5 ms·

The article claims a high cost of mirrors and steam turbines as the big negative. Without seeing the cost-breakdown of the entire project, it is hard to ascerta

by chime 13y ago

The article claims a high cost of mirrors and steam turbines as the big negative. Without seeing the cost-breakdown of the entire project, it is hard to ascertain if materials are indeed the largest factor or if it is something else like labor/salary or technology/licensing. If it costs $200m to build the tower and buy turbines, $2B/200,000 mirrors = $10,000/mirror. I find it hard to believe that cannot be lowered drastically once the research phase is over. While I feel Solar Thermal (more appropriately CSP http://en.wikipedia.org/wiki/Concentrated_Solar_Power http://en.wikipedia.org/wiki/Concentrated_Solar_Power ) is very promising, for some reason many don't. Google invested in this company ( http://www.google.org/rec.html http://www.google.org/rec.html ) but has stopped investing more into CSP, preferring photovoltaics ( http://www.evwind.es/2011/11/24/google-cans-concentrated-solar-power-project/14860 http://www.evwind.es/2011/11/24/google-cans-concentrated-sol... ).

raverbashing 13y ago
The article reads biased, exactly because of the points mentioned What else uses steam turbines? Coal plants! My bet is that the high cost is in the primary(?) heat circuit, the one that runs on molten salt Also, it looks like it's still one of the first designs, so it's bound to cost more.
- ndonnellan 13y ago
  Ivanpah is steam, not salt (in salt plants, the storage system is a large component of the cost). It's not quite as simple as which area is the biggest cost component; in my experience everything adds up. You see that giant structure next to the tower? That's an air-cooled condenser, much more expensive than a wet-cooling tower, but doesn't use any water. Wouldn't surprise me if that one component cost >$50m. For the record, power towers have been around since at least the 1980s; the technology isn't super novel, but it's all in the cost reductions since then.
  - raverbashing 13y ago
    Intersting! Well, I don't know the details, if it works directly with steam then the circuits must withstand a high pressure as well no? (which can justify the higher cost) But I thought that for thermal solar the most common case would be a molten salt circuit (high temperature low pressure) transferring heat to a steam circuit for better efficiency (and the heat storage capability of the salt) Ah yes, the technology is not that new (it's heating with the sun after all)
    - ndonnellan 13y ago
      You are correct on both points. But you have to keep in mind that these projects were conceived more than 5 years ago, when everyone was hesitant to build a $2b salt plant. When you combine big money with old engineers, you get very conservative (with respect to technology). For comparison, Solar Reserve has a salt plant under construction: http://www.solarreserve.com/what-we-do/csp-projects/crescent-dunes/ http://www.solarreserve.com/what-we-do/csp-projects/crescent...
  - pivo 13y ago
    Curious: Any idea why a water cooled condenser wasn't used? Water scarcity?
- danielweber 13y ago
  I read the comments first, and was expecting some very negative article. It's not. It talks about both the positives (this is solar power that can run at night) and the negatives (it's expensive).
pjc50 13y ago
I feel that's probably a low number for the cost of turbines; turbomachinery is always very expensive. The mirrors are "heliostats" that turn towards the sun. So they have a set of motors and control electronics which will be a large part of the cost.
- jws 13y ago
  You can do a pretty nifty heliostat with passive heat collectors and a closed loop containing a fluid which vaporizes at a convenient temperature for shifting the center of gravity. (And dampers so wind buffeting doesn't jiggle the panel too much.) That said, $30 of motors and $50 of controller board is probably cheaper and more accurate. The passive system is good enough for orienting solar panels, probably not for aiming a mirror.
  - jacquesm 13y ago
    You can't use passive systems like that for pointing a mirror at a stationary source since they will aim the mirror straight at the sun. The whole point is that the mirrors form a parabola and that you point each mirror slightly different. Passive hookups like you describe would aim all the mirrors in the same direction. I've done some toying around with this and came up with a system that can run off a limited number of motors. http://pics.ww.com/v/jacques/renewables/concentrator/8.jpg.html http://pics.ww.com/v/jacques/renewables/concentrator/8.jpg.h...
  - pjc50 13y ago
    Have you seen how large the heliostats are? They look over 2 meters on each side, and the surface looks like glass or stainless steel. That's several thousand dollars of light engineering right there. Plus shipping to a desert, plus installation crew. So you're looking at $100-$1000 of motor. Several hundred dollars for the controller board (how do they know where to point? Are they networked or something?), or thousands if you have to design it yourself. Then you have to cable power to it, all of which has to be up to code. The whole thing also has to be weatherproof in a desert and not blow over in the wind. (The passive one-axis heliostat does sound interesting for DIY systems, though)
