Rooftop solar panels considered harmful?

“The Hole in the Rooftop Solar-Panel Craze” is a Wall Street Journal editorial (May 17, 2015) that heaps scorn on the way that America’s crony capitalist system encourages domestic rooftop solar power. Here’s a sample:

Recent studies by Lazard and others, however, have found that large, utility-scale solar power plants can cost as little as five cents (or six cents without a subsidy) per kilowatt-hour to build and operate in the sunny Southwest.

Large-scale solar-power prices are falling because the cost to manufacture solar panels has been decreasing and because large solar installations permit economies of scale. Rooftop solar, on the other hand, often involves microinstallations in inefficient places, which makes the overall cost as much as 3½ times higher.

Yet the federal subsidies for solar amount to about $5 billion a year, with more than half of that amount going to rooftop and other, more expensive, non-utility solar plants. If the federal government spent the $5 billion instead subsidizing only utility-scale solar plants, I estimate that it could increase the amount of solar power installed in this country every year by about 65%. And without net metering and all of the other nonsensical state and local subsidies for rooftop solar, we could save this country billions of dollars every year.

The author doesn’t calculate the full amount of the wasted dollars because, presumably, it is too hard to find out what each of the 50 states is doing.

First, do we believe this guy? Brian H. Potts is the author and (1) he is a lawyer who works mostly for utilities, (2) he doesn’t look old enough to shave.

If Potts is right, is it reasonable for him to expect a program run by the U.S. government to be efficient? Car emissions reductions, for example, have been handled in what economists would call the dumbest and most expensive possible way. Instead of measuring emissions every year when cars are inspected and taxing each car owner according to miles driven and pollution emitted per mile, standards are promulgated for new cars and society has to wait 10-20 years to see an effect. The result is that a small percentage of older/mistuned cars generate most of the pollution (example study). Why wouldn’t we expect solar energy production to be handled in a similarly inefficient manner?

Related:

13 thoughts on “Rooftop solar panels considered harmful?

  1. Utiltied hate the solar panels on customer rooftops. They are trying hard (from Hawaii to Arizona) to just stop it all. Mr. Potts is writing from his wallet.

    I suspect if that money was suddenly “saved,” it would wind up in the pocket of the utilites, not citizens.

    My electric bill, even with my electric cars, was $12 last month for a 2,000 square foot house.

  2. If you were a Congressman looking to be re-elected (as they must every 2 years) and there was $10M of Federal solar subsidy available in your district every year, would you rather hand out the whole $10M to one utility (whose HQ is not even in your district) or give out 2,000 subsidies of $5K each to grateful voter families?

    I’m not sure I understand the economies of scale here – why is it cheaper to put all the solar cells in one distant place (with power line losses on the way) vs having the same cells right on the roof of your house?

  3. Presumably you need fewer resources, have better trained employees, more utilized equipment and assets, better siting, less use of gas to move maintenance/installation workers around, …

    Something like 3 guys can maintain 1 acre of centralized solar cells laying on an acre of desert, but 10 guys needed to maintain 1 acre of decentralized solar cells spread across 100 homes. And a hundred acres of centralized solar cells might come with one large tesla battery instead of needing 100 tiny tesla batteries.

    But then there are line losses…

  4. The proper question is not what technology to fetishize. The evaluation should look at how many tons of CO2 emissions are eliminated and at what cost. The metric of dollars/ton of CO2 is applicable to all technologies and all products. There is more than this to an evaluation, but the dollars/ton is very important.

    At the gross scale, you find that the Energy Star program is estimated to have reduced US emissions by 300 megatons/yr in 2014. Solar PV (all programs) installed in 2014 reduced emissions by 30 megatons/yr in the US. Finding cost information is extremely hard given the dozens of tax credits involved, and hundreds of different programs.

    Efficiency projects are usually the best cost/ton projects, consistent with the 10:1 ratio of CO2 reductions in 2014. One steel mill that redesigned and replaced its air pollution control system reduced electricity consumption at $8/ton equivalent in the 2014 TVA efficiency program. Since electricity cost was about $75/ton equivalent at than site this was also a huge cost saving. (The approximate conversion factor is 0.5 kg of CO2 per kWh in the US.)

    When you go down to the individual project scale, you do find PV programs that are cost competitive with the alternatives, as well as many PV programs that are ridiculously excessive.

    One of the worst is the MA $0.50/kwh renewable credit, which translates to $1000/ton equivalent. But PV powered railroad switches are less expensive than grid powered switches (due to construction issues) in a switching yard, and PV powered highway signs are less expensive than generator or grid powered signs.

