Kevin Drum re-posts a chart on wind power made by Stuart Staniford showing that the number of new wind power plants installed in 2010 was way lower than in 2009 or 2008:

i-28ccfbaa3f71c611e0781eb099524a88-sm_windfarms.png

This is meant as a starting point for discussion about the big economic issues that might’ve caused this. One of the many, many reasons I’ll never make it as a political pundit, though, is that when I see a graph like this, I’m inexorably drawn to speculating about aspects of it that really have nothing to do with the intended point. In this particular case, I look at this graphic and ask myself “Why are there so many wind power plants installed in the fourth quarter every year?”

Because, I mean, look at it. For 2010, the green bar representing the fourth quarter is bigger than the other three quarters put together. 2008 and 2009 come close as well, and in all the years where the installations are broken out by quarter, the fourth quarter has the largest bar of any of them. So, why would wind power installation peak in the fourth quarter?

I can come up with a bunch of reasons, but there are two that seem most plausible to me, one demand-related, and the other accounting-related. The fourth quarter is the start of winter in the Northern Hemisphere, and that’s a time when energy use goes way up– more people need to heat their homes, days get short so lights are on for longer hours, etc.If you were running a utility, and adding power plants, you would definitely want them to come on-line in the fourth quarter of the year, so as to meet that increase in demand, so it could be that projects that are mostly constructed in the summer months only get finished and hooked up at the end of the year, because that’s when it makes the most sense.

The other possibility is less interesting, really, which is that it’s just an accounting game. That is, there’s some immediate advantage to be gained in terms of taxes or whatever by having a wind plant come on in December of one year rather than January of the next, so they make a push to finish before the end of the year, skewing the numbers a bit.

If accounting were the whole story, you would expect to see the number of plants coming on-line in the first quarter depressed relative to the others (because those that would naturally have ended up there are hurried along to get them into the previous fiscal year). You might see a bit of that in the data from 2006, 2007, and 2010. Not so much in 2008 and 2009, though– the first-quarter bar for 2008 is about the same size as the other two, and the first-quarter bar for 2009 is the second-largest for the year.

I suspect it’s a combination of those two, plus a couple of other things (in much of the country, you’re unlikely to break ground on new construction in November, which will reduce the number of projects finished however many months it takes to build one of these later, which might account for the low summer numbers, and that sort of thing). I would guess that some of the abnormally large first quarters in 2008 and 2009 were projects that had been intended to finish in the fourth quarter of the previous year, but ran over a few months. The big run-up in 2007-2009 probably reflects oil prices going high enough that wind power started to look like a good idea, and the crash in 2010 is probably a delayed effect from the financial crash in 2008– projects that were started with bubble money have finished, and post-bubble, there was less money for new projects that would’ve finished in 2010.

An interesting question is whether the huge fourth quarter of 2010 is just an accounting effect, or a “bump” in projects started with stimulus money. If the latter, the 2011 total would probably be higher overall, probably with a larger first quarter (catching the projects that were aimed for late 2010 and missed). If the former, I’d expect a really small first quarter of 2011, with the overall pattern looking more like 2006 or 2007.

Of course, I’ll totally forget this long before the relevant data become available to check this. But it’s kind of fun to speculate about this sort of thing, just to see what you can come up with from a sketchy set of data.

(I need to come up with a good way to turn this sort of thing into a classroom exercise, because this is exactly the kind of mental process scientists and engineers should be doing as a matter of course…)

Comments

  1. #1 OgreMkV
    March 22, 2011

    You must be from the north. I live in Texas and it would make much more sense for wind stations to come on-line in Spring and Summer rather than winter for us. Most of the south doesn’t use electric heating, we use gas (that’s a whole ‘nother story though).

    What we need is power in the summer, when it’s over 100F for 45 straight days with no rainfall.

    So, my guess would be the accounting thing.

    I’ll also add that the dip in 2010 was because the subsidy for wind farms was about to go away until Obama got it renewed. IIRC, the new construction for wind farms in the US in the first 2 months of 2011 is already more than all of 2010.

  2. #2 Lyle
    March 22, 2011

    In Texas the problem is that until the transmission lines get built if you install a wind farm, you likely get your output restricted due to a lack of transmission capacity. If one looks at the ercot site, one finds that the next 2 years in TX will not have much wind installed until the CREZ transmission projects are put thru. Of course the NIBY folks who hate lines are in full cry. If you go to the committes links on the ercot site and look at the board of directors committe and the march 22 meeting, grid operations report you see on page 5 the wind curtailments (more wind power than ways to get it to the customer). Other pages show 36 gw of wind under study but only about 1 gw likley to go on line the next 4 years. No point in putting up a turbine if you can’t get the energy to the customer.

