Fight The Snobbery - Maybe Solar Efficiency Is Good Enough
    By Hank Campbell | January 9th 2012 11:41 AM | 10 comments | Print | E-mail | Track Comments
    About Hank

    I'm the founder of Science 2.0®.

    A wise man once said Darwin had the greatest idea anyone ever had. Others may prefer Newton or Archimedes...

    View Hank's Profile
    Solar photovoltaic (PV) mass-market polycrystalline panels are typically about 15% efficiency. Pretty terrible, right?  Maybe, maybe not.  

    Cars - both gasoline and electric vehicles, despite what the electric car hype machine claims - are only slightly more efficient at 15-25%, as is eating, and biology had all of existence to perfect that process. Ethanol and other biofuels, which were the darling of anti-science environmentalists in the Anything-But-Oil camp before electric cars because they are the ultimate solar-powered device, are a dismal <2% efficiency.

    Yet maybe an efficiency fetish is just as misguided as the Anything-But-Oil one. Regardless, efficiency is what people talk about and that may be a mistake - there are lots of other reasons why solar power in current form is unworkable but efficiency may be just as misguided when critics bash it as when zealots promote it.

    Efficiency can be improved, and has been in for NASA applications (to around 40%) but those manufacturing costs are insane. It doesn't keep the interns working at the Department of Energy who are squandering $35 billion on corporate welfare from insisting the Miracle of Capitalism will bring those costs down, ironically requiring massive taxpayer funding first to do so, but in the non-government world where we don't get to play with Monopoly money, actual science and economics are in force. So let's stick to data and the data says solar is not great, and maybe it never will be, but it isn't awful.

    Writing at Do The Math, Tom Murphy, associate professor of physics at the University of California, San Diego, tackles the evolution of PV technology and why the efficiency concerns are not all they are cracked up to be.

    He is right. Even current silicon is reasonably efficient compared to mature technologies so why Energy Secretary (and Nobel-winning physicist) Steven Chu was determined to fund copper indium gallium diselenide (CIGS) companies (not basic research, actual companies, to the tune of a million dollars per job in government funding) remains a mystery of how astute scientists can lose their minds when they enter politics, just like anyone else can.

    But solar won't get better any time soon, especially if we are rewarding old technology we know is not great with unlimited government money. "There are really good reasons why the efficiencies will not climb to arbitrarily high values. Basic physics stands in the way, and I am left impressed with what we’ve got," he writes.   But I have a whole book coming out in September that has a section dealing with the physics obstacles so read his piece in the interim.  He does a terrific analysis of how we got to where we are.

    Of course, cultural obstacles also remain if solar power is only almost as good as the low end of cars; just like you can't get a job writing at the New York Times by telling an editor you are as good as their worst employee, existing technology cannot replace an incumbent by being on the low end of what existing technology has. Cars replaced horses - even without a Department of Energy to pork barrel Ford Motor Company - because they were clearly better than horses even while the cost was far higher.

    Solar power, regardless of efficiency, has a much different problem - all the land it will take:

    Land area, shown as black dots, needed to produce 18 terawatts. Source: WikiMedia.

    It's only a dot, right?  Well, America has far more open space than Africa and many claim that is still not enough - and that dot, as Murphy notes, is more actual space than every paved road in America. Like when ethanol was implemented rather than gushed about, environmentalists will turn on solar rather quickly when they stop buying into environmental corporation marketing claims and see what this actually means for the environment.  You think roads need repaired a lot?  Wait until you get the maintenance bill for solar panels.

    Don’t Be a PV Efficiency Snob - Tom Murphy, Do The Math
    H/T Dan Drake, who sent me the link on GooglePlus


    Thor Russell
    I am sure that there is already more corn planted for ethanol than the size of that black dot. (i.e.enough land to power the whole world with solar not just the US, is wasted growing corn for ethanol) Feel free to look it up if you don't believe me. Given that we both agree this is essentially land wasted, why not just replace it with solar panels. It would probably be better for the environment too to have essentially nothing growing there than plants that cause environmental problems such as fertilizer runoff. ( I use this as a simplistic example to demonstrate where the land could be found without causing major disruption. A more balanced approach such as using that land to grow food and setting aside marginal land for solar would make more sense in the big scheme of things.)
    I really think you should back up your claims with data because I have seen costs for solar farms and the maintenance bill is very small compared to the cost to set it up, and in absolute terms also. Do you have any hard data to contradict this?

    Thor Russell
    There are no large scale solar farms in the US but you can't comment with vague claims like that more corn is planted for ethanol than would be needed for solar panels and then say "Feel free to look it up if you don't believe me." Where did you get that evidence?

    On small-scale solar, the maintenance costs are quite high for the energy they produce, so it is reasonable to assume replacement costs will not magically be reduced just because there is 5 miles of something rather than 1 acre.

    Ethanol has been a disaster because people subsidized and mandated it in defiance of the science - duplicating that in solar is not smart.  Again, being almost as good as the worst thing is not how change gets made.
    Thor Russell
    According to wikipedia, there are 10 million hectares set aside for corn based ethanol. This appears to be at least as big as the black dot you have shown, hence that claim that it could provide that much energy. If you know the size of that black dot in more detail, then it would obviously make it easy for us to compare directly.
    I agree that maintenance costs may be high for rooftop solar, but I have seen it claimed that the maintenance costs for solar farms are far less. I can't immediately find the link it was a while ago. I would have thought that it would be obvious that it would cost a lot less to maintain a solar farm on a flat piece of land than have to scramble on multiple roofs etc. I suppose I will have to try and find the link again. Do you have any data of your own? 

