The Motley Fool Discussion Boards

Previous Page

Personal Finances / Building / Maintaining a Home

URL:  http://boards.fool.com/my-thought-is-that-once-i-find-out-how-long-and-20120130.aspx

Subject:  Re: Radiant Heat System Date:  1/8/2004  8:32 PM
Author:  Goofyhoofy Number:  44394 of 128288

My thought is that once I find out how long (and what diameter) the hot water piping is, I can look up the expected rate of heat loss. Then I can pick a minimum desired temperature for the static hot water, and from that figure out how many gallons per day of hot water would need to be circulated, at most, to maintain the desired temperature.

Unless you have a lot of experience with this and similar HVAC stuff, I will guess that whatever number you come up with will be wildly wrong because of things which you do not anticipate. I'm sure you will have a nice theoretical package, but in practice I'll bet you do a lot of fudging to get it right eventually. Maybe not, and I'll be interested in how it works out.

I figured I needed a certain number of BTU's to be produced and a certain number of BTU's transferred under the floor. I didn't see how the "fins" would change that equation, except to disseminate the heat more evenly, which I didn't think a problem (and turned out not to be.) As it turned out, the number in the pipe transferred to the joist cavity (I think) but dissipated as it moved through the (so-called) thermal mass of the concrete board, thinset, and tile. I really thought [avg] 90 degree water would produce an 80 degree floor and it didn't. (The Ultra-Fin book gives that exact scenario as an example, and IT DON'T WOIK DAT WAY.)

I expect this will be a small number, and that a very small pump could simply run continuously (rather than being thermostatically controlled, like the Auto-Circ models).

I don't understand why that's preferable, but OK. My forced air system pulses, a hot water radiator pulses, but by the time the "heat" gets into the radiator the room stays a fairly constant temperature, because the radiator, er, radiates more slowly than the hot water cycles. Likewise the heat from the "pulses" through my copper pipes (by the time it gets up through the floor.)

So my thinking for the floor is that I could run supply and return lines from the water heater to the bathroom floor (it's not far) and run tubing back and forth under the floor, much like you did. Then connect the output of the circulation pump to the supply line, and connect the return line to where the circulation pump would normally connect. And, place a solenoid valve from the supply to the return line where they connect to the pump....
...Then control the valve with a thermostat in the heated floor, much like the NuHeat people do.


Again, it looks like you are building piece by piece, and at greater expense, everything that a thermostatically controlled pump does for about $120. If you want an Intermatic timer on it (slick) add $40. I don't know why you would do that (except for the joy of building it, which I fully understand, if that's your motivation.)

With the valve closed, water would have to flow through the floor system. With it open, the floor system would be effectively "short-circuited" and have little, if any, flow

In my original design I had a gate valve and a double path, from the stubout straight to the sink, or from the stubout through the underfloor piping and straight to the sink. Same kind of thing, I guess. I was going to control it by the season. Now I'll just unplug the water heater and maybe drain the pipe. Dunno.

I have looked at NuHeat. It seems to have two disadvantages. One, it is easily damaged during tile installation. I've had enough problems with tile installation to be leery of creating opportunities for more. Two, it's electric.

I agree on both counts. It is a pretty easy "aftermarket" system, though; it would have eliminated half the subfloor replacement and all of the piping, so I can see how it would save a contractor a lot of time. (The "electric" doesn't matter for me, we're "all electric" around here.)
Copyright 1996-2014 trademark and the "Fool" logo is a trademark of The Motley Fool, Inc. Contact Us