The gas is greener

How much steel and land do “environmentally friendly” energy technologies use? Is this “sustainable”? Energy guru Robert Bryce explains how un-green renewable energy is in this… New York Times op-ed.

Here’s the land use excerpt:

Consider California’s new mandate. The state’s peak electricity demand is about 52,000 megawatts. Meeting the one-third target will require (if you oversimplify a bit) about 17,000 megawatts of renewable energy capacity. Let’s assume that California will get half of that capacity from solar and half from wind. Most of its large-scale solar electricity production will presumably come from projects like the $2 billion Ivanpah solar plant, which is now under construction in the Mojave Desert in southern California. When completed, Ivanpah, which aims to provide 370 megawatts of solar generation capacity, will cover 3,600 acres — about five and a half square miles.

The math is simple: to have 8,500 megawatts of solar capacity, California would need at least 23 projects the size of Ivanpah, covering about 129 square miles, an area more than five times as large as Manhattan. While there’s plenty of land in the Mojave, projects as big as Ivanpah raise environmental concerns. In April, the federal Bureau of Land Management ordered a halt to construction on part of the facility out of concern for the desert tortoise, which is protected under the Endangered Species Act.

Wind energy projects require even more land. The Roscoe wind farm in Texas, which has a capacity of 781.5 megawatts, covers about 154 square miles. Again, the math is straightforward: to have 8,500 megawatts of wind generation capacity, California would likely need to set aside an area equivalent to more than 70 Manhattans. Apart from the impact on the environment itself, few if any people could live on the land because of the noise (and the infrasound, which is inaudible to most humans but potentially harmful) produced by the turbines.

Industrial solar and wind projects also require long swaths of land for power lines. Last year, despite opposition from environmental groups, San Diego Gas & Electric started construction on the 117-mile Sunrise Powerlink, which will carry electricity from solar, wind and geothermal projects located in Imperial County, Calif., to customers in and around San Diego. In January, environmental groups filed a federal lawsuit to prevent the $1.9 billion line from cutting through a nearby national forest.

Here’s the steel use excerpt:

Consider the massive quantities of steel required for wind projects. The production and transportation of steel are both expensive and energy-intensive, and installing a single wind turbine requires about 200 tons of it. Many turbines have capacities of 3 or 4 megawatts, so you can assume that each megawatt of wind capacity requires roughly 50 tons of steel. By contrast, a typical natural gas turbine can produce nearly 43 megawatts while weighing only 9 tons. Thus, each megawatt of capacity requires less than a quarter of a ton of steel.

Obviously these are ballpark figures, but however you crunch the numbers, the takeaway is the same: the amount of steel needed to generate a given amount of electricity from a wind turbine is greater by several orders of magnitude.

Such profligate use of resources is the antithesis of the environmental ideal.

20 thoughts on “The gas is greener”

  1. This answers some questions that I’ve had and confirms some suspicions.

    It is ironic to see the “green” people screaming for “renewable energy”—–until someone tries to produce it.

  2. Robert Bryce states that the Ivanpah solar thermal facility will require 3,600 acres for a 370 megawatt plant. That’s 9.73 acres per megawatt, 39,375 square meters per MW, or 39.375 square meters per kilowatt. However, it takes no more than 6 square meters per kilowatt solar photovoltaic (16% or more efficiency), and solar thermal is more efficient than solar photovoltaic. Access space is needed between rows of panels, but much more efficient use of land should be possible with solar power. Using existing rooftops and parking lots for photovoltaics would not require any unused land.

  3. It’s great to take a deeper look at all energy sources as Bryce tries to do, but for an energy guru, he leaves out some startling pieces. If we’re going to fully compare energy sources, then lets also compare land use for producing the gas in the first place as well as machinery, man hours and steel needed to collect and transport that gas to the power plants in the first place. Not to mention that dual land use is fairly easy to do with green energy with two specific examples being rooftop PV solar as well as putting wind turbines on farms.

    Again, I appreciate the deeper look into the topic and a few good points are raised, but equally good and obvious points seem to be left completely out. Until Bryce goes back and takes an equally deep look at BOTH types of sources, then his article is fairly useless.

  4. What are those turbines running on, air?

    This is a classic example of bad journalism that makes scientists like me cringe.

    If you are evaluating the amount of steel needed to generate a MW from a natural gas engine, one must consider the energy requirements to collect the gas as well. Natural gas wells are drilled using steel bits and lined with steel pipes. There is a lot of infrastructure involved in transporting the gas as well – large energy intensive compressor stations, many miles of pipelines, etc.. There are also other environmental issues, such as the use of fracking chemicals when drilling the well, continuous use of anti-corrosives, methanol and other chemical addititives, the large amount of methane (a potent greenhouse gas) that leaks out of the wells during construction and as part of normal operations, the vehicle emissions associated with transporting well fluid to water treatment facilities and it’s processing, and so forth. Additionally, there is a significant amount of land use associated with nat gas collection including the well sites, compressor stations, and road building. Just FYI, if say you wanted to write an article on this or something.

