White House goes Green – and into the Red

This is a guest post by Roger Andrews.

In August 2013, at the instigation of President Obama, 6.3kW of solar PV panels were installed on the roof of the White House. A video commemorating the event was recently released in which a number of administration officials waxed lyrical about what a fine example this installation sets and how solar power is now a reality and how it’s going to underpin the US’s transition to green energy and so on and so forth:

Inside the White House: Solar Panels

[Image: President Obama wonders why the panels are not pointing at the Sun and wonders how they will perform under a yard of snow in winter time.]

With such a buildup one might expect that the White House solar system would be an economic barn-burner, and while the White House doesn’t go quite that far it does claim payback in eight years, which is not too shabby: (“The project …. is estimated to pay for itself in energy savings over the next eight years,” the White House said.) But how good are the economics of the White House solar system, really? Well, they turn out to be an unmitigated disaster.

First I analyzed them on a stand-alone, unsubsidized basis. The White House declines to supply data on system cost or performance, so I based the analysis on the following assumptions for a DC residential installation of comparable size (links to data sources provided below):

  • 25- year panel life
  • Installed cost, 6.3 kW of panels = $31,500, or $5,000 kW (1, 2)
  • Average Washington DC residential electricity price = 11.88 cents (3)
  • System load factor = 13.0% (3)
  • Panel output decreases by 1% a year (5)
  • Replace inverter in years 10 and 20 (best-guess estimate from a number of sources)
  • Inverter replacement cost $5,600 (prorated from the quote for my solar system inverter)

(1) Solarpowerrocks
(2) How Much Solar Panels Cost
(4) Wikipedia: List of monitored photovoltaic power stations
(5) Solar Efficiency Losses Over Time And here are the resulting cash flows:

Without subsidies the system has a rate of return of minus 10.6% and a cumulative cash flow at the end of year 25 of minus $24,000. The system in fact never pays back regardless of how long it stays in operation.

So how does the White House come up with an 8-year payback? By including subsidy payments. Solar systems in Washington DC are eligible for subsidies, rebates and tax credits which when added together are arguably large enough to qualify as charitable donations. Here’s a summary of them, excerpted from reference 1 above:

Installing a typical 5kW solar system should start at about $25,000. Don’t freak – that’s gonna drop fast.

  • You start with that big rebate from the city government. You get $4,500 for the first three kws and $1,000 for the last two kws, so subtract $6,500 for a new starting cost of $18,500.
  • The feds are smart – they calculate your 30% tax credit on your costs after state rebates. So we take 30% of $18,500 and subtract another $5,550 for a new price of $12,950.
  • Don’t forget about those SREC (a solar renewable energy credit for each mWh of electricity delivered to the grid) payments. Using the current $320/SREC average, we estimate your first year SREC sales to bring back $1,872. Subtract that for a new cost of $11,078
  • Finally we subtracted your estimated annual electricity savings of $730, giving us a total cost after year 1 of just $10,348. That’s a discount of nearly 60% already.
  • With a conservative estimate for the future rise of electricity prices, you can expect your new solar power system to pay for itself in just 5 years!

The summary culminates with the graphic reproduced below, which shows that if you have ~$10,000 to invest then you can’t do better than to invest it in subsidized solar:

It’s not in the least surprising that the economics of a DC solar system show a huge improvement when these subsidies, rebates and tax credits are added in. The rate of return increases to 20% from minus 10.6%, the cumulative cash flow in year 25 to plus $42,000 from minus $24,000 and payback from never to a little over five years. (The White House’s eight-year payback probably reflects an unusually expensive installation. An extra ~$8,000 in installation costs would give an eight-year payback.) The table below shows the cash flows for the subsidized case:

The problem, however, is that the economic gains are contributed entirely by US taxpayers, who over the life of the White House system will fork out $66,000 in subsidies (calculated as the difference between the cumulative year 25 cash flows for the subsidized and unsubsidized cases), which works out to a subsidy of 0.42 cents/kWh – more than three times the DC average residential rate – for each of the 156,000 kWh of electricity the system generates over its lifetime. (The residential rate would in fact have to increase to 38 cents/kWh before the system earned a return of 5% without subsidies and to 27 cents/kWh just to make it break even at the end of year 25.)

