control nearly half the market for silicon modules: one is Canadian, four are Chinese, one is South Korean, and none is American. A locally sourced Wall would need to rely on second-tier operators such as California-based SunPower, which seems unprepared for an order of 10m panels having last year laid off a quarter of its workforce.
There’s also SolarCity.
But let’s be clear. It’s a coincidence that Tesla shareholders inked its merger with SolarCity eight days after Trump’s election victory gave the green light to the world’s oddest grand design. There’s nothing suspicious in Elon Musk’s unprompted and ideologically dissonant decision to join the White House economic advisory council less than a month later. And there’s no reason whatsoever to question the commercial motives behind SolarCity’s plans to build the world’s largest panel factory in upstate New York, thereby adding a US-branded gigawatt of capacity into a deflationary market that’s already awash with Chinese silicon. This is a serious analysis. It has no time for conspiracy theories. Take it to Reddit, wingnut.
Stage IV: the transportation
Tesla aims to make 500,000 cars a year by 2018. Each Tesla has the capability to carry one panel on its roof rack, probably. The cars’ self-driving capabilities, in combination with Trumpian regulatory rollback, might allow the fleet to be deployed 24-7.
A round trip between the border and SolarCity’s factory in Boulder takes 28 hours. Assuming half an hour each side for loading and unloading, and allowing for zero downtime, all the panels could be delivered by one annual yield of Teslas in just 47 days.
Alternatively, there’s a train.
Stage V: the teardown
Inconveniently, less than 2 per cent of the US population lives within 40 miles of the Mexico border. Bringing power to the people would likely require construction of an interstate electricity transmission superhighway, which is something the renewables associations have long been lobbying for.
A 2012 study by Parsons Brinckerhoff for the Institution of Engineering Technology put the cost of overland long-stretch superhighway cables at £1.6m per kilometre overground, or £4m including lifetime operation and maintenance. (Burying the cable costs about five times as much.)
Using those figures, and at current exchange rates, a single overground power line from running from New Mexico to North Dakota requires budgeting a minimum of $8.5bn.
Another inconvenience is the sun’s tendency to set each day, particularly if it’s going to be providing a substantial proportion of America’s electricity requirement. Once again there are some novel solutions to the darkness problem — such as pumping water at high pressure into spent oil wells when electricity is plentiful, then firing it back at a turbine when it’s not. There would be a lot of work required to avoid nightly blackouts, however, given the US deployed just 161 gigawatt hours of energy storage in 2015. That’s only enough to capture the annual output of 20 wind turbines.
Also beyond the scope of this analysis are the economic and social consequences of stranding quite.a few of America’s nuclear and carbon-burning power stations. Suffice to say: probably not good?
Of course, it was a silly premise to make the design this grand. Any proposal to incorporate renewables into the Wall’s construction, if one is ever to exist, would need to be informed by realism and practicality. Support from the campaigners and special-interest groups couldn’t be bluffed with grandiose plans, because everyone knows there’s no earthly reason to build a whole Wall out of solar panels. It’s a non-starter. Our working through of the mechanics of such a scheme is the opposite of fake news: factual nonsense.
Judgements on whether the same phrase applies to areas of current US government policy are, likewise, beyond the scope of this analysis.
Note: this post has been substantally rejigged to correct some extremely iffy area maths.
The View from Mexico City — The World Post
Some notes on the Worst Case Scenario — Charlie Stross
Recreating the Scrooge McDuck Gold Coin Swim — Billfold