solar thermal salt

=energy =suggestion =explanation =molten salt



This is a post about solar thermal power (STP), which implies that I still consider that worth thinking about. In 2019, that's a significant statement, and I should qualify it:

- STP is more expensive than photovoltaic (PV) power. The only reason to use it is to use molten salt thermal energy storage.
- If flexible power production is available (usually cheap natural gas or hydroelectric power) then that + PV + wind is cheaper than STP.
- STP requires a large scale and low interest rates to be viable. Past projects didn't do well enough for future projects to get big investments at low interest rates.
- STP doesn't work well when there are clouds. This makes locations more limited than PV locations.


The most notable failure of STP was Ivanpah, the largest solar power tower project. Yes, it produces power, but it was ~12x too expensive and killed future investment, so it should be considered a failure. 12x is a big hurdle, but things aren't as bad as they seem from Ivanpah, because:
- bad project management made it 3x as expensive as it should have been
- it didn't have thermal energy storage, which would've made it 2x as valuable
- the design used steam turbines, which was an obviously dumb idea


Really, power towers are "only" inherently ~2x too expensive.


Mirrors are not the problem. Flat mirrors are cheaper than solar panels, and STP is usually more efficient than PV solar power. What is expensive about power towers is:
- the solar receiver
- turbines
- heat exchangers
- mirror supports
- mirror drive systems


The solar receiver is a big heat exchanger that absorbs light and transfers heat to a (somewhat corrosive) molten salt. It could be much cheaper to instead use transparent quartz tubes containing molten salt with some graphite particles in it to absorb light. This "Direct Absorption Receiver" (DAR) idea has been considered since the 1980s.


So, what's the problem?


Those DAR papers usually assumed a Na-K-Li carbonate (mp 397 C) salt. But that's only really usable at 500+ C (too high a minimum temperature) and lithium is expensive.


In today’s commercial solar-thermal systems, the molten salts used are usually:
- [60% NaNO3 + 40% KNO] ("Solar Salt") (usable up to ~575 C)
- [53% KNO3 + 40% NaNO2 + 7% NaNO3] ("Hitec") (usable up to ~535 C)


The max temperatures of those are lower than is optimal, but they work well enough...but nitrates are oxidizing, which is a problem if you want graphite particles!


There are many molten salt eutectics, but there are also many criteria to meet:
- low melting point
- high boiling point or decomposition temperature
- low cost per heat capacity
- low viscosity per heat capacity
- low corrosiveness


Hydroxides are corrosive, lithium and bromine are limited and expensive, and chloroaluminates have low boiling points.


I think the best choice for a DAR salt for power tower STP with molten salt thermal energy storage is:
13.8% NaCl / 41.9% KCl / 44.3% ZnCl2 ("Zn salt") (mp 229 C)


With that, air-free salt is important to minimize corrosion. [Ni + ~18% Cr] alloy would probably still be needed for heat exchangers, which is unfortunate but acceptable.


Like many molten salts, that should be used well above its melting point to reduce viscosity, maybe at 375 C <-> 750 C. Operating well above the melting point also makes it less likely that salt will freeze in pipes, which is usually hard to fix.


That's good enough, but there's a hidden assumption here that only 1 salt type can be used. It's possible to use separate low and high temperature salts! Is there a reason to? Maybe.

27.5% NaCl / 32.5% KCl / 40% MgCl2 ("Mg salt") (mp 383 C)

has a higher melting point than Zn salt, but gives higher heat capacity and lower corrosion at high temperatures.


It's possible to do something like:
- heat CO2 with Zn salt to 680 C
- expand CO2 with a turbine to 420 C
- reheat CO2 with Mg salt to 800 C
- expand again


Is this better? Well, maybe. It depends.





back to index