Now it wouldn’t be too bad if the US taxpayer was underwriting only the losses incurred by the White House solar system, but according to the video a comparable system is being installed on a US rooftop once every four minutes, which translates into 131,400 White House-sized solar systems and about 800 megawatts of new solar capacity each year. It’s impossible to calculate exactly how much subsidy these systems would be eligible for over their lifetimes without doing a state-by-state analysis, but as Everett Dirksen once famously observed: “A billion here, a billion there, and pretty soon you’re talking real money.” (Hawaii, incidentally, is the only US state where unsubsidized rooftop solar is presently economic, but only because electricity rates there are about three times the national average. And is rooftop solar subsidized in Hawaii? Of course.) Roger Andrews is a British born, naturalised American mining consultant who is now semi-retired and lives on the West coast of Mexico where he spends some of his time sitting under a wavy palm tree blogging and drinking tequila.

This entry was posted in Energy and tagged , , . Bookmark the permalink.

37 Responses to White House goes Green – and into the Red

  1. Roberto Zavattiero says:

    I guess that on the chart you mean 6.3kWp installed not 6.3KWH

  2. peter2108 says:

    The calculation of $42K over 25 years assumes that the US Gvernment will not renege on the subsidies. This is a strong assumption!

    • Roger Andrews says:

      The government giveth, the government taketh away. But no reneging until Obama’s out of office, so it won’t be him destroying the planet.

  3. Euan Mearns says:

    Great post Roger. I had not realised the scale of subsidy required to make solar work. We hear so much about falling costs and nearing cost parity with traditional generators and so there is perhaps an argument to tolerate subsidies for a while while the industry gains traction and scale? Cost parity should of course include storage.

    Solar in sunny climates does make some sense to me, but it requires higher electricity prices like you have in Mexico.

    It does seem however that Obama has traded The American Dream for the European Nightmare.

    PS – you could have added 5% financing costs to get an even gloomier picture.

  4. Hi Roger,

    Thank you for an interesting calculation.

    “Hawaii, incidentally, is the only US state where unsubsidized rooftop solar is presently economic, but only because electricity rates there are about three times the national average.”

    It is common to use tiered pricing in California as a cross subsidy. Many homeowners are paying a marginal price of more than 30 cents per kWh. Solar works for them also.


    • Roger Andrews says:

      H Dave:

      You are quite right. Tiered pricing does that have that effect. However, my claim was based on average residential electricity rates, which according to EIA were 15.94 cents/kWh in California and 38.51 cents/kWh in Hawaii in March 2014.

      It is of course possible that the economics of the White House system are improved by tiered pricing in Washington DC, but if so the effect isn’t large. The “sales pitch” I quoted lists annual electricity savings of $730, which works out to about 13 cents/kWh, only a cent or so more than the price I assumed. If tiered pricing had added any more icing to the cake I’m sure the salesperson would have mentioned it. 😉

      • Euan Mearns says:

        Roger, is it not tiered pricing that makes your Mexican PV profitable? I’m actually in favour of social justice and have little objection to a power pricing system that ramps up for the wealthy. Like all such systems it has problems. In this case it would naturally make solar PV the domain of the rich – but it is expensive, so there is a degree of natural order there.

        If tiered pricing had added any more icing to the cake I’m sure the salesperson would have mentioned it.

        I’m quite sure the sales person does not know what a photon is, what electricity costs in DC and what tiered pricing means – I don’t 😉

        • Roger Andrews says:


          Re electricity prices in Mexico, I don’t have the exact numbers to hand but from memory it goes like this:

          Effectively all of the electricity consumed here is generated and sold by Comisión Federal de Electridad, a state-owned company. CFE has two basic residential rates – around 10c/kwh for low usage – just about enough to run a refrigerator, a TV and a few 60W light bulbs – and around 30c/kwh for anything over that. The 10c rate is heavily subsidized (CFE claims that its generation costs are closer to 30c than to 10c) to accommodate the poor people who couldn’t afford to light their homes if they had to pay the true cost of generation. The higher rate – well, that’s for the rich people like me who can’t live without large-screen TVs, computers, freezers, automatic garage doors and air conditioners.

