[Prev][Next][Index][Thread]

Norman Saunders on Heat Flow Within a House


93420 Heat flow within a house. 93202

   Copyright Norman B. Saunders, P. E. 15 Ellis Road, M3, Weston MA 02193
             may be freely quoted if in context with appropriate credit.

This topic has come up again with respect to the same real house used in:
93202 Solar Heating, Boston to Chicago.

First the heat flow in the ground around the house. The seasonal
temperature change at the surface diffuses downward, being delayed so
that at a 2m or so depth it is 6 months behind, and the temperature
change is also attenuated so that below 3m it is negligible.

All soil contains some moisture. When heat is moving downward, the
conductivity of the water speeds the temperature change slightly, even
as water's thermal capacitance slows it. A warm wet strata below is
another matter. Vapor is always forming, and because of its low density,
it rises through the inevitable pores, until the lower temperature causes
it to condense, releasing the considerable heat from change of state.
See Penrod, "Measurement of the Thermal Diffusivity of a Soil by the
Use of a Heat Pump," J. App. Phys. 21 May 1950.

Within the house itself, interior partitions run about R3, whereas the
external walls tend toward R30. Take the extreme case of a square house
of 9 rooms, each 12 x 12' with 0F outside, 68F in the central room,
and all air flow blocked, and consider heat flow only through walls:
the central side rooms would come to about 53F and the corner rooms
to about 48F. Practically, there is air flow between rooms and heat passed
between them through floor and ceiling as well.

A room most likely to need supplemental heat is the isolated corner room,
receiving heat through 2 walls and losing through 2, and hence at 61F,
reduced by window losses. So you send warm air to this den through 4 x 6"
oval duct with return under the door. Most interior doors have 3/4" cut
off the bottom to allow clearance for rugs, giving an open area larger
than such a duct.

The major need in winter is to move heat downward. The convector or
radiator warms air, which rises to, spreads across, and radiates heat
downward from the ceiling. It is much simpler to simply spread the
warm air across the ceiling. Of course the warmed ceiling allows heat
to flow upward; even a carpeted floor is likely only R2, so the room
above gets heated too. That brings up another point. ASHRAE et al,
give 6W/m^2 for the convective coefficient up from such as floor and 2
for the coefficient down from ceilings. The 6 is substantially correct.
The 2 is a crude approximation. Air flow removes heat from ceilings only
in so far as there are temperature differences between walls at the same
elevation that cause the air to circulate. The uniformity of joist
temperatures within a floor likely raises the thermal resistance of
downward heat flow by 0.1 m^2/W or about 0.5 R, relative to upward.

Yours,

Norman B. Saunders, P. E.

SSS                 Experimental Manor
Basics              15 Ellis
Glazing Rails       Sunshine Circle
North Window (TM)   Weston, MA 02193
Solar Staircase (TM)
Saunders Solar Systems