Steve Shantz wrote:
I'd like to do the same later (the staple up). Can you provide some
details about what size, thickness and number of tubes? Here in Atlanta,
I was thinking of one tube per joist section and running the aluminum
nearly from joist to joist (~14").
Great people to work with. They have about anything you want
I run two 1/2" tubes per joist (which are on 16" centers). The sheets
are 16" x 8" x .012", 3003-0 soft aluminum, and after forming, the
width decreases to about 14". I made my own jig to form the sheets. I
just insert a sheet into my jig, place the forming piece on top, and
smack it twice with a mallet. (Put about 25 cents in my pocket for each
sheet! Great therapy.)
I ran my sheets continuous down the joist, with no gaps between sheets.
This scheme uses more plates, costing more, but it gives much better
heat exchange to the floor. (I would relish feedback on the merits of
this idea. I'm a strong advocate of well sized heat exchangers, in
solar DHW systems as well as radiant applications.) I like the idea of
spending more up front, and saving in the long term because of the more
efficient heat exchange allowing cooler water temperatures and higher
They supply this material and size to distributors, so it is readily
Steve Shantz wrote:
I like the wider plate.
Have you done any temperature measurements at the plate edge and then
near the tube, and then perhaps at the joist? I wonder what the fin
efficiency is for this.
The cooler temp seems right to me all around. Better efficiency and
more storeage for the same size tank as you can take the storeage to a
I'm assuming you have tempering valves in the radiant loop. Where did
you get these and do you have any recomendations on whatever else is
needed? Some of the diagrams I've seen have all kinds of optional one
way valves and other whatnots. I've got a stack of PDFs I guess I need
I haven't done any temperature measurements, but I'm not worrying about
it. (Believe me, this is unusual, as I like to measure all sorts of
unusual stuff). The reason for my lack of worry is that aluminum is a
much much better conductor than the wood floor above it. 250 W/m C,
vs. ~0.13 for wood. That makes it almost 2000 times better, given
equal thickness. If one runs two tubes per joist, then the heat must
travel out about 4" to the edges of the aluminum plates. This 4" is the
effective 'thickness' of the aluminum. So if the wood floor is 3/4"
thick, and the aluminum effective thickness to the edge of the plates
is 4", the aluminum is still 37.5 times better than the wood at
conducting heat (if I understand the calculations correctly!) Even at
0.012" thick, the aluminum is a way better conductor than the wood, and
explains my observation that it stays quite hot out to the edges. It
is a detail that I personally don't think is worth worrying about.
Increasing the thickness of the plates will give higher cost, more
difficulty working with and installing the plates, and a marginal
increase in performance. Better to spend the mony running continuous
aluminum so that the entire surface of the floor is heated, and not
just 1/2 or 1/4.
I should say that I live in northern Indiana, where things can get
quite cold. In Atlanta, you may be able to cut back to 1/2 or even 1/3
coverage, as your heat load is not going to be nearly as high.
Actually, I don't have a tempering valve in the radiant loop. We have
an older boiler (about 23 years old, and slated for replacement soon),
and the gas valve is controlled directly by the thermostat. A 55 watt
circulating pump is always on. This method seems to work quite well.
As it gets colder outside, the house temperature drops, and the
thermostat responds by leaving the gas valve on longer. The end result
is that the water temperature naturally rises as the heat load on the
house increases. I'm sure the newer control loops would reduce some of
the overshoot I sometimes see, but overall, the scheme works quite
well. I've never seen this control scenario anywhere on the web.
To prevent overheating and damage to my PEX, I have a thermostat on
the hot water (discharge side) to limit the water temp to 140 F, and
the high limit switch on the boiler is set to 145. The room
temperature thermostat and the two boiler thermostats are all wired in
series, so if any one opens, the gas valve shuts off.
I seem to see a lot of new boiler control loops that rely only on the
boiler temperature controller to keep things in check. See
http://www.ecrinternational.com/secure/upload/document/403.pdf for an
Tempering valves are a mixed blessing. They do provide a degree of
protection, but at a significant increase in cost and complexity, and
seem to require water hotter than that actually used for heating,
resulting in decreased efficiency of the boiler. Additionally, I am
not too happy with their reliability, nor with their ability to respond
accurately and quickly to over-temperature conditions. I have
monitored the performance of some brand new 2" tempering valves at the
college where I work, and the temperature swings are typically +- 10 -
12 F. A valve set at 125 can pass water at 135 - 137, and limit it
to 112 - 115. Temperatures swing between these extremes. This doesn't
seem to me to offer much protection It would be interesting to hear
others experiences with tempering valves used for this purpose!