NREL study says solar thermal cheap

[eclipse]

A cost/performance comparison between power tower and parabolic trough concentrators was made by the NREL which estimated that by 2020 electricity could be produced from power towers for 5.47 ¢/kWh and for 6.21 ¢/kWh from parabolic troughs. The capacity factor for power towers was estimated to be 72.9% and 56.2% for parabolic troughs.[47] There is some hope that the development of cheap, durable, mass producible heliostat power plant components could bring this cost down.[48]

From the wiki
en.wikipedia.org/wiki/Solar_thermal_energy#cite_note-49

Anyone got a counter-study?

So I gather this is another case of it being affordable electricity IF we don't ask that electricity to be baseload? It's like the argument that says Solar PV is cheaper than coal because all they're asking it to do is reduce your electricity bill and afternoon peak demand on the grid, which to me seems fine for those who have the money to buy Solar PV in the first place. Let them! Just don't pretend it's going to supply baseload power or close any coal plants soon.

[David B. Benson]

One readily adds a thermal store to the concentrator generator. Up to 5--7 hours of supply has already been accomplished.

[anonposter]

How is solar meant to get a capacity factor greater than 50% without being placed in high orbit?

Are they using a different definition of capacity factor?

[edireland]

Presumably they have a giant tank of molten salt that allows heat to be "stored" on a diurnal basis.

Still doesn't help with seasonal variations.

[Graham Palmer]

Jul 13, 2013 16:51:41 GMT 9.5 @anonposter said:

How is solar meant to get a capacity factor greater than 50% without being placed in high orbit?

With a solar field multiple of 2 to 3 (i.e. a 100 MW solar field, but a 30 to 50 MW power block), with storage and natural gas backup (Gemasolar quote a 15% contribution of natural gas with an overall capacity factor of 75%).

www.nrel.gov/csp/solarpaces/project_detail.cfm/projectID=40

The trade-off of course is cost since the solar field and receiver make up around half the total cost. Gemasolar has a cost of around AUD $23,000 /kW under ideal solar conditions.

[anonposter]

So they are down-rating the plant from what it can actually collect and then using that new lower figure to calculate a less unacceptable capacity factor?

Seems a bit dishonest.

[David B. Benson]

Those NREL LCOE estimates are about 4 times smaller than current actual LCOE for concentrators. I question whether that much of a so-called learning curve will be possible.

[Graham Palmer]

Jul 13, 2013 20:56:14 GMT 9.5 @anonposter said:

So they are down-rating the plant from what it can actually collect and then using that new lower figure to calculate a less unacceptable capacity factor?

See page 8
www.irena.org/DocumentDownloads/Publications/RE_Technologies_Cost_Analysis-CSP.pdf

At a tehnical level, determining the optimal configuration of solar multiple and storage is really just part of the R&D. The issue you allude to is probably more a criticism of the spruikers who exploit the sleight-of-hand that is possible by claiming a high CF without highlighting the cost trade-off and reliance on gas backup, and also assume a future learning curve that usually ends up being optimistic. Ted Trainer has also brought attention to the embodied energy, which has so far received little research attention.

socialsciences.arts.unsw.edu.au/tsw/JandDreplytoreply.htm

Since all of these projects are heavily subsidised, it is difficult to know how commercial operators would ultimately configure them in a competitive market with a carbon price.

[cyrilr]

Counter-study? Not necessary. A simple look at how high Spain's feed-in-tariff must be to get even a few CSP plants deployed, shows that 5 cents per kWh is magic pixie dust. And a Gemasolar plant costing $23000/kWe is not going to produce power at 5 cents/kWh either. Not even for 20 cents per kWh, in fact.

Somehow we are supposed to believe a factor of 5 cost reduction. With a very conventional technology - mirrors, trusses, concrete, steam turbines. Not going to happen. The SEGS plants built decades ago were cheaper than modern plants, because of materials and labor inflation that is higher than any learning curve effect. Adding giant tanks of hot salt, with all the insulation, freeze protection etc. isn't helping the cost to go down, as evidenced by Gemasolar and Andasol plants.

[dwalters]

Cyrilr is of course absolutely correct. These low production costs is made up...without subsidies simply do the math.

Secondly the very FAKE "capacity factor" numbers are also the result of base 8 numerology or something similar.

Lets examine this: Solar hours per day at *name plate capacity*...say 100 MWs...is what can be achieved around noon time plus or minus 2 hours (on completely sunny days). Before and after that, of course, it's less. Total solar capacity is never more than .18 or slightly less than 1/5 of the day. That is, 20% of a 24 hour day. (this is true for PV as it is for CSP with thermal storage).

If the "100MW solar plant" is hooked into some sort of storage then the energy from the solar array, instead of producing steam to produce electricity, is shunted to the thermal storage. How much? Well, 100% minus 80% = 20%. That is, to achieve 24 hour output, generation has to be cut down by 80% or, in this example, 80MW of the 100MW capacity goes into thermal storage...and not out to the grid. The rating of the plant then falls from 100MW to 20MW on a 24 hour basis.

Now, to be fair, obviously solar PV is not aiming for 24 hours. It's aiming, really, to provide peak load demands which more often fall about 4 to 6 hours AFTER solar peak at noon.

The bottom line is that the costs to build what is essentially a 20MW CSP plant is insane because of the very low generation you get out of it.

And that's only with solar storage designed, on a *warm night* to carry you to about midnight or something similar. Wow, wow, wow. Chutpah. Imagine running any nation on this?

[Greg Simpson]

Total solar capacity is never more than .18 or slightly less than 1/5 of the day

From what I've seen 25% is readily available in most desert areas with flat plates properly tilted, or 6 kWh/m²/day. You can do better in the Sahara and some other very sunny places. See www.nrel.gov/gis/images/map_pv_us_annual_may2004.jpg for example. All I have bookmarked are USA maps, but I believe Australia is at least as good.

Also, solar thermal often uses two axis tracking, which can boost the capacity factor to 30%.

While I much prefer nuclear, this is the one renewable that I think could work, at least sort of.

[edireland]

People don't tend to live in deserts.
The power lines would be jugulars stretching from industrial centres into the desert.
Desert that would have to be garrisoned with vast armies of ground troops (especially in the Europe/Sahara scenario) that would effectively have to do Western Sahara-style fortifications on an unprecedented scale.

It would be expensive and would cause all sorts of othher problems.

[Greg Simpson]

It would be expensive and would cause all sorts of other problems.

I have no doubt that's true.