  5. “However, it takes no more than 6 square meters per kilowatt solar photovoltaic (16% or more efficiency), and solar thermal is more efficient than solar photovoltaic.”

    Not in the real world. You are assuming 1000 watts/meter which is only going to happen for a short period of time on any given day near the equator. In fact, the highest solar isolation in the U.S is in the south west. The peak average monthly isolation is 6 kWh per square meter per day. Even if you only consider the summer months the peak is only 8 kWh per square meter per day while the average during that time period is between 5 and 6 kWh per square meter per day.

    With 100% conversion efficiency during the summer you would get at best 36 kWh’s per day from a 6 square meter array. The 16% conversion efficiency only happens in the lab where the silicon substrate is held a constant temperature of 25 degrees C. It is a well known fact, that due to silicon’s negative temperature coefficient, that the efficiency of solar cells drops with increasing temperature. At 1000 Watt’s per square meter the solar cells are going to get awful hot. Having looked at data for dozens of real life operating solar arrays I very much doubt the efficiency at 1000 watt’s per square meter would even make double digits much less 16% efficiency.

  6. takes a stab at estimating the spacial power density of various sources. These include the transportation/extraction costs of gas/coal and biomas.
    Power Source Power Density (W/m2)
    Natural Gas 200 – 2000
    Coal 100 – 1000
    Solar (PV) 4- 9
    Solar (CSP) 4- 10
    Wind 0.5-1.5
    Biomass 0.5- 0.6
    I also believe there are some issues with farm-wind power combinations:
    How much does the tower and access roads disupt the tiling (drainage), tillable area, and (if applicable) irrigation.
    The solar (roof, parking lot) dual use does however have some potential to make a particular installation more attractive (ie roof rack system shades house reducing cooling load) Can it make up the spread in comparison to conventional sources? – No!

  7. Greg, you are talking about kilowatt hours. Bryce’s article talks about megawatts of solar capacity, not megawatt hours, and my reply limited itself to kilowatts, not kilowatt hours. A one kilowatt system, 16% efficiency, requiring six square meters of panels, is very common, look up the specs for solar panels made in the last few years. Sunpower E18 and E19 panels have 18% and 19% efficiency. True, in most locations in the U.S., a one kilowatt system will generate only 4 to 6 kilowatt hours AC electricity per day.

  8. Skeptic, what you fail to realize is the specs from the manufacturers are for the substrate being at standard temperature of 25 degrees C. That will never happen in the real world as the panel substrate will heat up with a corresponding drop in efficiency. Capacity (kilowatts) is useless as the sun doesn’t shine half the day. If you want to know how much the panels really produce you have to look at real life data, not some lab generated specs.

  9. Greg, I “fail to realize”? That’s not true, I do know the STC specs.
    “That will never happen in the real world”? That’s not true, photovoltaic panels can function at their rated capacity in the middle of a sunny day in mid-latitudes, if it is a cool day with air circulating. That’s frequently the case for my photovoltaic installation.
    “Capacity (kilowatts) is useless”? That’s not true, it is a starting point to calculate how much the panels produce in real life. I discussed capacity in kW or MW, as that is what was done in the original article.

  10. Skeptic
    The problem is that the photo voltaic array will only produce the maximum output for two hours near noon. The rest of the time the output is reduced and for at least 12 hours per day the output will be zero. This will not produce a stable power grid. The solar cells must be cleaned daily, slowly lose generating capacity and only last about 20 years. There isn’t any environmentally friendly energy storage system to fill in when the sun doesn’t shine so the whole photo-voltaic generating system is a waste of time and energy. We need to start using the LFTR reactor technology that was developed during the 1960’s and stop wasting money on 14th century windmills and useless photo-voltaic arrays.

  11. I don’t think the panels need to be cleaned daily. Ours are on the roof, are only “cleaned” when it rains, and I haven’t seen any decline in two years. We have a few percent return on investment, amortized over 20-25 years, but that does depend on rebates and how much you pay for electricity from the grid. Solar is obviously not going to be the entire answer, 24/7.

  12. Not here in the frozen North. Minnesota has three feet of snow on the ground for neary half the year. In addition we’ve got way to many overcast days as well.

    Now, down southwest… maybe.

    Still, ‘clean’ coal, natural gas, and nukes can provide more power than we can possibly use and we’ve got a thousand year supply of the stuff. The power density of the land use for conventional is orders of magnitude smaller than all these green flights of fancy.