          Is this tiered pricing? I guess it’s a form thereof. Does it make solar the domain of the rich? Of course; if you’re poor there’s no point having it. Is my rooftop solar system socially just? Well, the only segment of society I’m taking money from is the Mexican government, but the Mexican Climate Change Law of 2012 encourages everyone to go green, so I guess my solar system is OK with them, and when I installed the system I pumped several thousand dollars into the Mexican economy that wouldn’t otherwise have been there. So once again I plead innocent to all charges 😉

          Re the Washington salesperson. I don’t doubt that he or she didn’t know what a photon is (I’m not sure anyone does), but I’ll bet you a bottle of the best tequila (palm tree not included) that he or she knew exactly what DC electricity prices were. The people who installed my system had them nailed down to the nearest centavo.

    • kakatoa says:


      Page 8 in the following document denotes the current residential rate structure for the Big three (PG&E,SCE,SDG&E) CA utilities.


      Residential rates increased recently leading to a marginal cost of $.3595/kWh for Tier 4 usage with PG&E. The Average price for a kWh for E-1 residential customers just hit $.20 kWh for non-care PG&E customers.

      • kak,

        Thank you for the fabulous link. My utility is municipal (Pasadena) and I did not realize that the huge cross subsidies go back only to 2001.


        • kakatoa says:


          I would prefer to have SMUD as my service provider as it would mean 20 to 25% less cash flowing out of my wallet for my kWh usage. One of their hydro plants is about 10 miles from our place, as is one of PG&E’s. I would have to pay a separate fee to SMUD though for my little PV system to cover some of their fixed costs to provide service- which includes their acting as the battery when I send energy into the grid via my net meter. I believe this fee, for any size PV system, is going up to $20/month. I would gladly pay this fee in trade for how they set baseline levels.

          SMUD is aware that folks that use electric clothes dryers use more kWhs of electricity so they adjust their baseline levels for folks that use this means of drying clothes- vs say using PG&E’s natural gas to accomplish the task. Unfortunately, PG&E does not take this use of energy into account in their electrical energy baseline levels…. When you add in the energy needed to run our well you can understand why we decided not to trust PG&E, CARB and the CPUC to fairly set prices for folks that live in the country without the “G” in PG&E’s services and put in a PV system 8 years ago. We missed the $.49 kWh tier 5 prices that led to the rate revolt in the Central Valley a few years back. Legislated updates now allow the CPUC and PG&E to increase Tier 1 and 2 kWh prices. PG&E is recommending (2014 GRC) that the CARE rate schedule have their subsidies reduced to 30 to 35%. It will be interesting to see how this plays out.

          I am not sure if you have had a chance to read CARB’s update to the climate change scoping plan:

          For the life of me I can’t figure out how my transportation costs to travel a mile are going to be less in 2020 than they are today, but from a macro perspective this is stated in the plan; B. Foster Resilient Economic Growth pg 27:

          “ The combination of California’s vehicle GHG and Zero Emission Vehicle (ZEV) standards and policies adopted under AB 32—including the Low Carbon Fuel Standard, SB 375, and Cap-and-Trade—will reduce per-capita fuel costs and GHG emissions from light-duty vehicles and fuel use by about 30 percent from current levels in 2020, and by about 50 percent in 2035 (see Figure 5). Additional measures to reduce emissions could further reduce fuel costs, as well.”

          By 2050 I am likely going to be dead, but I was wondering how the next owner, or renter, of the property I live on will address their mobility needs and somehow manage to do this while also having reduced fuel costs. My gut tells me that fuel costs is the wrong metric to be looking at here…. levelized cost to travel a 100 miles or something like that seems more appropriate. Oh well, that will be their problem. I guess the new resident at my place could always upgrade the electrical service to a higher Amperage service to allow for charging of an EV. As to what it will cost ($/kWh) to charge that EV the plan doesn’t cover. This does seem a bit odd as a lot of modeling and real world data is available……….

