BNC's solar slam was misleading

How long before it's free?

Back in May, Brave New Climate published a thought experiment related to the potential of solar power in Florida. It was a guest post by Peter Morcombe, a climate "skeptic" who comments as "gallopingcamel." It came to my attention recently when Judith Curry highlighted it.

Imagine a future in which so-called “environmentalist” politicians are given the mandate to prohibit the construction of nuclear and fossil fuel power plants in Florida. As wind and hydro are not suited to Florida, the only remaining option would be solar power.

Mr. Morcombe is proposing a thought experiment to evaluate the potential of renewable energy in Florida, and, by extension, the world. So the critical element of the scenario is: no new fossil fuels or nuclear. He then asks us to look at what that would mean in 2100. But there is a problem with Mr. Morcombe's first assertion; that wind and hydro are unsuited to Florida. In fact:

A recent report by Oceana, an ocean conservation organization, estimates wind energy potential on Florida's offshore Atlantic coast as 10.3 gigawatts, enough to supply 16% of electrical generation. Consistent with this report is a study by Navigant Consulting, sponsored by the Florida Public Service Commission (PSC). The study suggests that although onshore winds tend to be weaker in Florida due to lack of terrain and open prairie spaces, Florida is surrounded by a robust offshore wind resource. Offshore wind is desirable since wind farms may be positioned in relatively shallow water offshore adjacent to coastal electrical load demand centers of major metropolitan areas, but far enough to be unseen from the coast.

Offshore wind is expensive right now, but then, so is solar. Mr. Morcombe is wrong to ignore this potential energy source in his ninety-years-from-now thought experiment. Indeed, accurately predicting power generation and consumption ninety years into the future can be expected to be difficult. Ninety years ago nuclear power was undreamed (1) of, wind and solar had never been used to generate electricity. There was exactly one geothermal plant in existence, which opened in 1911 in Italy, who remained the sole producer of commercial geothermal energy for the next forty-seven years.

What other sources of renewable energy might be important in ninety years' time?  There's tidal power, stratospheric wind, and, just off the coast of Florida you have this:

Note how the Gulf Stream hugs practically every inch of the Florida coast, begging someone to pop a turbine in the middle of it.

You know you want to.

Why does Mr. Morcombe venture so far into the future with his hypothetical solar scenario, especially as he does not seem too interested in technology or infrastructure that might become available between now and then?

Probably to get to the "sticker shock" number here:

Florida is planning to increase its peak electrical generating capacity from 52 GW in 2009 to 62 GW in 2018.  This works out at an increase of 1.9% (1,100 MW) per year.  The growth projections may be conservative given the following factors:

1. Electrical consumption grew at an average rate of 3.6% p.a. from 1980-2005.
2. Florida’s population is expected to grow from 18.8 million in 2010 to 23.8 million in 2030, a growth rate of 1.2%/year.
3. With environmentalists in positions of influence (as we now see in Germany), there would be pressure to phase out fossil fueled automobiles in favor of electric automobiles.
4. Comfort in Florida depends on an ample supply of cheap electricity to keep our air conditioners going from May through October.

Even at the conservative growth rate of 1.9%, Florida would need a peak generating capacity of 289 GW by the year 2100.  Thus over the next 90 years we might expect 289 – 52 = 237 GW of new generating capacity to be created.  What would this mean if the generators were all solar powered? The Martin solar plant shows that 18 MW of solar power requires ~500 acres of mirrors so the needed capacity would correspond to 6.6 million acres for mirrors alone, without allowing anything for energy storage.   Would that have a significant effect on land usage?

Even assuming an affordable solution to the problem of storing energy to cover the times when there is insufficient solar power to match demand, it is not credible to suggest that 19% of Florida be covered in mirrors; clearly solar fails the “scalability” test.

This is the core of Mr. Morcombe's argument: a fifth of Florida covered in mirrors! Those daffy greens! But to get there, he cheats in several ways:

1. Starting with a real predication for the next nine years, he projects that rate of growth will continue unchanged for the next ninety years.

It's never wise to naively assume a linear trend will continue indefinitely. Mr. Morcombe, arguing from an inflated estimate, then calculate that Florida's electricity usage will more than quintuple in the next ninety years, based on what is expected to happen over the next six.

It's nice to hope, but don't count on it.

