re: above ground passive cold house
15 dec 2001
david delaney wrote:
>the design problem: to build a cold house the interior
>of which will have a temperature between 1c and 15c (2f
>and 60f) all year. the locale has 4700c (8500f) heating
>degree days per year. average january temperature is
>-11c (12f), often reaching -18c (0f). average july
>temperature is 21c (72f), reaching 35c (95f) several
>days a year. the ground is very wet and poorly drained,
>so i do not want to build below grade. i do not want to
>use any active elements at all in the design.
no active elements. that's tough. keeping the inside 34 f min
when it's 0 f outdoors makes a 34 f difference, and 60 f max
when it's 95 f outdoors makes a 35 f difference. this sounds
like a simple dc circuit with some resistors.
>in this locale, the ground temperature three or four feet below
>the surface remains at 10c (50f) year round...
and the indoor temp swings 16 f below and 10 f above that over the year.
>the design i like at the moment is as follows. a concrete floor slab
you might consider above grade, so condensation has somewhere to go.
pile up 1' of earth from the excavation for footings and put the slab
over that. or put it on concrete block piers, over a 1' crawlspace,
with a vapor barrier on the ground, and use the walls vs floor to
conduct coolth up from the ground. you might want to make the piers
a lot deeper, to get in touch with more cool ground.
you might add some (passive) automatic foundation vents to let outdoor
cold air flow through the crawlspace when it is available. you know these
things, with bimetallic springs that operate louvers? about 8"x8"x16".
home depot sells leslie-locke afv-1b vents for about $12. you can rotate
the springs on their mounts to vary the opening temperature, or remove and
reinstall them backwards so the louvers open on falling vs rising temps.
you might use 2 vents in series, one that opens when outdoor air is less
than 40 f, and one that opens when indoor air is greater than 40 f.
you might also look at john hait's book. a box with thermal mass inside
and insulation outside will tend to be cooler than the outdoors if it has
holes at the top, so warm indoor air can escape to the outdoors and be
replaced by outdoor air when it is cooler outdoors. the holes at the top
might be fitted with automatic foundation vents to stop this process when
it is less than 40 f indoors.
>vapor/moisture barrier under the slab, but no insulation, because
>i want the earth below to serve as a heat sink in summer and
>a heat source in winter.
it's easier for the ground to provide heat in the winter than coolth
in the summer, since warm air rises, and because upwards heatflow in
soil has another mechanism beyond conduction to help: evaporation of
water from lower layers, migration of vapor to upper layers, and
condensation in upper layers. you seem to be aiming at a system in
which the only heat store is the earth, vs a building with containers
of ice that freeze in the winter and melt in the summer, and so on.
we might assume the ground temp under the building never changes
over the year...
>concrete footings descend from the edge of the slab four or five feet
>into the ground. a short or tall concrete wall ascends from the edge
>of the slab to support or form the structural wall of the cold house.
>the ouside of the footings, the outer edge of the slab and the outside
>of the ascending wall are well insulated (and the roof of course).
you might use thick and tall vs thin and short concrete walls to conduct
the ground coolth to the top of the space, since warm air rises...
>i will design the insulation of the walls, footings, and
>roof by using the maximum r values required to satisfy
>both of two cases: 1) a winter case at an interior air
>temperature of 2c (36f) and an exterior air temperature
>of -20c (-5f). 2) a summer case at an inteior air temperature
>of 13c (56f), and an exterior air temperature of 35c (95f).
a little tighter than your initial specs.
>if this generates ridiculously thick insulation, i will have to
>relax the requirements, do a more sophisticated analysis, or give up.
the required insulation thickness depends on the concrete wall
thickness and the wall and slab area in contact with the cool
ground and the ground conductance...
>i think i know how to write approximate terms of the
>necessary heat balance equations for the roof, walls,
>slab edge, and footings, but i havn't a clue when it
>comes to estimating the heat flow into and out of the
>earth through the floor slab (which just happens to be
>the most critical term of each equation:-)
table 7 on page 22.13 of the 1993 ashrae hof:
"typical apparent thermal conductivity values for soils, btu-in/h-ft^2-f"
recommended values for design (a)
normal range low (b) high (c)
sands 4.2 to 17.4 5.4 15.6
silts 6 to 17.4 11.4 15.6
clays 6 to 11.4 7.8 10.8
loams 6 to 17.4 6.6 15.6
(a) reasonable values for use when no site- or soil-specific data are
(b) moderately conservative values for minimum heat loss through soil
(eg use in soil heat exchanger or earth-contact cooling calculations.)
values are from salomone and marlowe (1979)
(c) moderately conservative values for maximum heat loss through soil
(eg use in peak winter heat loss calculations.) values are from salomone
and marlowe (1979)
i'd pick the "low" column, or some higher value than that, since your
ground is wet. eyeballing fig 6, a graph of apparent moisture content vs
conductivity, we might pick 12 btu-in/h-f-ft^2, ie r1 per foot, altho
it seems that would only matter for upward heatflow...
most of the soil's thermal resistance occurs in the first foot of depth,
but there is more in the second foot, which has more cross sectional area,
and so on, expanding in shells, in something like an underground cup or
hemisphere, concave side up.
if an 8' cube with very thick concrete walls and 4' footings and 64 ft^2
in touch with 50 f soil and thermal conductance 320/rv to outdoor air
were 56 f inside on a 95 f day, we might have something like this:
i 56 f
50 f -----www-----*-----www-----95 f
i = (56-50)64 = 384 btu/h, so (95-56)320/rv = 384, and rv = 32.5, eg
6.5" of styrofoam. might work...
i'd be tempted to dig 2 shallow wells with an auger, line them with 2
4" perforated pvc pipes, and use a 4 watt pump to move some groundwater
up through a pond under the ceiling, or through some vertical pipes in
or near the walls. you might keep the wells and riser pipes filled with
a float valve from the house water supply.