Google Invests $168MM in World's Largest Solar Power Tower Plant | Hacker News Reader

Google Invests $168MM in World's Largest Solar Power Tower Plant

mashable.com

55 pointsshimonamit23 comments

That is such a smart design (using mirrors to focus sunlight onto one point rather than using lots of solar panels spread about). How long have scientists had this idea?

I remember seeing 1970s vintage books with these sorts of things in them.

The main difficulty is maintaining a high temperature system with lots of moving parts, but in that respect it's not much worse than a fossil fuel plant.

Edit: not forgetting the usual source reliability / distribution problems for renewables.

A good while. There's actually a plant of this design (smaller though, only 20MW) outside of Seville, it's been there for a while.

Also, to the sister post, there are relatively few moving parts in the collector itself. Furthermore, these are all conditions that we've dealt with already with traditional turbine systems (they get pretty hot too), and aircraft design. The harder part is getting the sun tracking right. The Seville (and I guess this one too) plant actually has to point a bunch of mirrors off the tower (they park them into a tower of light above the tower... looks really cool) at peak, since it actually gets too hot (it would melt concrete or something ridiculous I heard...)

http://en.wikipedia.org/wiki/PS20_solar_power_towers

Seville has one

http://news.bbc.co.uk/1/hi/sci/tech/6616651.stm

but the idea is as old, Archimedes is said to have used mirrors to set fire to a roman fleet.

The first experimental solar power towers were built in the 1980s. The first commercial plant was constructed in Seville, Spain and became operational in 2007.

I remember them being in Sim City 2000. ;-) So, it's been a while.

I am actually disappointed in the video. There were no 3d models, and the project talked very little about the engineering effort. Instead we are given very stoic music and a bunch of PR.

http://en.wikipedia.org/wiki/Archimedes#The_Archimedes_Heat_...

Two millennia or so. Power generation has always been applying new ways to boil water (save wind and hydro).

Actually I misconstrued what the design was. I thought the mirrors would project a higher concentration of sunlight onto a traditional solar panel (except one designed to accept much higher input/density), thus saving the cost of making many panels and many periodic replacements. Mirrors are presumably cheaper and/or more durable than solar panels.

The New York Times was writing about concentrated solar power as early as 1868, in the context of the imminent exhaustion of coal.

http://select.nytimes.com/gst/abstract.html?res=F40A10FB3854...

Has anyone come across a calculation on how this affects the local climate? Won't this system be sucking 390MW of energy out of the environment to convert to mechanical energy?

The effect must be noticeable. I get the same feeling about wind. If solar ever became large scale wouldn't we be directly countering global warming? And won't solar panels/mirrors be directly competing for the solar energy that wind uses?

It's quite tiny as a fraction of the total solar energy. Remember also that all energy we are collecting by burning coal, gas, and oil today was once solar energy as well.

Total energy consumption of the world: 474 exajoules. Total solar energy absorbed by earth: 3,850,000 exajoules

But your argument is even more rudimentally answered as such: energy captured by solar panels is still in the Earth's atmosphere. It'll appear later as heat anyways, once its energy is used practically. If you want to counter global warming with solar panels directly, they would have to be situated above the atmosphere, and NOT transmit any energy down to the surface.

There are a number of solar thermal plants in development now. It is worth noting that it is possible to store the energy for use in dark times by including storage tanks in the molten salt loop, though I don't see that in this plant's description and it isn't important until there is a significant solar contribution to the total energy production.

Original announcement: http://googleblog.blogspot.com/2011/04/investing-in-worlds-l...

A better summary, as it gives you some specifics, it's a 392MW (gross) generating plant (I found another link that also mentioned 370MW nominal). Compare to a typical coal-fired plant at 500MW, up to 1,000MW or more.

I guess that means that 370MW is the nominal optimal generating power? From what I read, thermal solar has a capacity factor of only about 20% [1], which means this solar farm will average out to about 70MW, which isn't that much.