          One does wonder why the update to the scoping plan doesn’t include a graph noting the expected costs for a kWh of energy between now and 2020 and then to 2035. They have rather firm estimates for this. Many, most?, of the must take PPA contracts have been put in place for the utility scale RE to meet the 33%RES……

  5. Craig says:

    But wait a minute. Aren’t fossil fuels subsidized a bit by solar and wind? In general, wouldn’t the price of fossil fuels be much higher and harder to bring to market if alternative energy systems like solar and wind were not already carrying a portion of the load? And of course, fossil fuels are subsidized by efforts to raise gas mileage on cars, and more efficient light bulbs and refrigerators do the same. Even double-paned windows keep the price of natural gas down further than it might be.

    It would be curious to take about ten things that have been around for thirty years and estimate how they have affected the price of fossil fuels. You might also have to include the savings on an extra two or three wars. Interesting calculations.

  6. Paolo Pulicani says:

    Here in Italy: cost for 6.3 kWp rooftop mounted solar panel $ 17,200 (to the final customer).

    Cost breakdown (per kWp):
    Modules: $816 (€ 600)
    Inverter: $300 (€ 300)
    Cables, switches, mounting frames: $272 (€ 200)
    Installation: $408 (€ 300)
    Company revenue (20%): $544 (€ 400)
    VAT (10%): $ 272 (€ 200)
    Totale (per kWp): $ 2,720 (€ 2,000)

    • Euan Mearns says:

      Paolo, thanks for this information. I always thought Roger’s initial price was very high, but it is backed up by “data”. Can you clarify what “Company revenue” is. And it would also be interesting to know what kind of subsidies are available in Italy for installation and power generation.

      A recurring theme on Energy Matters is that one size does not fit all. With lower costs and higher electricity prices and a sunny climate it is clear that solar may be more viable than in a cloudy climate with low electricity prices and high installation costs.

      The intermittency problem will remain until affordable grid scale storage is developed. But we already have grid scale storage in the form of U isotopes.

    • Roger Andrews says:

      I have estimates of installed costs for solar PV ranging from $1,000/kW to well over $10,000/kW (both in Australia).

      Here’s the best information I’ve been able to find, although it’s a little dated:


      Installed cost doesn’t actually make much difference to comparative (i.e. subsidized vs. unsubsidized) economics. At Paolo’s installed cost of $2,720/kW the taxpayer still forks out $59,000 in subsidies over the life of the White House system.

  7. Dennis Coyne says:

    Hi Roger,

    The analysis would be a little more believable if you had not assumed constant electricity prices.
    The EIA data for DC http://www.eia.gov/electricity/data/state/
    shows a 1.07 cent per kwh rise in price for full service residential electricity each year over the 2004 to 2010 period, in 2012 prices were 12,25 cents per kwh. A more realistic future scenario would look at the payback with prices rising by 0.5 cents per kwh each year over 25 years or linearly from 12.5 cents per kwh to 25 cents per kwh. In addition installation costs will fall to European levels (40% lower than median US costs) over time reducing the initial system cost $20,000. Also the use of micro-inverters eliminates the need for inverter replacement, they typically have a 20 to 25 year warranty and also tend to increase system output by 15% or so.

    See http://runonsun.com/~runons5/blogs/blog1.php/solnews/solar-trends-storage-lower-cost

    for more on micro-inverters.

    • Roger Andrews says:


      You’re quite correct. US electricity rates have indeed been increasing at about a percent above inflation since 2004, but I used constant dollars in the analysis to avoid having to make any more assumptions about what’s going to happen in the future than I’d already made.

      Anyway, I plugged in your suggested price increase of 0.5c/year to see what it did. I didn’t change the inverter replacement costs because as far as I can tell from to the video the White House system has string inverters and not single-panel microinverters. I also made no allowance for potential future installation cost reductions because the installation costs are already sunk.

      The impact was to raise the unsubsidized rate of return from minus 10.6% to minus 4.4%, with payback still nowhere in sight, and the subsidized rate of return from 20.0% to 23.4% with the same ~5 year payback as before. Project-life subsidies underwritten by the taxpayer, however, increased from $66,000 to $77,000.

      Incidentally, the link you posted takes me to the Google home page. 😉

      • Roger Andrews says:

        Correction: The subsidized rate of return goes from 20.0% to 21.5% with the same $66,000 taxpayer outlay as before.

