I think the idea was to use the water largely as the heat reservoir/medium, rather than the rocks themselves (mine wallrock in the more normal groundsource. It would probably be cheaper to bore a 90m borehole adjacent to a warm stope (inc rocks) than it would to collar a shaft to make it safe to install a means of access to see if a tube installation in the water (at what level?) would be feasible (ie:- the shaft wasn't choked 3ft below water level). You'd also have to be sure that bits of wall rock/collar, etc wouldn't come unhinged and smash up your installation. Fine in Mt Wellington, but not so fine up at Penberthy Croft.
There's a few points here, if you have chemical oxidation of pyrite and biological oxidation to a lesser degree, this creates a load of heat, presumably, due to the low SHC of rock and low surface area, this takes a while to cool down. So you'd presumably get left with quite a large reservoir of warm water. The heat won't leak away and it's unlikely (is it?) that you'd get appreciable convection to the surface and water/air exchange in the shafts.
Several years ago, I found myself enjoying the very warm water in New Roskear Shaft in Camborne. I presume this was due to the residual heat in the mine warming the water. (I don't imagine there is much oxidation of pyrite underwater, via whatever means). Since roskear is water-upcast, most of the nice warm water ends up in the adit, so it's probably violates the problem I'm thinking about.
That problem being you take your U tube, whack it in a wet shaft and turn your heat pump on. You'd get localised cooling. The question would be, would the cold water sink and you get some sort of convection set up, or would you end up with a thermal gradient in the shaft....? Whether this happened or not would dictate the surface area of your U tube. The requirements could be hugely different, due to what happens in practice. It would be exceedingly complicated to model (would it?) the behaviour of different temp fluids in different size (area and depth) shafts. It would be quite interesting though.
I remember seeing the rosemanowes quarry hot rocks project spraying everywhere when I was very young. I gather this was not a success. I wonder if the problems of a low surface area (despite mega deep borehole) and the low SHC of granite ultimately limited the output.
I know Mount Wellington is being used by Kensa to abstract heat. I wonder if he has done the maths for it being the usual temp of wallrock (10 ish on ave) rather than the warm water in the shaft. Or whether he's just put down a mega tube, knowing that there will be plenty of energy in the return fluid.
I imagine health and safety would have something to say about getting a set of tubes installed in most (quite deep) shafts, considering how gnarly most of them are. We have enough trouble going down with no method statement, no risk assessment and no bloody great tubes!