A different way to look at solar

With government help, solar power becomes a natural resource to be tapped.

The typical way we think about solar is in terms of solar power, that is, as another way to generate electricity. This electricity is intermittent—there isn’t any at night. The standard solution is then to store electricity (e.g., in batteries) for use at night. Storage is expensive and reduce an already-low EROI to an even lower value as shown in the figure below. EROI is return on investment measured in energy units.

Figure 1. EROI values for various energy sources

We should be able to obtain a crude relationship between EROI and investment cost. For example, this report provides investment costs (I) for nuclear power expressed in dollars per watt ranging from $2-12. One watt of capacity, run at a 90% utilization would produce about 7.9 kwh of energy per year. Over a plant lifetime of L years, I dollars of nuclear plant will produce 7.9 L kwh of energy. Given a price of P dollars per kwh, we can convert the investment cost I from dollar per watt capacity to kwh per watt capacity (I/P). EROI is then the output of 7.9 L divided by I/P or

1.      EROI = 7.9 L / (I/P)  = 7.9 ∙ P ∙ L / I

If we assume L = 45 years and I = $2, the EROI value of 75 given in Figure 1 for nuclear implies an electricity price of $0.42 per kwh in this analysis. With P = 0.42, EROI of 4, and L = 20 years we get an investment cost (I) of $16.6/watt for solar PV. The value for EROI was obtained from a 2013 paper which used information from 2005-8. Solar PV investment costs in 2010 were more than $5/watt and rapidly declining, suggesting that were much higher before 2010. Given much lower investment costs today, EROI for solar PV is likely to be correspondingly higher. As more operating experience has accumulated, the expected lifetime of solar PV has also lengthened, which is another reason to believe EROI for solar PV today is quite a bit higher than 4.

What hasn’t changed is the intermittent nature of solar PV. The effective use of solar PV for electric power will require “smoothing” of the highly variable electrical output from solar PV to fit electrical demand. This requires what we in the chemical process industry call a “wide spot in the line” also known as surge capacity. This is simply temporary storage of output of one process step until the next process step can take it. Here is it the mismatch between what the solar PV is producing and the demand for electricity, requiring temporary storage of excess electrical power, in batteries, pumped hydro, or some other option. Another possibility, might be to add electrolyzers to a solar array to convert electricity and water into hydrogen and oxygen:

2.      9 kg  H₂O + 57 kwh à  1 kg H₂ + 8 kg O₂

If this is done the variable electrical output is converted to a variable hydrogen output, which can be stored and released as needed like any other gaseous chemical product. The hydrogen produced by electrolysis would be compressed and fed into a pressurized dispatch vessel, making the solar installation sort of a “hydrogen well” that produces steady daily stream of hydrogen whose rate varies with the seasons. This hydrogen could be used to produce “green” synfuels for use in applications like aviation, for which there are no green options.

Using this idea, I constructed a very simple cost model to compare nuclear-derived hydrogen produced where it is to be used with solar-derived hydrogen from such a “well” transported to the point of use. Solar PC electricity is much cheaper than nuclear. On the other hand, solar energy is diffuse, requiring 4-7 acres per MW electric power (enough to produce 150 tons/yr of hydrogen) and so generally cannot be built on the user’s site. This is the reason for my use of the hydrogen well analogy. Like oil and gas, one must collect it where it is and then transport it to where it will be used. The solar wells would be located in a region with lots of sun and low-value land, like the Chihuahuan Desert in the American Southwest. The hydrogen would be conducted by pipeline to manufacturing plants where it would be used to produce fuels and chemicals.

American demand for aviation fuel in 2019 was 18.3 billion gallons (6 billion kg). Just based on the stoichiometry, it would take at least 2.6 billion kg/yr of hydrogen to produce this much fuel. A minimum of 18 GW of electrical capacity would be needed which translates to some 150 square miles of solar farms in an optimal siting. Alternately, 4-5 large nuclear plants could manage the whole thing. The example I had the mind was wells located in Southwestern Texas, around 500 miles from end users in the Houston area petrochemical industrial complex, or perhaps in much closer Midland TX. I compared the cost of hydrogen produced from solar PV as a function of distance from use point with hydrogen produced from nuclear power constructed at the point of use with the results shown in Figure 2 below.

Figure 2. Cost of solar-generated hydrogen as a function of distance from use point compared to nuclear.

Nuclear power-derived hydrogen has a highly variable investment cost depending on where it is installed. In authoritarian societies like China nuclear investment costs are as low as $2 per watt for which my simple model gives a hydrogen cost of $1.87/kg, lower than lowest value for solar, but PV costs for them are probably also lower, so PV can still make sense for them. In any case they are building a lot of both. Nuclear costs are higher in liberal democracies, ranging from $2.5-4 in Japan and Korea to $8-12 in the West. 

My PV analysis was based on hydrogen wells each employing a 250 MW solar array (installations this size already exist). Each would be capable of producing enough hydrogen to serve 1.4% of the aviation fuel market. Now suppose the government-built projects like this as a form of stimulus for dealing with economic downturns rather than counterproductive policies like QE. Money would be created and used to build a number of hydrogen wells in the deserts of the Southwest. Once created the government would lease the wells out to businesses seeking to exploit this resource. The cost of the hydrogen to the end user would come from pipeline costs and the lease rate, which would be set by the government. Figure 2 shows the cost (to government) of the hydrogen from one of these wells would be about $2.15/kg, implying the lease would need to be at least $322,500/year for the government to break even.