- 13y ago
xutopia 13y ago
Photovoltaics is more inline with Google's needs because it also offers increased cooling if placed above data centers. Could this be a reason why they favour PV?
- fludlight 13y ago
  Three reasons why PV works better for rooftop: 1.) Thermal only works on a massive scale. You can't fit 1000 large mirrors, a tower, a huge turbine, and the requisite plumbing on a commercial rooftop. 2.) Even if you could, the city wouldn't let you because the mirrors would distract aviation and anger neighbors. 3.) Technological advances and a capacity glut in PV manufacturing brought the price of panels down so far that it doesn't make economic sense to build thermal anymore.
redwood 13y ago
A reality in California is the high cost of regulatory siting. In this case the desert tortoise issue was costly. In many other cases, you see multi-year court battles (some in regulatory proceedings rather than traditional legal courts). A lot of the time those pushing the fight are not environmentalists but rather NIMBYs who use the laws/regulations to try to prevent progress. Same with High Speed Rail, of course. The delay and uncertainty adds a higher-than-expected cost since the investment appeal erodes. It's interesting that California's pro-environmental laws can sometimes be a hurdle to "green" projects in this sense. Different sides of the environmental issue spectrum.
- jonah 13y ago
  Finding and relocating all the tortoises and keeping an eye out for missed ones while the construction was going on is expensive but not relative to the total project cost. Source: My uncle does this work and was on this project.
  - redwood 13y ago
    That's true, the tortoises cost $22 million, a small percent of the overall cost. But siting, getting approval, finding a place where few folks will fight you, etc etc make it add up in terms of time and uncertainty and then when you do find a place you can move forward on there's a good chance it'll be remote enough to be extra costly to connect.
    - jonah 13y ago
      Granted. However, the large, open, sunny spaces in CA tend to be far from large population centers. He's now working on the project to replace/upgrade the original transmission lines from Hoover Dam which will pick up this power as well. I'm in agreement with the other commenters saying energy production should be distributed. However, this'll require a "smart grid" to balance things out - which we don't have - as well as potentially being less efficient.
- fludlight 13y ago
  Environmental concerns about displacing protected wildlife can be mitigated by installing relatively inexpensive things. Ditto for concerns about displacing low-income families. The real problem is that electricity in California is cheap and it is impossible to build a profitable utility-scale solar farm without monetary government incentives or a down-round. The former is winding down and the availability of the latter is constrained by a negative feedback loop. The expiration of federal, state, and utility incentives, mainly ARRA Section 1603, really screwed up the solar market in most of the US. Uncle Sam used to give developers of renewable projects (solar, wind, geothermal, biomass, hydro, etc) 30% of the total cost of the project in cash within a month or two of completion. Congress refused to renew that program after Solyndra, even though it was probably responsible for building several nuclear generators worth of renewable capacity. Companies can still get tax credits, but these newly-formed single-asset corporations that typically own the farms don't have any tax liability to offset. They can monetize the tax equity by doing some screwed up partnership flip structure, but that is time intensive and costly. States also used to have incentives which ranged from state tax credits to RECs. These tax credits are even harder to monetize since now you need to find an equity partner who has a multimillion dollar state tax liability in, say, North Carolina. The market for RECs collapsed right after the utilities came close to meeting their renewable portfolio standards. Most of the current pipeline of utility scale projects in the continental US involves a cramdown of the initial investors and developers. Everyone is looking for "broken" projects where someone else already spent the time and capital to get the project shovel ready. At the smaller end, there are even some contractors who financed the construction loans themselves and are being crammed down by the takeout guys. At the same time, many lenders have discontinued new lending and monetizing tax equity is expensive and difficult. There are some exceptions such as Hawaii where 90% of their electricity is generated from oil so electricity prices are sky-high (think 5x+ California), but the local utility can't keep up with the requests for permits so projects wait in an ungodly long queue.

