Back here - https://news.ycombinator.com/item?id=39929842 - the US Govt. was calculating that their (funded) pilot project for heated sand energy storage could get 50% real-world efficiency. Or 55% if they used a more-complex turbine cycle.

If true - then this new U-M thermophotovoltaic system will either need a whole lot of improvement, or its use will be confined to corner cases.

Energy efficiency is nice, but capital efficiency is more important.

There are a lot of grids that have excess generation from solar or wind that drive prices very low on a routine basis. As long as the difference between the low and high price is enough, the question is more about capacity than efficiency, IMHO.

That other system seems mechanically complex, you've got to move sand through silos. This system seems a lot simpler, make a big box of sand/whatever, run heating elements through it, put modified solar cells on the outside, apply heat when power is cheap, draw power when power isn't cheap.

Might be able to do it at smaller scale too.

I was hoping Battery in every home would have solved this issue. We would basically have a distributed battery storage system in every home to balance capacity.

But it seems we are still far from it happening. Tesla PowerWall 3 isn't that much of an improvement.

Tangentially Exhaust Heat Recovery is a very interesting concept for ICE vehicles. The side effect of internal combustion engines is the most efficiency you can get out of the Otto cycle is like 40% of the heat shoots right out the tailpipe, into the coolant etc.

Toyota does this on the Prius where it passes coolant through the exhaust heat to recover it and help get the car up to cycle faster, but there's other cases that have been floated. For example, the reverse of a Peltier cooler is a TEG, which generates electricity from heat directly and use that to charge the battery

Unless the temperature differential is huge, hundreds of degrees C, the efficiency of these is tiny - around 5-7%. Incorporating things like stirling generators could theoretically get up to around 35-45% efficiency, but that adds weight and complexity, with the maintenance and production costs offsetting the potential economic gains in reduced fuel usage.

Better thermoelectric generator technology is possible - there are lots of exciting possibilities with graphene, and heat pumps, and so forth, but saving $10,000 in fuel costs over 5 years for $15,000 worth of technology in a car doesn't make sense. Auto manufacturers optimize for price and performance and space and maintenance and all sorts of competing factors, so it's not always as easy as engineering a technical solution and slapping it on top of everything else that goes into a product.

We already have heat recovery systems in production models, though not for generating electricity of course. Another driving feature is emissions, not cost. Meeting emissions requirements drives up the cost and complexity of the vehicle.

TEGs are essentially unworkable if you want to produce kilowatt-class energy in a size that can fit a car.

You don't want to produce... kilowatt class energy, you would want to offset the losses of running accessories directly off the driveshaft. Spinning an alternator takes energy. Spinning parts also have wear issues like bearings etc.

Secondly, in the Prius approach, it's about getting the engine to optimal temperature to reduce emissions and increase efficiency

Yeah, but all these things (even the heat transfer from exhaust toncoolant in the Prius) need to transfer many kilowatts of energy to actually work.

A kilowatt-class liquid heat exchanger is the size of a large coffee cup. A kilowatt-class TEG is the size of a car engine. See for instance the paper linked below, Fig. 12, their module is 1 m x 0.3 m x 0.35 m.

For comparison the Ford 1.0 Ecoboost engine is famously said to fit on an A4 sheet of paper and produces 100 kW. So adding a TEG you would more than double the size of your engine for a 1% efficiency gain.

https://pangea.stanford.edu/ERE/db/WGC/papers/WGC/2015/26009...

How about in any industrial building that has a smokestack?

Resistors can easily be scaled up. The heat storage medium can be easily scaled up. But what about these thermophotovoltaics whatever they are? Can they be cheaply mass produced?

When they reach lower temperatures like 900C, we could finally replace the inefficent hot water turbines in coal power stations.

It’s probably not cost effective, but could these heat storage devices be miniaturized and used in a house or camping setting?

Do you want to carry around a 1400C object and its necessary layers of insulation?

(The square-cube law means you get a greater volume of storage per area of insulation at larger sizes, so I'd expect them to be big)

Yeah, it's a dead end. That French guy, whatisname, J-B Fourier, said pretty much 99% of what you need to know about this way back in the early 1800s.