    • Euan Mearns says:

      Hi Dennis, thanks for this valuable insight. Any idea why prices are so much higher in the USA when the opposite is typically the norm? Power costs will certainly rise and the hardware costs will fall.

      But all this stuff currently exists in a ZIRP environment (zero interest rate policy) and replacing subsidies with tax and interest kills it all stone dead (including shale). Tax and interest are a product of/ enabled by cheap fossil fuels which as you know are running scarce.

  8. John B. says:

    From the article: “Hawaii, incidentally, is the only US state where unsubsidized rooftop solar is presently economic, but only because electricity rates there are about three times the national average.”

    The article fails to mention the “reason” why electricity rates are so high in Hawaii – it’s because they burn crude oil for electrical power.

    So therefore, solar power is now cheaper than crude oil. Which means Peak Oil doesn’t matter, and never did.

  9. Roger Andrews says:

    It’s true that electricity rates are high in Hawaii because most (actually over 70%) of the electricity is oil-fired, but that doesn’t necessarily mean that solar power is cheaper. With a feed-in tariff of 21.8c/kWh Hawaii rooftop solar is in fact one of the world’s most costly generation sources.

    Peak Oil wouldn’t be a problem if everyone drove a solar-powered vehicle, but few of us will be alive to see that. In the meantime the world still gets almost a third of its energy from oil, so we’d better hope that it lasts a little while longer.

    • John B. says:

      1. The feed in tariff amount has nothing to do with the cost of solar power.

      2. Solar power doubles every 2 years. So unless you’re planning on dying in the next 14 years (I’m not), yes you will be alive to see it.

      • Roger Andrews says:

        “1. The feed in tariff amount has nothing to do with the cost of solar power.”

        The 21.8c/kWh feed in tariff (27.4c/kWh with state and federal tax credits added) is the price Hawaiian Electric pays for solar power delivered to the grid. Levelized costs for Hawaii solar are in the same range (Hawaiian Electric median estimate 28c/kWh, my estimate 25c/kWh).

        “2. Solar power doubles every 2 years. So unless you’re planning on dying in the next 14 years (I’m not), yes you will be alive to see it.”

        That works out to an increase of a factor of 128 by 2028. In the unlikely event this can be achieved solar would still fill only about 16% of expanding world energy demand in that year (right now it fills 0.2%). But not to worry. If solar keeps doubling every two years after 2028 it will supply 110% of world energy demand by 2034, although I doubt that I will be alive to see it.

      • kakatoa says:


        Your thoughts on tariffs and costs: “The feed in tariff amount has nothing to do with the cost of solar power”
        are opposite of my experience.

        My experience, as an owner of a residential PV system, as an attendant at a few California Energy Commission meetings on the cost of utility scale PV projects, and as an interested fellow on how CA is going about meeting it’s 33%RES though it’s long term PPA’s leads me to say the opposite of your comment.

        I have sat in meetings where the costs to meet the states RES was stated as being irrelevant to the goal of reducing greenhouse gas emissions- really tail pipe/stack emissions as the C02 accounting didn’t include the new transmission lines or the backup scalable power idling away for when a cloud comes by. The costs would be dealt with in the tariffs. At the time those costs were noted at $200.00 to remove a ton of CO2 from the electrical generation mix in CA. NOT covered in these costs was the new transmission lines required to bring the energy to the grid.

  10. Mike Parr says:

    Doing a hybrid project on an island (oil fired generation & we will add PV + wind). For the first project & we will produce kWhrs at a 20% discount to that from oil (with a circa 60% RES penetration) IRR 20% and – 100% private financing and zero, I’ll repeat, zero, subsidies. RES in the right place (certainly islands – maybe elsewhere) is far cheaper than oil. On a related note PPAs in Texas for PV at 5UScents/kWhr, something similar for wind. Sure these are subsidised (production tax credit) – just like the fossil guys get their subsidies

    • Euan Mearns says:

      Mike, a recurring theme – one size does not fit all solutions. The guys in the Space Station appreciate their solar panels. But the Mars Rover has a nuke on board. If the owner of the diesel generator (?) also own the renewables then you have a system I would advocate for all where the owners of renewables also provide the back up / balancing generation capacity. If the renewables are owned separately then the owner of the diesel generator will lose market share and money and may have to raise prices to stay solvent.

