Running the Electric Grid with eSolar

As I often do on a Saturday morning, I was up early reading through energy headlines. I happened across this story on eSolar:

Bill Gross’s Solar Breakthrough

“We are producing the lowest cost solar electrons in the history of the world,” Bill Gross is telling me. “Nobody’s ever done it. Nobody’s close.”

“We have a cost-effective, no-subsidy solar power solution and it’s for sale, anywhere around the world,” he says.

The article was intriguing, and inevitably led me back to eSolar’s website to get a better idea of whether the claims appear to have merit. There, I watched the slide show on the technology, and caught this bit: A single unit generates 46 MW of clean electricity on a footprint of 160 acres.

While this doesn’t help me figure out whether they can deliver on the hype, it does enable me to update a couple of essays that I have written before:

A Solar Thought Experiment

Replacing Gasoline with Solar Power

In the first, I made an attempt to calculate the area that would be required to equal the entire installed electric capacity of the U.S. – using only solar power. (Yes, I understand that this number falls to zero at night). The numbers quoted above from eSolar – combined with the latest data on installed electrical generating capacity – enabled me to update that calculation.

Per the EIA, total installed electrical generating capacity in the U.S. is approximately 1 million megawatts. If we scale up eSolar’s claim of a required footprint of 160 acres to produce 46 MW of electricity, then it would require 5,435 square miles of eSolar technology to equal current U.S. electrical capacity. This is a square of 73.7 miles by 73.7 miles. This is greater than the 2,531 square miles calculated in the previous essay, but that essay only considered the area for solar panels. The present calculation encompasses the footprint of the plant.

Looking back at the gasoline calculation, I came up with 1,300 square miles required in my previous essay to replace the energy gasoline provides. Using the current eSolar numbers changes that number to 2,413 square miles, or a square of 49 miles on each side.

Of course all of the normal caveats apply as spelled out in the previous essays. The key point is not to read these sorts of thought experiments too literally. I tend to do them to get my head around the scale of certain problems. Complaints of “the cost is too great” or “the power is intermittent” – addressed by caveats in the previous essays – completely miss the point of the essay. It is sort of like trying to figure out how much biomass would be required to power the world. If the calculation is 10 times the current annual output of biomass, then that’s not going to work. If it is 1/100th the current annual output of biomass, then that might work (again, pending lots of other things working out).

In this case, I find this eSolar thought experiment encouraging insofar as the required land area isn’t a clear knockout.

29 thoughts on “Running the Electric Grid with eSolar”

  1. I think I would tend to look at it as "how many sq miles to provide "Peak Power" to Cities, and towns, below a certain Lattitude.

  2. It seems like about 1/3 of my, yearly, electric bill in the midsouth is from air conditioning during the summer months (admittedly, the Midsouth might be a little too far North for CSP.)

    However, let's take the lower 1,500 counties (about half of 3,000 U.S. counties.) Figure you want to produce 300,000 megawatts. What's that? 200 megawatts/co?

    About Two sq miles. Average county is about 1,000 sq miles. It should be pretty easy to find Two Sq Miles of worthless dirt in an area of 1,000 sq miles.

    I think land use isn't much of an issue.

  3. Bill Gross is an interesting guy, but remember, he is trying to raise money, and he has had a few bombs–he was behind eToys, the cratered website. Something happens to people in VC–they sometimes think first 1) Can we raise the money, then 2) Will it work?
    Energy is hot now, not VC guys are funding energy projects.
    I still wonder if solar eats up too much land. Why is solar better than the very small footprint of a mini-nuke plant?
    But, I concur with RR's sentiments: We have multiple energy options, and they are not of the question. Indeed, we have many practical options for generating electricity.
    Some may be a bit more expensive than other methods, but nothing to become a doomster about. After all, architects are designing buildings to use far less power than before. Even if electricity rates should double, most people and businesses, if they tried, could cut power consumption by half.
    The ultra-fascinating play in the future is the PHEV. If a viable PHEV can be made, then we are home free. We know we can generate power, from conventional plants, from nukes, even from wind and solar.
    For the next several decades, we have enough fossil fuels, certainly in the United States, with our epic supplies of natural gas (God does indeed look after drunks, fools and the United States).
    But if the PHEV can be made to work (not just technically, but commercially) we are truly home free.
    Global demand for liquid fuel would drop every year, not rise, even as biofuels and CTL and GTL increase every year.
    So, let's watch this eSolar plant, but remember, we enjoy a plethora of electricity generating options. The liquid fuel knot may be cut open by the PHEV. Time will tell.

