Become a fan of Slashdot on Facebook


Forgot your password?
Google Power Technology

Google Invests In World's Largest Solar Power Tower Plant 387

cylonlover writes "Google has chipped in a US$168 million investment in what will be the world's largest solar power tower plant. To be located on 3,600 acres of land in the Mojave Desert in southeastern California, the Ivanpah Solar Electric Generating System (ISEGS) will boast 173,000 heliostats that will concentrate the sun's rays onto a solar tower standing approximately 450 feet (137 m) tall. The plant commenced construction in October 2010 and is expected to generate 392 MW of solar energy following its projected completion in 2013."
This discussion has been archived. No new comments can be posted.

Google Invests In World's Largest Solar Power Tower Plant

Comments Filter:
  • Drop in the bucket (Score:5, Informative)

    by elrous0 ( 869638 ) * on Wednesday April 13, 2011 @03:33PM (#35811254)

    $168 million sounds like a serious investment, until you consider that this thing is projected to cost $1.37 *billion*.

    • by w_dragon ( 1802458 ) on Wednesday April 13, 2011 @03:40PM (#35811340)
      Over 10% is hardly a drop in the bucket.
      • by elrous0 ( 869638 ) *

        I think the summary should have included the total cost. One could read it and come away thinking that Google was completely bankrolling the project, when this is actually just a fraction of the money that will be required to build it.

    • If you were buying a $600,000 house, could you justify buying a $70,000 car because it's a drop in the bucket? Google hardly has a majority contribution (plurality maybe? Haven't read the article...), but about 12% is a non-trivial contribution in my book.
    • by kevinNCSU ( 1531307 ) on Wednesday April 13, 2011 @03:47PM (#35811420)

      $168 million sounds like a serious investment, until you consider that this thing is projected to cost $1.37 *billion*.

      You a Chemist? I don't know what the hell kinds of buckets you use but mine tend to carry more than 9 drops ;)

      168 mil / 1.37 billion = a little more than 12%. I'd consider 12% of my salary or budget a pretty significant investment, and if I was taking a test I'd consider a question worth 12% of the grade worth a pretty significant investment in doing well on it.

      • by elrous0 ( 869638 ) *

        I think the summary unfairly implied that this was a majority investment. 12% is, to my mind, a very small investment. But, drop in the bucket or not, it's still nowhere even close to the kind of investment that will be required to actually see it through. I think that needs to be clarified.

    • by elrous0 ( 869638 ) *

      Actually, I wanted to correct this. $1.37 billion was just the loan guarantees it was given for construction. After researching it a little more, I found out that this doesn't cover all the actual cost. Actual construction cost is estimated to be more like $2.14 billion.

    • That's the amount of the federal loan the company got. Add to that Google's $168m, and add other investments, but they won't say what the projected actual cost is. And the effective generation rate of the ISEGS is about 15%, which takes into account darkness, cloudy days, etc. They say the output is 392MW, but you need to discount that to get the effective delivered capacity of 60MW. So if the cost is $1.5B then the cost per kW is about $25,000, which is way high. Nuclear plants are up to about $10,0
      • by thermopile ( 571680 ) on Wednesday April 13, 2011 @04:46PM (#35812070) Homepage
        Wait, wait, wait. I was an engineer closely involved in a review of this project, and the BrightSource engineers were vehement in their protests against this kind of argument. I feel compelled to share their thinking.

        This is NOT intended to be a 24/7 power supply. It is only a "surge" power supply, intended to produce (and sell) power when power is most needed: during the afternoon hours when things get really hot in LA and everyone starts cranking their A/C units. In fact, the heliostats are arranged to favor the afternoon sun -- if you look at the pictures, you'll see the heliostat is not a perfect circle. There are more mirrors on the east side of the tower, so that when the sun is in the west, more light gets reflected back onto the tower.

        They openly admit they couldn't compete if they were trying to be a 24/7 power supplier. And that's not the point. They don't have energy storage (molten salt, etc.) to be able to keep producing heat at night -- that would be additional infrastructure to support selling power when there's a lower profit margin. They can sell power at a higher price when power is most needed, in the afternoons.

        Others on the internet have accused this plant of being a "natural gas plant" in disguise, which is laughably wrong. The natural gas boiler is *tiny* and serves only to warm up the boilers faster in the morning hours.

        • A large part of the power demand in southern California is for air conditioning, so a power system that produces its power in the daytime works just fine for most of the demand. (Also, the local climate tends to be hot days but much cooler at night, unlike say the humid Southeast where it stays hot at night.)

    • by tmosley ( 996283 )
      Wait, the total cost is then about $3.50 per watt? Christ, they should have just subsidized cheap panel purchases. If this kind of scale up can't create more economy than that, then this approach is not going to work.

      The money would have been better spent on outfitting solar power manufacturers with solar cells, so people stop complaining about how solar panels are really fossil fuels. Some people just don't understand the difference between marginal costs (like fuel that is dug out of the ground), an
      • by 0123456 ( 636235 )

        This difference means that, given a sufficiently long lived solar installation, the initial cost doesn't matter. You will make your money back eventually, and once you do, you have a free source of power.

        The only problem is that most solar installations aren't 'sufficiently long lived'.

        • by Jeremi ( 14640 )

          The only problem is that most solar installations aren't 'sufficiently long lived'.