      Sure these are subsidised (production tax credit) – just like the fossil guys get their subsidies

      Would you like to elaborate what subsidies the fossil guys get.

    • In Energy in Australia I covered some case studies, one of which was the King Island diesel/wind/solar system. Other similar examples in Australia include Esperance, Hopetoun, Denham, and Rottnest Island. The Flores and Graciosa Islands in the Azores, Portugal, also provide similar, highly integrated examples of wind/diesel systems in isolated small grids. These systems make a lot of sense due to the high cost of diesel generation and wind/solar provides a straightforward fuel-saving role. Comparing the cost of providing electricity on King Island and mainland Australia, I concluded:

      The per capita cost of the electricity system on King Island is around seven times higher, and when taxable income is included, the relatively affordability (excluding subsidies) grow to a tenfold difference.

  11. Mike Parr says:

    Oh please Euan, don’t come over all innocent – the USA fossil people get all sorts of subsidies – public record & all that – e.g. the fracking people – master partnership agreements ring a bell? Or let’s try this:

    .March 2014: IMF joins World Bank, IEA, OECD etc in calling for subsidies on fossil fuel to be eliminated…….says subsidies for petroleum products, electricity, natural gas and coal equal about 2.5 % of global GDP, or 8 % of government revenues. Advanced economies pay out about 40 % of the global total, the report says, listing the top three subsidizers as the US at $502 billion, China at $279 billion and Russia at $116 billion.

    Or perhaps this: (2012) Environmental Law Institute: Between 2002 and 2008, RES (in the USA) received $12.2 billion in government support, with $6 billion in direct spending and $6.2 billion in tax breaks. Fossil fuels had $70.2 billion in the same time, $16.3 billion directly and $53.9 billion in tax breaks

    I could go on & on & on – but I think I have made the point, glass houses etc.

      • Mike Parr says:

        Gosh ….do I believe a Forbes contributor? (who does not cite sources) or,…. the IMF, World Bank, IEA, OECD – yes I can see how that is a tough choice.

        On a “sort of related subject” just remind me how much energy (money?) goes up the smoke stack when converting subsidised fossil fuel into electricity? Whilst it does depend on the tech (40% up the stack for CCGTs) it certainly equals things up (old coal 65%?).

        Of course the other question never discussed, is why after 100 years++ of extraction do fossil fuels still need subsidies – I can only conclude that it must be socialists that are giving them all those subsidies. Shame on these socialists and on socialist magazines such as Forbes … oh er hang on….

    • In Australia, primary producers (farmers and miners) get a rebate on diesel excise meaning that they don’t pay the 38 cents a litre for diesel that everyone else pays – this is the largest of the “fossil fuel subsidies”. This is a legacy of the tradition of a pioneer country where primary production is considered essential. I’m not an economist, and most would argue against subsidies, but its also pretty obvious that removing the rebate would increase the cost of food (say), decrease other taxes, and the whole thing would come out in the wash. It’s hard to see a 300 kW combine harvester running on solar panels so this isn’t a fossil versus renewables debate. Diesel is already a large cost so farmers are already efficient.

      This isn’t an argument in favour of the subsidy but making the observation that farmers need diesel one way or the other if we are going to eat. On the other hand, the Renewable Energy Target subsidises the purchase and maintenance of imported wind turbines, but life would go on as normal if the turbines were turned off. Many people seem to be happy to pay for the wind subsidy and think that farmers shouldn’t get the rebate. I don’t know if there is a single right answer.

  12. Kit P says:

    The biggest mistake in Roger’s cacl is the cost of electicity in Virgina is about 7 cents/kwh on my bill.

  13. Kit P says:

    Just wanted to see if I could post from China where I am working at a nuke plant. The cost of making power is about the same anyplace in the would. The bigeest factor for those with high retail power cost is hidden taxes. The second problem in Roger’s caclualtion is a 25 year of solar panels. If you buy a system I think you be lucky if your system is still working in 5 years. The concept of zero maintence costs is bogus. Five years after being the victum of a solar scam the bubble has burst but you can still point to your roof to show how great an envirmentalist you are.

Comments are closed.