    I am not recommending US policy, just pointing out that it could be done. I think in terms of solar, the baton has passed to China now, it produces more than the US, and its costs appear to be dropping faster. Whatever financing arrangements are going on in China they seem to be able to afford and indications are they have a policy to aggressively push them. Whether things stay that way remains to be seen.
    It appears that solar now has a clear advantage over fossil fuels without subsidies in the developing world if the financing can be made right. (finance that would make commercial sense)
    No doubt the Chinese solar companies will be trying to full this need.
    Thor Russell
    the baton has passed to China now, it produces more than the US, and its costs appear to be dropping faster.
    They can have it.  The US government seems to think there is a 'race' to produce solar but the marketplace does not - the Chinese are subsidizing it so heavily they will lose money.  It's like subsidizing old computers and insisting we will own market share forever if we lose money on them now.  In order to make it look like a race, crazy politicians are also subsidizing the purchase of panels they insist we need to produce cheaper than China; so we subsidize both ends of the economic chain. Madness.

    Current solar tech may be as efficient as it will ever get, which means if we instead let China lose money on solar panels they sell to us (we aren't in a race to produce cheap radios, we simply buy them cheap from China also) while we invest in basic research to produce some technology which will be better, we will win in the end.

    There are 297 m^2 of impermeable surface area per person in the USA. Full sunlight is 500 watts / m^2 or 148.5 KW per person.

    Would it be hard to convert some of that surface area to solar panel/paint substrate? In my location they have panels over parking lots and the shade keeps the cars cool. I don't think it's too hard nor is it expensive to maintain unless we drive our cars and trucks on the panels.

    I also found this using the internet search: In this study, relating turf grass area to an estimate of fractional impervious surface area, it was calculated that potentially 163,800 km2 (± 35,850 km2) of land are cultivated with turf grasses in the continental United States, an area three times larger than that of any irrigated crop.



    Here's the information that goes with the figure (found at

    Required land areas

    If the Total Primary Energy Supply (TPES) was to be generated entirely from sunlight, a certain amount of the earth's surface would be needed for capturing.  The size of that area largely depends on the efficiency of the conversion technology.  An example with a conversion efficiency of 8 % is shown in the figure at the top of this page:  Six discs, each large enough to produce an average power output of 3 TW, are distributed across the world in deserts, areas that have plenty of sunlight and little population.  The desert locations, their sizes, the average sunlight intensity, and the required areas are listed in the following table:

    Location / DesertDesert size
    Required area
    W / m2
    Africa, Sahara9,064,960144,231260
    Australia, Great Sandy388,500141,509265
    China, Takla Makan271,950178,571210
    Middle-East, Arabian2,589,910138,889270
    South America, Atacama139,860136,364275
    U.S.A., Great Basin492,100170,455220

    Now, 10 million hectares is equal to only about 100,000 km2.  So the "black dot" on the U.S. is 70% larger than that, and will only generate an average of about 3 TW, at 8% efficiency (or 5.6 TW at 15%).


    Thor Russell
    Thanks for that. I see that each black dot is 3TW, not enough to power the whole world, but just that chosen continent. I am a bit curious about the 8% figure, I am it includes gaps in between the solar panels? 
    Also I am pretty sure that SunTech power is now mass producing panels with 24% efficiency/20% module efficiency. I read an article somewhere about how they had managed to take some advanced tech from an Australian lab and quickly mass produce it. 

    Thor Russell
    The article does not seem to say that.

    8% is because no one can afford to actually roll out the top-end solar panels.  8% is what is commonly used now.   44% is really the maximum available at all so by the time inefficiencies are considered, 16% will be about the best.  But 8% is realistic today.

    Thor Russell
    I don't know where you can buy 8% panels anymore: 

    I can't see how it would be cheaper to use them either. Costs of $1-1.2 per Watt are now quoted for solar panels and installation is now about a similar price per watt to the panels in land based installations. 
    Even if you go with $1.50 per watt for the panels at 15% and $1.50 for installation, then if you have an 8% panel, isn't the equivalent installation cost per watt going to about double? Even if those 8% panels where pretty much free, wouldn't that then make installing the same wattage still cost $3 per watt as you would need twice as much installing/material etc?
    Thor Russell
    Timothy Connet
    Let's have some fun with solar power and figure how large of an area would be needed to power the USA.

    The source of how much electricity the USA uses will be from:
    "The U.S. aluminum industry directly comsumes 45.7 x 10 (9 power) kilowatt hours (0.16 quad) of electricity annually or 1.2 percent of all the electricity consumed by the residential, commercial, and industrial sectors of the U.S. economy."

    (I am guessing this is from around the year 2007.)

    45700000000 * 83 1/3 = 3,808,333,333,333 kilowatt hours

    One square meter of solar panel will produce 100 watts per hour. (1000 watts @ 10% efficiency) or (1300 watts @ 8% efficiency)
    A day will have at least 10 hours of light a day.
    A day per meter of panel will produce one kilowatt.
    A year, being 365 days, will allow a one square meter of panel to produce 365 kilowatts per year.

    3808333333333/365 = 10433789954 meters of solar panels
    This converts to around 4028 square miles or an area of 64 miles by 64 miles.

    If you wish to indulge a fantasy, imagine this vast array floating above the clouds on thin film vacuum balloons at 60,000 feet to avoid land issues and above the clouds to gain the full exposure of the sun.
    One can dream.