I'm curious about "why sand", is it better at holding heat than say water?
Sand has way less heat capacity then water per kg (about half).
Water can be heated to 95C with standard unpressurized vessel. Sand in this application is heated to 600C.
Sand is denser then water (kg/m3).
For the same heat energy stored this comes out to about 2.5x more volume of water(95C) compared to sand(600C).
Water and Sand are both dirt-cheap.
Hot water can be managed with standard plumbing equipment.
Sand needs some high temperature piping (hot air to water heat-exchanger, resistive heat tho heat up the sand).
How well both contain the heat is primarily dependent on the isolation. Which favors the smaller footprint of sand, but needs to isolate a higher temperature difference...
So yes, the volume of 95C water would be much greater than that of 600C sand, but if volume wasn't an issue you could do it much more efficiently. Alternatively, you could use battery storage for just the electrical capacity required and not the (much higher) thermal capacity which may be more cost effective when you look at the conversion.
[1] https://www.sciencedirect.com/science/article/abs/pii/S03605...
Using a heat pump will increase the yield. Usable temp range from 95C water all the way to 0C ice in theory (latent heat).
And a modern isolated home helps, but seasonal heat water storage is basically a big tank, with a house build around it.
https://www.energie-experten.ch/de/wissen/detail/waermespeic...
But sand doesn't freeze (or perhaps, sand doesn't melt) under normal temperatures. That's probably more important for an icy region.
Exactly to your point, though, one of the great things about using sand for this application instead of water is that you can probably just shut the thing off for maintenance without having to worry about draining all of the sand out. If it freezes up due a bit of residual moisture content it's not going to expand nearly the same way that a silo full of water would, and it should be easy enough to thaw out just by putting some heat into it.
A quick caveat/clarification: It's only true if you're pushing the system over the 100°C mark. Otherwise a volume of liquid water--with its greater latent heat-capacity--will outclass the same volume of sand.
Water's heat-capacity is 4.186 J/g°C, while estimates for sand run towards ~0.830 J/g°C. If we also assume the sand is 1.6x denser, then our below-boiling water still comes out ahead at ~3.15x the joules per volume.
There are hints [0] this system tops out around 600°C.
Is that just speculation or did you read it somewhere? IIRC the original motivation of PNE was a bunch of engineers at uni speculating on how to build the perfect building for engineers, and making it self-sufficient would require handling its own heating, which they originally thought would be best done with a big hot-water tank to store the heat. No turbine was suggested, IIRC.
That means that the hot middle of a sand pile stays hot, instead of advecting to the outside, where it cools by conduction.
And you might vent steam, but you should probably take into account that while water < 45 degrees or so is pretty innocent, steam will strip flesh from bone starting at 180 degrees or so, it won't "just" burn you.
A given mass of water:
- At 20C is less dense/higher volume than water at 4C
- As you cool it the density increases and the water shrinks
- At 4C the water is now at maximum density and minimum volume
- As you continue cooling it from 4C, it starts expanding again
- At 3C the water is lower density than it was at 4C and is again expanding
- At 0C where the water starts to transition to a solid the volume expands significantly
I guess I wasn't clear that the expansion only starts again as you go from 4C -> 0C in the liquid phase.
Which means that you have rather little of delta to work with. And at upper end it becomes somewhat risky to have large container of water that is beyond boiling in normal pressure...
With sand you can use very simple heat-exchangers. No need to use exotic heat pumps that require extra energy...