The primary application for this hydrogen is making synfuels. A recent analysis suggests that hydrogen costing $2/kg would give synfuel costing about $5.4-5.9 per gal, which is far more than the current price of around $3. Suppose the government offered initial lease rates at a discount to encourage the development of a green aviation fuel industry. If the government initially sold the gas for nothing the cost of hydrogen to the synfuel maker would just be the transport cost, which the model suggests could be $1/kg or less for shorter distances. According to the analysis cited earlier a hydrogen cost of $0.8/kg would be needed to produce fuel competitive with today’s price. Free hydrogen produced by government-owned hydrogen wells could be used by synfuel companies to produce aviation fuel cost-competitive with the fossil-derived variety. After an initial free lease, the government would then open future leases to bidders, in an attempt to establish a market price for solar hydrogen “in the field.”

Wells could be constructed on worthless land in other sunny parts of the country that are reasonably close to industrial centers. Once a solar-generated synfuel industry has been established, there will be considerable support for solar energy in Texas and the American Southwest, which should reduce opposition to the “killer ap” that would get this new industry really moving—a carbon tax. A carbon tax raises the price of non-green fuels relative to green ones. Once the existing petrochemical industry has developed green synfuels made using low-cost “government hydrogen” that can serve as drop-in replacements for their fossil fuel-derived nongreen products, the government can start taxing carbon, making synfuels preferable to the fossil variety. A strong demand for hydrogen will result and lease rates on government wells will rise. At some point lease rates will justify the cost of creating these wells, private industries will then start building their own wells, and the government can sell the ones they built to recover their investment cost and a reasonable profit.

At this point, the market price for hydrogen will be high enough for nuclear hydrogen to make sense for portions of the country that lack optimal solar resources but have the infrastructure and know-how to make synfuels. The government, by building these wells, will have enabled the birth of one aspect of a new green energy leading sector. This idea was inspired by Henry Clay’s American plan that called for government to construct infrastructure that would aid in economic development.

As for the cost of the program, there isn’t any. Building these wells would be a stimulus project. America is in a secular cycle crisis. Historically there are only two paths out of these times. One is through political crisis such as civil war, revolution or state collapse, and the other is through successful resolution of economic crisis. America has seen both kinds; the 18^th and 19^th century crises were resolved through revolution and civil war, respectively, and the 20^th century crisis was resolved through the Great Depression and World War II. If we choose the political path, the result will be widespread death and destruction. Even when handled in a more orderly fashion as in our own Revolution and Civil Wars, the cost in lives (750K dead in the Civil War) and treasure (ca. $7-10 trillion) was enormous. Building enough hydrogen wells to produce free hydrogen to make all the aviation fuel we need would be around 1% of the money spent on quantitative easing (QE) over 2008-2022.

Readers might note that QE did not involve government spending per se, just money creation, but government spending (fiscal stimulus) is also money creation. It works a bit differently than QE. QE is spent directly into the financial economy, where it mostly stays, producing inflation in asset prices, not goods and services. One of these assets, housing, is also a consumer good, whose rising price may be preventing family formation, which is not a good thing. Fiscal stimulus for infrastructure like hydrogen wells is spent directly into the economy, where it can temporarily bid up prices of infrastructure inputs. However, this new money is extracted from the economy as profits, which are sent to financial markets via dividends and stock buybacks. The money ends up in the same place, it simply takes a more circuitous route. Fiscal deficits generally do not produce inflation unless they are large relative to dividends + stock buybacks. As I see it, it makes more sense to have money destined to end up in financial markets employed for something useful on its way than to send it there directly via QE.

What this means is a program of hydrogen well building is something that could be pursued by a Democratic administration facing an economic downturn. Wells built in Southeastern New Mexico and Southwestern Arizona to feed industry in Texas and California would produce direct stimulus in Blue New Mexico and Purple Arizona, serving as pork for Democratic legislators as did New Deal projects in the South. Creation of a new green synfuel industry in Texas, could convince more Texans that environmentally Green can translate into Greenbacks. A long-term goal of the Democratic party has been to turn Texas blue, perhaps introduction of a money-making Blue policy can do the trick. As for Republicans, they oppose doing anything about global warming. Much of the focus of climate denialists (whose analysis forms the basis for the Republican climate policy stance) has been on the lack of a clear relation between adverse weather events and global warming, which ignores the real threat from climate change.

Rising temperatures mean some regions of the world will become uninhabitable, resulting in large numbers of climate refugees. Among them will rise large numbers of anti-Western political actors who have nothing to lose (think Hamas times a hundred). Sooner or later, they will get their hands on WMDs or seize control of a nuclear-armed state like Pakistan, which is one of the countries likely to be seriously affected by climate change. How long before agents bent on suicide terror attacks against the West get ahold of nukes? Already today we have seen seemingly irrational behavior from nuclear-armed actors. Russia has embarked on a war of conquest at a cost of more than 300 thousand casualties to expand their territory by 0.35%. Given their collapsing demographics Russia simply cannot afford to spend its young men so wastefully for so little, and yet that is what it has done. Finally, the US is presently going through a secular cycle crisis, in which societal fission and internecine war becomes more likely with each year. Rising levels of greenhouse gases promise to return humans to an era of climatic variability in which environmental forces combine with the normal social forces to push harder for societal collapse and widespread conflict. These are the reasons for concern about climate change, not the direct effects on of rising temperatures on US weather.