re: low temp radiators
2 dec 2002
>> warmboard is another alternative, thick plywood with a heat-spreading
>> layer of aluminum on top and grooves for hydronic tubing.
>this does not perform as well as covering the tubes in a cement screed.
i'll do some more checking. the warmboard site says its heatflow is 8.12
times better than gypsum concrete with tubing on 12" centers, and one of
their engineers at 877 338 5493 told me it can move 25 btu/h-ft^2 with
100 f water, vs 120 for aluminum heat emission plates and 160 for concrete,
citing the wirsbro cdam manual.
but page 279 of appendix f of the manual says tubing on 12" centers in a 4"
bare concrete slab only requires 90 f. maybe he meant to compare warmboard
with a "poured floor underlayment" with 125 f water. hydronic slabs also
have downward heat loss, moreso than fin tubes, i suppose.
>> a concrete hydronic floor might cost $5/ft^2 and need 160 f water for
>> 25 btu/h-ft^2. then again, it can provide desirable overnight heat storage.
>if you are buildng the floor anyhow the extra is minimal.
fin tube radiators seem cheaper and allow deeper temp setbacks at night, vs
mass floors. warmboard joints can't be staggered over joists for greater
strength, as one might do with 2 subfloor layers...
>where it scores well is that it can move solar gained heat from warm rooms
>to cooler rooms. this again requires the control system to be designed to
>suit. if there is useful solar heat the boiler or zone should be held off
>and zones opened up to take the warm solar gained water.
somehow, that doesn't seem worth the trouble. it would tend to happen
naturally even without the extra controls. physicist tim ellison intended
to build a house this way, with a radiant floor for heat collection and
distribution, but with a very small temp swing and low upper temp limit,
his 3 3000 gallon slab water tanks wouldn't store much heat.
>grainger sells 8' lengths of heatrim-american 3/4" fin tube pipe [including
>the sheet metal housing] for $52.27. they move 310 btu/h-f-ft with 140 f
>water at 1 gpm with a pressure drop of 0.047 "h2o/ft.
vs 0.457 "h2o/ft for 1/2" pex tubing.
>the thermal conductance is 310/(140f-70f) = 4.4 btu/h-f-ft...
>if it moves 310 btu/h-ft in slow-moving air, how much more can it move if
>it's mounted in an air inlet at the base of an 8' closet or wall cavity,
>with an equal-sized vent at the top?
if a linear foot in a 3" slot with 0.25 ft^2 of free area heats "convector
air" by dt f, cfm = 16.6x0.25sqrt(8'dt) = 11.74sqrt(dt), with velocity v
= 4cfm = 46.95sqrt(dt) fpm or 0.534sqrt(dt) mph, which makes the tube-to-air
conductance roughly g = 2.2(2+v/2) = 4.4+0.587sqrt(dt). if (140-70)g = cfmdt,
70(4.4+0.587sqrt(dt)) = 11.74sqrt(dt)dt, so dt = (26.2+3.5sqrt(dt)^0.6666.
plugging in dt = 10 f on the right makes dt = 11.16 on the left. iterating
makes dt = 11.28, then 11.29, so g = 4.4+0.587sqrt(11.29) = 6.37 btu/h-f,
and it might move (140-70)g = 446 btu/h, a 44% gain.
to make this work, we might install a fin-tube without the back plate, over
a 3" slot at the base of an internal wall, and leave the metal damper above
the tube closed. the fin-tube could also contribute some heat to the room on
the other side of the wall, and warm the wall for radiant heating. the fins
are only 2.125" square. we might put bare fin tube completely inside a wall,
with a 1.44" notch at the base of each 2x4 stud.
then again, why waste that internal wall cavity space by drywalling both
sides? rich komp uses it all for bookshelves, in his house...