My greenhouse active thermal mass project is all about regulating a room's temperature by clever use of stored heat. Unlike storing electricity, storing heat is trivial - a large amount of water can store heat for far less money than storing the same amount of energy as a battery. There is, however, a number of gotchas. The most important of which are that heat transfer is proportional to the difference in temperature, and heat always passes from the hotter body to the colder body. Heat won't pass from the cooler to the hotter.
Another annoyance is that heat transfer tends to require large surface areas, large mass flows and long periods of time. Which means that experiments must be performed over time periods of a day or more. This means that only a computer has the required patience to control it. I've built a python based PID controller for the greenhouse system and pfh and I have been working on a smarter controller.
We've now mostly ironed out the bugs and worked out what things work and what things don't - expect to see a few interesting publications about this project.
The system has used 0.7kWh over 3 days and moved about 170kWh of heat (assuming my measurement system is accurate - still not sure). That's an overall COP of about 240.
I met the guy who grew my silver perch (by coincidence!) and he came around on monday to look at my setup. Now he's planning to build a bigger version to regulate his fish shed. We're going to visit tomorrow and hopefully flesh out a complete plan.