by chime 13y ago

raverbashing 13y ago

The article reads biased, exactly because of the points mentioned What else uses steam turbines? Coal plants! My bet is that the high cost is in the primary(?) heat circuit, the one that runs on molten salt Also, it looks like it's still one of the first designs, so it's bound to cost more.

ndonnellan 13y ago
Ivanpah is steam, not salt (in salt plants, the storage system is a large component of the cost). It's not quite as simple as which area is the biggest cost component; in my experience everything adds up. You see that giant structure next to the tower? That's an air-cooled condenser, much more expensive than a wet-cooling tower, but doesn't use any water. Wouldn't surprise me if that one component cost >$50m. For the record, power towers have been around since at least the 1980s; the technology isn't super novel, but it's all in the cost reductions since then.
- raverbashing 13y ago
  Intersting! Well, I don't know the details, if it works directly with steam then the circuits must withstand a high pressure as well no? (which can justify the higher cost) But I thought that for thermal solar the most common case would be a molten salt circuit (high temperature low pressure) transferring heat to a steam circuit for better efficiency (and the heat storage capability of the salt) Ah yes, the technology is not that new (it's heating with the sun after all)
  - ndonnellan 13y ago
    You are correct on both points. But you have to keep in mind that these projects were conceived more than 5 years ago, when everyone was hesitant to build a $2b salt plant. When you combine big money with old engineers, you get very conservative (with respect to technology). For comparison, Solar Reserve has a salt plant under construction: http://www.solarreserve.com/what-we-do/csp-projects/crescent-dunes/ http://www.solarreserve.com/what-we-do/csp-projects/crescent...
- pivo 13y ago
  Curious: Any idea why a water cooled condenser wasn't used? Water scarcity?
danielweber 13y ago
I read the comments first, and was expecting some very negative article. It's not. It talks about both the positives (this is solar power that can run at night) and the negatives (it's expensive).

pjc50 13y ago

I feel that's probably a low number for the cost of turbines; turbomachinery is always very expensive. The mirrors are "heliostats" that turn towards the sun. So they have a set of motors and control electronics which will be a large part of the cost.

jws 13y ago
You can do a pretty nifty heliostat with passive heat collectors and a closed loop containing a fluid which vaporizes at a convenient temperature for shifting the center of gravity. (And dampers so wind buffeting doesn't jiggle the panel too much.) That said, $30 of motors and $50 of controller board is probably cheaper and more accurate. The passive system is good enough for orienting solar panels, probably not for aiming a mirror.
- jacquesm 13y ago
  You can't use passive systems like that for pointing a mirror at a stationary source since they will aim the mirror straight at the sun. The whole point is that the mirrors form a parabola and that you point each mirror slightly different. Passive hookups like you describe would aim all the mirrors in the same direction. I've done some toying around with this and came up with a system that can run off a limited number of motors. http://pics.ww.com/v/jacques/renewables/concentrator/8.jpg.html http://pics.ww.com/v/jacques/renewables/concentrator/8.jpg.h...
- pjc50 13y ago
  Have you seen how large the heliostats are? They look over 2 meters on each side, and the surface looks like glass or stainless steel. That's several thousand dollars of light engineering right there. Plus shipping to a desert, plus installation crew. So you're looking at $100-$1000 of motor. Several hundred dollars for the controller board (how do they know where to point? Are they networked or something?), or thousands if you have to design it yourself. Then you have to cable power to it, all of which has to be up to code. The whole thing also has to be weatherproof in a desert and not blow over in the wind. (The passive one-axis heliostat does sound interesting for DIY systems, though)
13y ago

xutopia 13y ago

Photovoltaics is more inline with Google's needs because it also offers increased cooling if placed above data centers. Could this be a reason why they favour PV?