To illustrate how quickly such predictions become ridiculous, consider the chart of PV solar, above, which shows prices (since 2007) falling at 20% per year. By my calculations, if solar power prices continue to fall as they have been, a Watt of installed solar PV, presently costing about one and a half dollars, in 2100 will cost a mere 3.5 * 10^(-9)! A gigawatt of solar PV in 2100 will set you back . . . $3.50. The entire power needs of the state could be covered for the price of a single used car!

Probably not, though. Linear projections of growth or reductions over long periods of time are not reliable.

2. He use the peak demand for electricity, ignoring power storage and power sharing.

The sun doesn't shine all the time. So if you plan to get most of your power from solar cells, you need to either store some of that power, or share it with other people who can share power with you when you need it. Both of those things are perfectly possible with current technology; both will be vastly easier in ninety years' time.

Mr. Morcombe's way of accounting for this makes no sense. He takes the average expected output of a given amount of solar, and compares it to peak demand. But that solves nothing, because a considerable portion of the time production will be below average. You still need some form of storage or sharing. On the other hand, once you have some means of storage or sharing, Morcombe's estimates are wildly pessimistic; he is comparing peak demand to average production, instead of average demand to average production, or even peak to peak.

The actual average electrical power consumed in Florida is 16GW, not 52GW. And when the sun doesn't shine? HVDC cables to sell power to the rest of the country during high-output periods (when power tends to be more expensive) and buy it back during low-output periods (when power is cheaper because demand is generally less.)


3. Morcombe ignores energy efficiency.

The closest Morcombe comes to addressing efficiency is his tone-deaf assertion that "Comfort in Florida depends on an ample supply of cheap electricity to keep our air conditioners going from May through October." As opposed to, say, building smarter housing or using more efficient air conditioning.

Remember the key element of Morcombe's thought experiment:

Imagine a future in which so-called “environmentalist” politicians are given the mandate to prohibit the construction of nuclear and fossil fuel power plants in Florida. As wind and hydro are not suited to Florida, the only remaining option would be solar power.

Morcombe ignores the idea that we might use less electricity. America is one of the least energy-efficient countries in the developed world:

Japan gets roughly two-and-a-half times as much economic production from a given amount of energy. So if you were determined not to build any more coal, gas, oil or nuclear plants, using energy more efficiently would be an obvious alternative to blanketing the hinderlands with solar panels. Based on the efficiencies achieved in Europe and Japan, cutting per capita use in half is certainly realistic with current technology, let alone nine decades of progress later.


4. Morcombe ignores progress in solar technology.

Besides ignoring other forms of renewable energy, Morcombe assumes that the solar cells of 2100 will be no more efficient or productive than today's, despite the dramatic improvements of the last thirty years. Just in the time since Morcombe's anti-solar screed was published, development of cells that are 16% more space-efficient has been proposed. A technique using "bumpy nanoparticles" has boosted thin-film efficiencies by 8%. And so on.

So what happens if we run Mr. Morcombe's thought experiment with slightly more realistic assumptions? Do we still end up with a massive slice of Florida covered in solar cells?

Let's leave his sky-high 1.9% growth rate alone. We will compare average supply to average demand, however (16GW, not 52GW at the start: 16/52 = 0.307). We'll give solar panels that 16% increase in space efficiency (1/1.16 = 0.862) and a conservative 20% boost in sun-catching efficiency (1/1.2 = 0.833). We're accepting the unrealistic growth rate of 1.9%, but we're going to presume that in ninety years we will use power as efficiently as most of Europe does today (1/2 = 0.5). We're left with 19% * 0.307 * 0.862 * 0.833 * 0.5 = 2.1%. Minus a little more for energy efficiency's impact on current consumption, plus some transmission losses shipping power around, minus any other renewables. Call that a wash: 2.1%.

Put half of it on building roofs and half in the hinderlands. Easy. And beware climate "skeptics" bearing assumptions.
--------------------
1.) Readers from the future, be warned. This sentence has an expiration date which is fast approaching. In 1913 C.W. Leadbeater published Man: How, Whence, and Whither? It posited "atomic energy." In 1914, HG Wells got in on the act with The World Set Free, imagining atom bombs as well as atomic power plants. So nuclear power was undreamed of ninety years ago, but only just.