  4. From the Article: "The Sierra SunTower alone uses 24,000 mirrors, made by a contract manufacturer in China."

    Great! So the big breakthrough in Concentrating Solar Power is the use of slave labor in China.

    If the US is going to try to reduce its dependence on imported oil, the last thing it should do is replace it with a dependence on imported mirrors from a somewhat unpredictable regime which already has the US over a balance of trade barrel.

    Sandia National Labs built one of these solar towers in New Mexico probably about 20 years ago. You can still see it when flying out of Albuquerque. Nothing new here — except for the slave labor aspect.

    And, at the risk of seeming irritable, the solar tower still does not address the primary issue of what to do when the sun goes down.

    Final nail in the coffin — Google is investing in it. Kiss of death!

  5. You DO seem a little irritable, Kinauchdrach. (think maybe a couple of prunes would help?)

    🙂

    However, I'm having a hard time finding a point for disagreement.

  6. I think the PHEVs will work, Benny. I just wonder, in the long run, if they'll even be necessary.

    There are small cars on the road, today, that will get 30,000 miles from an acre of oil palm, and the cars that can get 28,000 miles from an acre of corn (after accounting for the DDGS) are just around the corner.

    I don't worry much about the Long term. I do fret about the short term. It looks to me like, starting around 2011, or 2012, we might have a few tough years, until we can get our feet under us, and get "up and running" on petroleum alternatives.

  7. I keep wondering how hard it would be to put power towers or Stirling dishes or the like over parking lots and shopping malls.  Don't those areas need shade?  Why isn't anyone proposing to get a two-fer?

    I can see if the scattered light off the target itself is a hazard to people nearby, but I've not read anything about it and a quick search finds nothing.

  8. The problem with CSP is it does not work for making electricity. I have read all the studies. Operators of steam plants like to brag about performance. Nobody is bragging. Of course the absence of evidence, is not evidence.

    The first issue is environmental impact. The environmental impact of the steam plant is going to be at best the same as a nuke plant per MWe.

    The second issue is the engineering challenges. Engineering wise steam plants are subject to thermal fatigue. Storage is needed to keep the plant operating 24/7. The next issue is scalability. A 1600 MWe nuke plant has one reactor, one turbine, 4 steam generators, 4 reactor coolant pumps, 3 feed pumps and so forth. This CSP has one turbine per 46 MWe and so forth. The limitation may be do to the amount of heat that is lost over the distance between the collector boiler and turbine.

    These CSP do make interesting background for movies.

  9. "I keep wondering how hard it would be to put power towers or Stirling dishes or the like over parking lots and shopping malls."

    Concentrated energy always needs to be treated with a lot of respect. How many people want to go shopping underneath giant tanks of superheated water, with big steam turbines spinning at high speed?

    Then there are the other real-world factors. How do you stop birds nesting in the structures & crapping on the mirrors? Or bats? Industrial facilities sometimes have to do things that the general public would rather not know about.

    And the complexity. Each of those 24,000 Chinese mirrors requires an independent two axis tracking system, with motors & sensors. All exposed to ambient conditions. Lots of maintenance work there, which could interfer with shoppers or car-park patrons below.

    All this for a few MegaWatts? Sounds like a very poor substitute for a well-sited compact nuclear power plant.

    Excuse me — I need to run out and get another gallon of prune juice.