fludlight 13y ago
Three reasons why PV works better for rooftop: 1.) Thermal only works on a massive scale. You can't fit 1000 large mirrors, a tower, a huge turbine, and the requisite plumbing on a commercial rooftop. 2.) Even if you could, the city wouldn't let you because the mirrors would distract aviation and anger neighbors. 3.) Technological advances and a capacity glut in PV manufacturing brought the price of panels down so far that it doesn't make economic sense to build thermal anymore.

redwood 13y ago

A reality in California is the high cost of regulatory siting. In this case the desert tortoise issue was costly. In many other cases, you see multi-year court battles (some in regulatory proceedings rather than traditional legal courts). A lot of the time those pushing the fight are not environmentalists but rather NIMBYs who use the laws/regulations to try to prevent progress. Same with High Speed Rail, of course. The delay and uncertainty adds a higher-than-expected cost since the investment appeal erodes. It's interesting that California's pro-environmental laws can sometimes be a hurdle to "green" projects in this sense. Different sides of the environmental issue spectrum.

jonah 13y ago
Finding and relocating all the tortoises and keeping an eye out for missed ones while the construction was going on is expensive but not relative to the total project cost. Source: My uncle does this work and was on this project.
- redwood 13y ago
  That's true, the tortoises cost $22 million, a small percent of the overall cost. But siting, getting approval, finding a place where few folks will fight you, etc etc make it add up in terms of time and uncertainty and then when you do find a place you can move forward on there's a good chance it'll be remote enough to be extra costly to connect.
  - jonah 13y ago
    Granted. However, the large, open, sunny spaces in CA tend to be far from large population centers. He's now working on the project to replace/upgrade the original transmission lines from Hoover Dam which will pick up this power as well. I'm in agreement with the other commenters saying energy production should be distributed. However, this'll require a "smart grid" to balance things out - which we don't have - as well as potentially being less efficient.
fludlight 13y ago
Environmental concerns about displacing protected wildlife can be mitigated by installing relatively inexpensive things. Ditto for concerns about displacing low-income families. The real problem is that electricity in California is cheap and it is impossible to build a profitable utility-scale solar farm without monetary government incentives or a down-round. The former is winding down and the availability of the latter is constrained by a negative feedback loop. The expiration of federal, state, and utility incentives, mainly ARRA Section 1603, really screwed up the solar market in most of the US. Uncle Sam used to give developers of renewable projects (solar, wind, geothermal, biomass, hydro, etc) 30% of the total cost of the project in cash within a month or two of completion. Congress refused to renew that program after Solyndra, even though it was probably responsible for building several nuclear generators worth of renewable capacity. Companies can still get tax credits, but these newly-formed single-asset corporations that typically own the farms don't have any tax liability to offset. They can monetize the tax equity by doing some screwed up partnership flip structure, but that is time intensive and costly. States also used to have incentives which ranged from state tax credits to RECs. These tax credits are even harder to monetize since now you need to find an equity partner who has a multimillion dollar state tax liability in, say, North Carolina. The market for RECs collapsed right after the utilities came close to meeting their renewable portfolio standards. Most of the current pipeline of utility scale projects in the continental US involves a cramdown of the initial investors and developers. Everyone is looking for "broken" projects where someone else already spent the time and capital to get the project shovel ready. At the smaller end, there are even some contractors who financed the construction loans themselves and are being crammed down by the takeout guys. At the same time, many lenders have discontinued new lending and monetizing tax equity is expensive and difficult. There are some exceptions such as Hawaii where 90% of their electricity is generated from oil so electricity prices are sky-high (think 5x+ California), but the local utility can't keep up with the requests for permits so projects wait in an ungodly long queue.