    Solar panels on houses are a complex system problem, involving much more than just the cost of the PV. There are issues like voltage regulation stability, reactance stability, peak power demand, demand management, transformer capacity, safety cutoffs, ground current, transmission capacity, etc. that the utilities must deal with.

  5. Rooftop solar is more expensive than centralized solar, just like it is more expensive than centralized natural gas power generation, but that is irrelevant. When you factor the.costs of transmission and the utility’s profit margin, it is more cost-effective for the consumer. The main limitation is lack of energy storage options. I wouldn’t be surprised if it takes over completely in Hawaii within a decade or two.

    http://spectrum.ieee.org/green-tech/solar/residential-solar-power-heads-toward-grid-parity
    http://bert-hubert.blogspot.com/2014/11/tin-cans-can-openers-and-solar-power.html

  6. It’s interesting because the article dwells on (and I think conflates) two things in favor of large solar projects:
    1. Economies of scale
    2. Advantageous siting

    While completely ignoring the benefit of decentralization of electricity generation.

    A given set of solar cells (the cells themselves) is going to cost roughly the same amount when bought in bulk whether it’s a utility buying a few acres worth or a large reseller buying the same amount. If they are destined for a mega-installation in the Mojave desert, the utility will be marking them up to pay for the capital costs of the land under the installation, and the mounting, inversion, control, and transmission equipment that’ll be installed (which the utility will then own). If they are going onto residential or commercial roofs, the “land” is free, but there will be markups from the solar cell distributors, and the installation will be built out using slightly more expensive components by much smaller construction firms. I suspect that there is going to be a higher cost/watt when it’s distributed this way.

    However, the article is comparing the entire market of roofs where people live to ideal sites in the mojave desert. As long as the cost/area of the solar cells themselves is a substantial portion of the cost of the energy, it will make much more financial sense to put solar cells on your roof in Phoenix or Los Angeles than it will in foggy Monterey, but I think if you lump Monterey in in the comparison of big vs. small solar, you’re putting your finger on the scale a bit.

    It also ignores the fact that the Utilities will own the power generation in one scenario, and in the other, the homeowners, and the power generation becomes geographically distributed, no longer relying quite as much on the transmission infrastructure.

  7. Given the current economic distortions (net metering, federal tax credit, etc.) and other factors such as low interest rates, home solar generation is certainly viable, especially in states like Massachusetts with high electricity costs. We went live this month.

    And there are intangible benefits, like giving me an excuse to reach for an editor and play with the API to generate pretty graphs. Sample here: http://blogboxen.com/pages/solar.html

  8. A $12 electric bill is great except if it cost you $35,000 to save $150/month.

  9. I am leasing the panels. The lease costs me $120 a month. My electric bill was nudging over $300 a month (to $325 during the summer). I don’t believe SolarCity makes this deal anymore.

  10. $150/month from a $35K investment is just over 5% per year, tax free.

    In our case, ignoring SRECs, the guaranteed production from year one will provide a rate of return that is over 8%… With SRECs, best guess will be complete payback in 5 years. I suspect a bit longer, but still within reasonable time-frames.

    But please note: that is with the 30% Federal Tax CREDIT.

  11. I’m a little confused by the “untuned cars” in that example study. I don’t think cars have needed to be “tuned” for about 25 years. The last car I had without a engine computer that handled it was a 1988 model. Since then the maintenance has been changing spark plugs every 100,000 miles and new filters. I thought cars generally aged out of usage at about 17 years now? Seems like car tuneup should be an archaic phrase by now.

    The study did mention “black smoke”, so I assume they meant old diesel trucks. Engine management computers and fuel injection became common for those around 2000, so hopefully those will age out over the next few years too.

  12. John V: Replace “untuned cars” with “old cars” then. It doesn’t make sense to have a system that does not discourage people from driving old cars that pollute much more than the latest and greatest.

  13. An emissions tax might make sense now, but not so much when the current system was designed 50 years ago. At that time there was no auto emissions testing infrastructure, no internet for moving data around, minuscule computing power for managing the data, less difference between clean vehicles and gross polluters, no clear technical path for achieving the goals, auto manufacturers formed an oligopoly much less responsive to market pressure, and the public would have been less receptive to the inconvenience of periodic testing. Sure, alternatives to the current system might be better for the current task of managing a fleet of clean vehicles, but I’m not convinced that the task of cleaning up the fleet (which saved the auto industry from itself) was accomplished in the “dumbest and most expensive possible way”.

Comments are closed.