          The average energy payback period for a solar installation is 1-4 years. The expected lifetime of a solar installation is 30 years. Source [].

    • by Thelasko ( 1196535 ) on Wednesday April 13, 2011 @04:35PM (#35811936) Journal
      392 MW sounds like a lot, until you consider that's only ~8% of Fukashima. []
      • by Jeremi ( 14640 )

        392 MW sounds like a lot, until you consider that's only ~8% of Fukashima.

        It's actually a lot more than the output of Fukashima these days.

      • So 12.5 of these things would take about 182 square kilometers and produce the same power as Fukashima. Of course, Fukashima now takes about 1260 square kilometers and growing, so the solar farm seems pretty good against your choice of nuclear plant to compare...

  • by mangu ( 126918 ) on Wednesday April 13, 2011 @03:33PM (#35811260)

    That's 14.57 square kilometers, the size of a small to medium-sized town, maybe 20000 to 50000 inhabitants.

    • by h4rr4r ( 612664 )

      So? Lots more desert than that. Also in the USA a town with 20k folks is probably more like 20sq km.

    • by rahvin112 ( 446269 ) on Wednesday April 13, 2011 @04:00PM (#35811560)

      Do you have any idea how big the Mohave is? You could fit several European countries in it. It's not even the largest, just the one with (IIRC) the lowest rainfall and cloud cover with bonus points for being the closest to the major CA population centers.

      We have about 6 deserts in the US that could fit dozens of facilities this size with a minimal wildlife impact (they spread the concentric circles of mirrors out by about triple the mirror size). In fact I wouldn't be surprised if we could build mirror farms like this in rural deserts and end up with an area the size of France covered in mirrors. People really fail to grasp just how big the American southwest is.

    • by tmosley ( 996283 )
      They are only producing 10 kW per acre. That's pretty shitty, in my book.
  • is expected to generate 392 MW of solar power


  • I'm sure there are numbers, but from a completely un-informed standpoint it seems to me that the paraboloc trough designs where a slurry tube runs through a mirrored trough would be cheaper to produce and maintain? []
    • At a guess,
      1. it's cheaper to build a flat mirror than a curved one.
      2. A flat mirror can track the sun in two dimensions
      3. greater chance of a coolant leak with the miles of pipe in the parabolic design
      4. greater concentration of heat means greater efficiency
  • That seems a bit low.

    3,600 acres = 14,568,683 m^2

    ~1,000 w / m^2 incident solar energy in the Mojave

    That would give an total solar energy input of 14,568 MW of power to this installation.

    392MW / 14,568 MW = 2.7%

  • by Dutchmaan ( 442553 ) on Wednesday April 13, 2011 @03:59PM (#35811550) Homepage

    Paranoids to the right...

    Fanboy's to the left...


  • How come CERN seems to have money coming out their asses, to bang Large Hadrons together? Now, if they could just bang two Hydrogen atoms together, producing a butt-load of heat . . . now then we're talking!

  • Is a solar plant cost effective yet against traditionally fueled plants?

    A new 2GW nuclear plant costs [] around $10B, this plant is $1.7B for 400MW. Since it's solar, divide rated power by around 3 to account for nighttime, so it's more like a 133MW plant.

    So, cost per watt of nuclear is $10B / 2GW = $5/watt versus 1.7B / 133MW = $12.75/watt

    Can they sell electricity at a high enough price to recoup their costs? Are operating costs for a solar plant like this much lower than for a nuclear plant? The sheer size o

    • by Mr Bubble ( 14652 ) on Wednesday April 13, 2011 @04:28PM (#35811872)

      You are not factoring in the money it cost to mine the uranium, transport the uranium, store the nuclear waste and decommission the facility. Not to mention the costs of all the Fukushimas yet to come.

  • The whole plant (3 units) is expected to generate about 1.2GW at peak. That's about one modern nuclear unit.

    Over a full day, a solar plant generates maybe 1/3 of its peak power. That's OK, though. For areas where air conditioning is the peak load, a solar plant produces max power just when it's needed. A reasonable near-term goal would be to get Southern California's entire air conditioning load (10 to 15 GW) onto solar power.

    This is solar's big advantage over wind power. Wind power is highly variable,

  • by theJML ( 911853 )

    Anyone know what the Return on Investment is for this? I mean, Beside the intangible "We're saving the earth" publicity... Sure they can sell some of it back to power companies, and perhaps gain some carbon credits... but I'm sure they'll also use it as power for a server farm. I have to believe there's some amount of time this pays for itself with any of those options, but the article is a bit light on those details.

    • Well, it depends on the ratio of the growth of costs of other electricity generation methods. The cost of generation for this installation is fixed over the lifetime of the plant, so its long-term cost is basically the cost of return on a conservative bond. Cost of a coal/oil/coal/nuclear powerplant depends on the future cost of fuel, the future cost of environmental remediation/wast disposal. So if oil keep going up in cost, this plant just gets cheaper and cheaper (relatively), but if it drops back dow

  • by gestalt_n_pepper ( 991155 ) on Wednesday April 13, 2011 @05:40PM (#35812608)

    Liquid hydrocarbon depletion will outrun all our attempts to replace the 160 exajoules that oil adds to the world's energy supply each year.

    At least they're trying though. That's more than I can say for the USA federal government.

"If it's not loud, it doesn't work!" -- Blank Reg, from "Max Headroom"