Quora: Could Solar Panels Pay for Trump's Wall?

Mexico border wall
A general view shows a newly built section of the U.S.-Mexico border fence at Sunland Park, opposite the Mexican border city of Ciudad Juarez, Mexico, on January 26. Jose Luis Gonzalez/Reuters

Quora Questions are part of a partnership between Newsweek and Quora, through which we'll be posting relevant and interesting answers from Quora contributors throughout the week. Read more about the partnership here.

Answer from Josh Velson, part data scientist, part chemical engineer:

Would President Trump's proposed US-Mexico border wall be able to pay for itself if it were covered with solar panels? It would never pay for it under any rational system of accounting. The resulting solar installation would, at minimum, cost about four times as much as the wall itself and have orders of magnitude higher land use impact than the wall could reasonably cover.

Sure, you might pay off the cost of the solar PV systems involved - might, because building a linear system obviates many of the advantages of scale that come from a large solar farm. However, if you use proper accounting methods that include the opportunity cost of building such a wall as is standard with any form of project financial accounting, you will find that you incur a strong negative return in multiple different ways.

The Federal Government doesn't generally seek a large financial return on investment, so its hurdle rate for project finance is likely to be the return that it incurs from instead using the funds to pay down the national debt, or equivalently choosing not to issue the bonds from financing this through a deficit. For a project that nonpartisan sources have estimated will cost on the order of $10 billion (the true number is higher, but let's keep going), this cost will amount to $300 million in interest per year with twenty year treasury rates at 3% (they are currently at about 2.75% as of this writing).

You also have to factor in depreciation of the assets and sustenance capital. This wall isn't going to maintain itself. For the sake of argument, let's assume that depreciation (which accounts for the drop in value of an asset as it is being used, but isn't an actual cash flow) doesn't matter. You still have to put enough money into the project so that it retains its original use, otherwise known as sustenance capital. In high-level cost engineering, this is often taken to be a percentage of the original capital cost. To be generous, let's assume 5%. That's another $500 million.

You also have to consider operational expenses. A security wall is not a passive investment. I have no way of estimating operational expenditures and the like, but credible estimates put costs of operating at about $750 million per year. Subtracting our sustenance capital, that's $250 million per year.

So you'd need to generate at least $1.05 billion in revenue per year to even pay for continued operation and break even on interest paid, before you even pay down the principal. We then have to set a timeline for payoff. Let's say twenty years, which last I heard is the generally accepted lifetime of a solar PV system before replacement panels are required. The annual amount of capital to be paid down is then $500 million.

A generous price to get for power in the United States is ten cents per kWh (the average retail price in the US is around seven to nine cents per kWh depending on location, and the wholesale price given to the power producer is lower). You would need to generate 15500 GWh of electricity per year, which is approximately three times the output of the entire United States' solar installations during the month of July 2016, or about 1.5 times as much solar power as was generated in the entire United States in 2015 [1]. The US NREL [2] estimates that a fixed panel installation at the approximate latitude of the border wall in San Diego, CA will generate about 1607 kWh of electricity per kW of installed capacity if using commodity panels, which brings the nominal installed capacity requirement of such a facility to 19.3 GW [3]. NREL has further estimated that in the first quarter of 2016, the cost of a utility scale solar system was about $1.42 per Watt of installed capacity [4]. The cost of the system is thus about $27.41 billion[5], or nearly double what Congressional Republicans estimate that the entire wall would take. That cost in turn must be taken into account using the same sustenance capital and opportunity cost metrics we've been using - if we didn't increase the solar area, it would lose about $641 million a year.

To save some boring iterative work, I decided to just calculate what amount of panels would get us to break-even in a spreadsheet.

Assuming the same capex payback rate, same cost of capital, and no sustenance capital requirements, the equilibrium figure for break-even solar installation comes down to a whopping 32.9 GW of installed capacity. Such a system would cost about $46.7 billion in addition to the OOM $10 billion for the wall.

Yowza! This is roughly the unit cost of 4 Gerald R. Ford class aircraft carriers, or about 4 times the annual cost to taxpayers of the entire F-35 program.

And let's not forget that a typical solar installation makes money instead of merely breaking even on costs. The more money you want to make, the larger it will get - and that's not even considering that we assumed a generous electricity price.

The land use impacts of such an installation would not be trivial. In one of my earlier references, NREL calculated that a fixed axis solar installation will have a generation-weighted land use impact of about 4.77 ac/(GWh/yr). Some quick calculation suggests that over the 1954 miles of the US-Mexico border, the resulting solar field would have to be 1065 feet across, or roughly a fifth of a mile [6]. A reasonable figure for the wall and associated exclusion zone might generously be maybe ten feet. So plainly speaking, it would already be impossible to simply cover the proposed wall from end to end and have it remotely close to break-even, unless this wall is literally a fifth of a mile thick, at which point we're basically talking some East Germany type thing (and since the $10 billion order of magnitude capex figure likely scales with construction materials and we estimated a length of about ten feet, the capex estimate for the wall component is off by about two orders of magnitude itself, i.e. $1 trillion in stead of $10 billion).

As a former renewables project cost analyst, I cannot support this proposal.


[1] Electricity Data Browser

[2] http://www.nrel.gov/docs/fy13ost...

[3] Computational Knowledge Engine

[4] http://www.nrel.gov/docs/fy16ost...

[5] Computational Knowledge Engine

[6] Computational Knowledge Engine

Would President Trump's proposed US-Mexico border wall be able to pay for itself if it were covered with solar panels? originally appeared on Quora - the place to gain and share knowledge, empowering people to learn from others and better understand the world. You can follow Quora on Twitter, Facebook, and Google+. More questions:

Quora: Could Solar Panels Pay for Trump's Wall? | Opinion