  10. I posted This Link Wednesday, on the "Rate Crimes" Thread.

    It's the eSolar, Sierra "Power Tower." It IS operating, and

    the PG&E guy says it's the "Cheapest" power they can buy during the time that it's making electricity. I don't know if that "price" is "subsidized," or not.

  11. I agree, K, that it's probably better to leave it on the ground. A couple of acres off to the side would be available in many cases. As for "rooftops:" I think I'd stick with PV.

    I like the Big Nukes for baseload; but, I can see nothing but benefits to having some solar for "peaking." (If it "works" at a reasonable cost, of course.)

  12. Don't get me wrong, I wouldn't tear down a reasonably clean Coal Plant. The corn, and soybeans can always use the extra CO2.

  13. “It IS operating, and “

    So Rufus, how much electrician is being produced? If you buy a new car and it stops running on the way home from dealer, IS it OPERATING?

    You did not read the new story very critically. Not that I would expect a journalist to provide the kind of information that I would be looking for.

    As far as statements by California utilities. They have contracted to buy renewable energy. If the suppler spends a ton of VC money and does not deliver, the utilities have not lost anything.

    California consumers are better off if CSP does not make much electricity. Then they do not have to pay for it. The converse is true for nuclear. If nuke plants are off line, it cost millions a day to buy replacement power.

  14. Kit, I'll take your word for it. I, certainly, don't know what ANY of the facts really are.

    I do have one question, though. Those Nukes really can't be ramped up, and down on a 24 hr cycle to handle the "Peak" hours, can they?

    And, if not, wouldn't it be good if you could get solar to handle the "Extra" load during that time?

  15. I still think the future belongs to deep geothermal. Once big oil gets the kinks out,they'll be off to the races. It's baseload power,clean,and scalable.

  16. "Concentrated energy always needs to be treated with a lot of respect. How many people want to go shopping underneath giant tanks of superheated water, with big steam turbines spinning at high speed?"

    All of the big hardware can be in its own building in the corner, like the stormwater retention pond.  Only the mirror field would be over the parking and on top of the occupied buildings, shadowing the area below and keeping it cooler.  Who doesn't want to park in the shade?

  17. Water use with a CSP is a big thing – in the range of 2.5 to 3.0 liters mer kW.

    In this case everyone is spreading a lot of BS about recycled water etc and it will be good and water is recharged to the reservoir after use. The water is lost through evaporation and blowdown.

    Actual operating data for a 5 mW PV unit is available at http://www.tep.com/Green/GreenWatts/solaroutput.asp
    As far as hours of output etc it is a 'little bit' comparable.

    They could use air cooling – hybrid air/water cooling or a Heller cooling tower but those are all additional capital cost.

    One of the next gen bigger CSP plants in utilizes approx 6 hours of thermal storage to extend the daily operating hours.

  18. Maury said…

    "I still think the future belongs to deep geothermal. Once big oil gets the kinks out,they'll be off to the races. It's baseload power,clean,and scalable."

    ———————————–
    Maury, If you like geothermal. here's something:

    "Three Companies Join to Build 200 50-MW Geothermal Power Plants in Indonesia."

    Arrow Resources Development (OTC:ARWD), Four Crystal Funding (OTC:FCRS), and GTherm have partnered to develop and build 200 50-megawatt geothermal power plants in Indonesia………….

    http://www.energyboom.com/geothermal/three-companies-join-build-200-50-mw-geothermal-power-plants-indonesia

    John

  19. Kit P wrote: The problem with CSP is it does not work for making electricity. I have read all the studies. Operators of steam plants like to brag about performance. Nobody is bragging. Of course the absence of evidence, is not evidence.

    Kit P's inability to find evidence is not the absence of evidence.