Coal has a capacity factor of about 60% and nuclear 90% [2], so this is similar to a 120MW coal plant or a 80MW nuclear power plant. Of course, you can't really compare them only on total power provided, as solar provides most of its power during the day, which is also the time in which the energy demand is the highest.

[1] http://en.wikipedia.org/wiki/Solar_thermal_energy

[2] http://www.eia.doe.gov/cneaf/electricity/epa/epaxlfile5_2.pd...

That wikipedia article seems to be citing somewhat conflicting references. Further down it says "The capacity factor for power towers was estimated to be 72.9%" [35] (http://www.nrel.gov/solar/parabolic_trough.html).

A heliostat based design is going to significantly raise the amount of usable insolation. And the mojave desert is probably close to an ideal location for such a plant.

Quoting your [1] (end of power tower desing paragraph http://en.wikipedia.org/wiki/Solar_thermal_energy#Power_towe... ):

>The capacity factor for power towers was estimated to be 72.9% and 56.2% for parabolic troughs.

so the 20% figure is simply incorrect, and is probably the one applying to rooftop solar heaters, which are a totally different type of design.

The NREL figures are for plants with heat storage (large tanks of high-temperature molten salt which accumulate heat, to power turbines at night). Without heat storage you only have 20-25% at most. Ivanpah has a capacity factor of 31%, but that's not solar; part of that comes from burning natural gas.

http://www.nrel.gov/csp/solarpaces/project_detail.cfm/projec...

The 20% efficiency is, as stated in the other comment, about correct for a plant without thermal storage. See far example [1]. What happens with the higher efficiency ratings is that they put more mirrors on the tower, or have a smaller turbine. That is, the ratio between thermal input and electric output increases. They can do this because any heat that isn't used by the turbine right at that moment can be stored temporarily and then used later on when the incoming heat isn't enough to power the turbine at 100%.

Which means that with an efficiency rating of 73%, you still need many more mirrors than with the capacity factor of 25%, but you just don't upgrade your turbine, and instead rely on thermal storage. Which one is preferred (bigger turbine or thermal storage) depends on the cost of turbines and thermal storage, and on whether you want to have baseload power or higher output during the day.

As a note, the 72.9% and 56.2% you quoted were estimates made in 2003 for the year 2020, so I wouldn't put to much trust into them.

As for Ivanpah itself, it appears to not have any thermal storage planned (which would have to be the biggest thermal storage ever created to be useful), according to [2], so it would be limited to about a 25% capacity factor.

[1] http://www.nrel.gov/docs/fy11osti/49303.pdf

[2] http://www.basinandrangewatch.org/IvanpahFSA-reliability.htm...

How much would a typical ~500MW coal-fired plant cost to build and operate (including cost of coal over the lifespan of the plant)? No sarcasm, I’m interested :)

The price of thermal coal (USA) is averaging around $2.26/million btu [1]; the average heat rate (inverse of efficiency) of coal plants is 10,414 btu/kWh electricity [2]. So the fuel component of cost is about 2.35 cents/kWh, or for a 500 MWe plant with 73.6% capacity factor [3], about $76 million/year.

There's also substantial capital costs. According to [4] (table 3.1 p. 19 [5]), representative costs for a new coal plant would be around $650 million for 500 MWe, and a total (levelized) cost of around 4.8 c/kWh. This is from a slightly lower coal price ($1.50/MMBtu).

[1] http://www.eia.doe.gov/cneaf/electricity/epm/epm_sum.html

[2] http://www.eia.doe.gov/cneaf/electricity/epa/epat5p3.html

[3] http://www.eia.doe.gov/cneaf/electricity/epa/epata6.html

[4] http://web.mit.edu/coal/

[5] http://i.imgur.com/eJlUL.png

Also compare it to the existing 10-20MW "solar tower"-type plants. It is a serious step forward!

Reply on news.ycombinator.com