  13. Springerville Arizona Solar Array sits 6500 feet above sea level, an advantage not afforded most solar installations, with a capacity rating (standard test conditions) of 4.59 MW. Today at 12:30 the solar isolation was 1023 Watts per meter and the array output was 3.962 MW. IOW, the output was about 84% of the rating under standard test conditions. The average performance from 2004 to 2006 was about 80% of the rating under standard test conditions. The standard test conditions is not starting point, it is marketing misrepresentation used to inflate the real life performance.
    To make a comparison with traditional sources of generation the true capacity has to be calculated from real data. In 2006 the Springerville Array generated a total of 7,765,000 kWh. Since there are 8760 hours in a year the average output for the Springerville array was 886.4 kWh per hour. Skeptic, you appear not to understand the relationship between Watts and Watt-hours. If I had a generator that put out 886.4 kW’s it would generate 886.4 kWh’s in one hour. If that generator did that 24 hours for 365 days it would generate the same kWh’s the Springerville plant does in a year. So, if I wanted to replace the Springerville Array with a traditional generator it wouldn’t have to be very big. A 1MW generator, even allowing for 2 weeks down time, would generate 8,424,000 kWh per year or about 8% more than the 4.59 MW solar array. To put it in more succulent terms. The capacity rating of a solar array, in what is arguably one of the most ideal locations, has to be 5 times larger than a traditional generating plant for the same yearly output. And, unlike the solar array, the traditional generator will produce power when it is needed.

  14. Greg, in an earlier comment, I said a 1 kW system would generate about 4-6 kWh per day, and I do understand the difference between Watts and Watt-hours. Agreed, a 1 kW generator should generate close to 24 kWh a day, much more than a 1 kW photovoltaic system.
    Re: your comment that in the middle of a sunny day today, the output was only about 84% of the STC rating: the panels are tilted at 34 degrees (the latitude of Springerville), so that they would be at right angles to the sun at noon on the equinox, almost three months ago. Currently, the sun is about 22 degrees above that, which explains part of the loss. Probably there are other ways the system does not meet STC conditions, to explain the rest of the loss. There are other rating systems which may be more “real world”, but STC is commonly used.
    Nothing surprising about the above facts, that’s the way solar works. Bottom line is the financial and environmental costs. Here’s a summary with some information about the finances and a few links:

  15. Our solar cost in excess of $11,000, I got a gov’t subsidy so I paid about $2,500. I get a feed in tariff 3 times the cost of power I use – 60cents feed in, 20 cents usage per kWHr.

    I was 55 when it was installed – I live in Queensland Australia where we have strong sunshine – we live on the Sunshine Coast.

    I will be lucky if I live long enough to recoup my $2,500 investment – the Gov’t’s $8500 will never be recovered.

    The investment seems a poor one to me – they’re Sharp panels and they only guarantee them for 10 years – I know they’ll probably last more than that but they need to last for 50 or more to be viable economically.

    I think the popular green terminology is for these having no running costs but I do not see them solving our energy needs.

  16. I think this is an irresponsible article in the way that the data is presented and his conclusions made. I did some sums myself, based on and came up with this:

    Total land dedicated to agriculture in California: 25.3m acres (out of total 100m acres, i.e. 25%) = 39,000 square miles, equivalent to 1625 Manhattans

    Assuming his assumptions and calculations are correct:

    8500MW solar requires 129 square miles
    8500MW wind requires 1,675 square miles
    Total extra land – 1,804 square miles = 1.1m acres = 1.1% of total land = 75 Manhattans

    So, this total may seem relatively high, but important to note is that land with turbines on, while they may not be all livable, they are certainly all usable. North Germany is covered in wind farms and cow farms, in the same fields. When looking for a suitable image to share, I came across this National Geographic article with a similar photo from Canada

    So, really, the extra land required equates only to 129 square miles for solar farms (which doesn’t take into account solar energy generated on roof tops as others have mentioned) = 0.0012% of land.

    This is a small price to pay for energy security, jobs and money (and that’s not even mentioning climate change)

    And good, up-front planning can ensure that ecologically sensitive areas are avoided for infrastructure, including power lines. If all the renewable energy industry followed this model, they would find that environmental organizations would support rather be against renewable energy projects. We need renewable energy – but in the “right” places, from technical, economic, social and environmental perspective. Superficial, distorted articles like this one in mainstream media do not help us to get to this negotiated balance.

  17. Think’s reference to the vast amount of steel used in drilling a well is misleading. The largest amount of steel used is the drill stem and bits which are recovered and used over and over. They are not left in the well as he implies. The only steel left is the casing which constitutes a tiny fraction of the steel used in turbine and produces 24/7, wind or no wind. Wells require pipelins, wind turbines require power lines and towers.

    Also solar does not produce 24/7. It produces nothing at night or when covered with snow or frost. Also greatly reduced output when it is cloudy. It is a viable supplement in the sunny SW, but not a replacement.

  18. You can speak sense to a Green neo-Druid. But not much, and not for long before he starts spouting neo-Druidic dogma. Most thave no technical background at all, aside from a few buzzwords as Mr. Skeptic has absorbed. He can’t even distinguish between a KW and a KW-h .

  19. Well, I guess what is really needed are nuclear reactors and reprocessing of the fuel rods. Nirvana!

Comments are closed.