  3. #3 Tom
    March 22, 2011

    I’m guessing taxes. I just read an article that pointed out how much solar got installed in the last quarter of the year in Germany the past few years, because the feed-in tariffs have gone down each year. So if your tax credit is expiring and especially if your country has no long-term commitment and they drag their feet with short-term renewal of incentives (like the US), you probably make sure you’re done while the tax credit is still there.

  4. #4 Eric Lund
    March 22, 2011

    We can definitely rule out seasonal demand as an explanation, and the feasibility of starting construction in November is not an issue in most non-mountainous places outside the Northeast, upper Midwest, and Alaska. As Ogre @1 says, there is significant summer demand for air conditioning–not just in Texas, but over a majority of the country, including places like DC and Chicagoland. Most places with significant demand for winter heating tend to use fuel rather than electricity for heat as the latter is generally quite expensive (the exception is the Pacific Northwest, with its abundant hydroelectric power). The biggest spike in lighting demand would be at Christmas time, and even that is probably not as large as air conditioning demand. So I’d go with accounting as the explanation.

  5. #5 formula 21
    March 22, 2011

    In Texas the problem is that until the transmission lines get built if you install a wind farm, you likely get your output restricted due to a lack of transmission capacity. If one looks at the ercot site, one finds that the next 2 years in TX will not have much wind installed until the CREZ transmission projects are put thru. Of course the NIBY folks who hate lines are in full cry. If you go to the committes links on the ercot site and look at the board of directors committe and the march 22 meeting, grid operations report you see on page 5 the wind curtailments (more wind power than ways to get it to the customer). Other pages show 36 gw of wind under study but only about 1 gw likley to go on line the next 4 years. No point in putting up a turbine if you can’t get the energy to the customer.

  6. #6 Juice
    March 22, 2011

    Speaking of the economics of wind power:

    http://www.wind-works.org/articles/EnergyBalanceofWindTurbines.html

    According to the German study by Gerd Hagedorn, wind turbines produce 4 to 33 times more energy during their expected 20-year lifetimes than that used in their construction. Coal plants produce 64 times more energy and nuclear 108 times more than that used in their construction, notably because they consume a fuel. Current photovoltaic technology produces 1-3 times the energy represented by their materials. When fuel is included, coal and nuclear plants deliver only one-third of the total energy used in their construction and in their fuel supply because fuel consumption dwarfs the amount of energy in the plant’s materials.

    Wind power is basically the least efficient of all electrical grid power sources.

    This came from a pro-wind website, so of course they will add a bit of obfuscation about the fuel. Radioactive nuclides are decaying at all times whether we are taking advantage of the release of energy or not, so it’s totally irrelevant when it comes to the energy balance of the power plant.

  7. #7 Stephen
    March 22, 2011

    @Juice:

    It’s very relevant about whether we use the nuclear fuel or not – Uranium has a tremendously long half-life, the natural decay heat in it is very small.

    In a reactor we moderate and control a nuclear reaction in it which consumes the fuel much (many thousands of times) faster than if it were simply left in the ground.

    Also, from the quote you give, PV gives the least return on energy invested (discounting fuel) and coal the least return on energy invested (including fuel). In neither case is wind the worst.

  8. #8 Juice
    March 23, 2011

    No, Stephen. We’re talking about overall energy balance. What’s the balance of power produced versus the power needed to construct and operate the power plant. For the nuke plant it would be the power required to build the plant and then mine and refine the fuel. The quote conflates the total energy efficiency of fuel use with the energy required to extract and refine the fuel. There’s no way that it takes 3 times as much power than produced to extract and refine any kind of fuel for a power plant. That doesn’t even make sense.

  9. #9 CCPhysicist
    March 25, 2011

    I also argue that taxes explain what you see. It makes good economic sense to spend money as close to when you will get a benefit from it. As charities know quite well, this is often right at the end of the tax year, closest to when you will get a refund as a result of that spending.

    In fact, wasn’t there an extension of special tax credits into the first part of 2009 from one of the stimulus bills? That might explain that one anomalous Q1.

The site is currently under maintenance and will be back shortly. New comments have been disabled during this time, please check back soon.