    SEGS Solar Energy Generating Systems performance 1986-2002 on page 9 of
    http://www.ornl.gov/sci/engineering_science_technology/world/renewable/Trough%20Technology%20-%20Algeria2.pdf

    Nevada Solar One Production June 2007 through April 2008 on pages 24 thru 28 of
    http://www.pole-derbi.com/photo_derbi/2%20Gilbert%20COHEN%20_%20DERBI%20Perpignan%202008.pdf

    Where can I find this detail of information for an individual nuclear reactor, such as Oyster Creek 1?

  20. Rufus, nuke plants load follow just fine. That is what the French do. In the US we use SCGT.

    E-P, my company used to build CSP to run HVAC. We installed one on a parking lot cover in California.

    Clee asked, “Where can I find this detail of information for an individual nuclear reactor, such as Oyster Creek 1?” Steam turbine blading has huge kinetic energy. If one lets go, be someplace elese.

    NEI provides an informative market report that includes daily power information about all nuke plants in the US.

    http://www.nei.org/resourcesandstats/documentlibrary/publications/energymarketsreport/energymarketsreport/

    BTW, Oyster Creek was built in 1969 and it is starting it 40th year of operation.

    Thanks for the solar links Clee but if I say something does not work you avoid providing links that prove my point. Clee's links provided no information on reliability.

    What we can conclude is that CSP could work if we ignore all those pesky engineering details like reliability. CSP is an exercise in wishful thinking.

  21. re: Benny Cole
    I still wonder if solar eats up too much land.

    5,435 square miles of eSolar.

    Roughly 43,000 square miles of arable farmland were devoted exclusively to US corn ethanol production in 2007. [Source]

    And we got 335,000 square miles of "Warm Desert" in North America. [Source]

    How long is it gonna take until your drop that silly land area argument?

  22. Kit P – What are the reliability concerns you have? Those 'pesky little engineering details'?

    Steam turbines are steam turbines

    Steam piping is steam piping

    The tracking systems should not be a big deal.

    The control systems overall are old hat.

    Water consumption – until they start being forced to use a low consumption cooling method (which are available and proven) – is a concern.

  23. Russ, every change in solar radiation induces a thermal cycle. There might be a little more to a steam that you know about.

    Why do you think CSP is worth the effort?

  24. I saw a show about lifting E-Solar's collectors to the top of the tower. The engineering standards reminded me of one of Bill Gross's dotcom startups, which was a bit scary.

    Rough estimates tell me E-Solar's design is not efficient. I think the linear fresnel guys (e.g. Ausra) will beat them. But it's good to have lots of competing designs. Reliability will come with practice, it's not like our first 20 GW of nukes delivered 99% uptime.

    CSP-PV will beat CSP-T on cost. You replace the entire liquid loop (collector, pipes, turbines, etc.) with a small area of triple-junction PV. Not only is it cheaper and more reliable, but efficiencies can hit 40% vs. 15-30%, giving you more kWhs from your huge investment in trackable mirrors and such. But CPS-T has an ace in the hole — storage. We'll see how it plays out.

  25. Kit P, thanks for the link. Very nice daily availability report. It appears I've confused "production" with "performance" when there's more to performance than just production. Having production every year after start up does at least mean each CSP plant was working during the year. They don't operate at night, but then they aren't meant to be base load plants, they're meant to provide power during peak demand hours.

    "Steam turbine blading has huge kinetic energy. If one lets go, be someplace elese. "

    I confess I am surprised that a blade would break off of a 2-year-old turbine taking Cook Nuclear Plant Unit 1's 1-GW capacity out of service for a year, possibly longer. I'm also surprised that Cook 1 was not running from Sept 1997 to Dec 2000 either.

  26. 46 MW/160 acres = 71 watts/meters squared = 7.1% solar efficiency. Is this correct? That is far higher than the 3% efficiency that most thermal solar company facilities achieve. How is eSolar so high?

    One of my friends with eSolar has told me that eSolar's cost per watt including the balance of system but excluding the cost of land = less than $1/watt. This is amazing. Is this what the rest of you are hearing as well?

Comments are closed.