Talk-Polywell.org

Self-Charged Graphene Battery Harvests Electricity from Thermal Energy of the Environment

...the theoretical power density is about 73.3 KW/Kg

but that's probably per Kg of graphene, not a complete, fully-assembled battery cell (not clear from the text).

The best lithium-ion battery power density seems to be around 21.4 KW/kg.
http://en.wikipedia.org/wiki/Power-to-w ... ll_systems

This is pretty special. Are these batteries going to be expensive?

I'll guess about a $bazillion per D-cell until Chinese factories kick in.

For me its looks like this are against the second law of thermodynamics. They may have a black ritual to bound a Maxwell demon to the graphene or I have not understand this correctly.

I could be reading it wrong, but it seems no different than solar panels really. You're taking something already there in the environment to recharge the battery. The question then would be how efficient is it to do so and is it economically feasible to do so.

Nothing is stored except heat in the fluid. Kinetic energy (heat) of ion impacting graphene is converted to electron energy.

If there is no temperature differential (heat flow) involved, then it's voodoo.

Since the thermal energy they're using penetrates, you're not as worried about open surface area. And if power goes up with heat, then this could also be used as a sort of thermocouple to get energy from a thermal reaction.

It's a good thing that calorimetry is a non-controversial subject.