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

A bang-bang house



I've been thinking more about thermal storage for heating and cooling...

For many years now, people like Steve Pitney have been building thermal
storage tanks for off-peak electric heating out of plywood boxes lined with
EPDM rubber, which costs about 28 cents/ft^2 and comes in rolls up to 20'
wide and 100' long. It takes 2 or 3 people to lift a roll like that.

                                                       28'
For a new house, we might have         pppppp.......................
one or two EPDM rubber-lined plywood         gi i i                .
boxes in the basement, say 12' long by       gi B i                .
7' high x 2-4' thick, on top of some         gi B i                .
railroad ties on 4' centers sitting on       gi B i                .
grade. The plywood or OSB could go on        gi B i                .
top of the ties, and the tanks would    <- S gi B i                .
run east and west inside the south wall      gi B i                .
of the basement, with an 8" air gap          d  d d <- glass doors .  32'
underneath. There might be a single          gi B i                .
cement block on grade in the center of    -->gi B i<-- 4-8'        .
each 4' square to help support the plywood.  gi B i                .
There would be a few small fans at the       gi B i  ssssssssssss  .
top to circulate air from south to north     gi B i  s  septic  s  .
over the top of the boxes, down from         gi B i  s   tank   s  .
the top of the box to the bottom on          gi i i  ssssssssssss  .
the north side, under the boxes, back        .....................
up the south side between the south          p            p
glazing and some insulation, and around      p            p
again. The boxes might be framed by          p  duckweed  p
through-bolting a couple of vertical         p     or     p
2x4s onto each end of each railroad tie,     p  reed bed  p
and screwing the plywood to those with       p            p
a plywood nailer, with drywall screws,       p            p
with a longitudinal 2x4 along each     pppppppppppppppppppp
inside corner at the bottom.
The EPDM rubber would be in one big sheet, folded up like a Chinese takeout
box, with no seams. The box would weigh 200-300 psf on the bottom, with less
force on the sides.

As an alternative, these two thermal masses in the basement might be some
2' diameter x 3' tall 55 gallon drums full of water stacked up 4 high,
horizontally, to make two walls 3' thick and 12' long and 7.19' high, each
containing 22 drums, each weighing about 10,000 pounds, with the drums resting
on grade, and on each other in a hexagonal array, with a wood frame to contain
them. This would be have more surface area than the plywood tank, and leaks
would be less worrysome, and it could be easily dismantled...

This solar closet might be a good place to put a clothesline or a hot tub
or a small woodstove or some fin tube pipe near the ceiling to heat water
in a conventional water heater on the floor above.

(i is insulation above. B are the boxes. g is glazing. p is a shallow solar
reflecting/wastewater treatment pool, ~ 1' deep x 16' wide x 64' long, divided
in 4 16' x 16' sections, wrapped around the house. The pool would reflect
about 60% of the sun onto the house with a layer of ice on top, and about
6% with no ice. The ice thickness would be controlled by the septic tank
effluent temperature. The septic tank might be cooled by a carefully-
designed DHW heat exchanger or some outside ventilation air.)

The basement could get some nice light from the floor above...

I'm just starting to look at cooling season Philadelphia TMY2 hourly weather
data from Eric Swanson.

I'd like to do a little passive cooling simulation, ventilating a house well
at night to store coolth in its thermal mass, and reducing the ventilation
during the warmer part of the day.

Some interesting daily highs and lows in the cooling season here are:

day  high  low

141  85 F  58 F        In this Typical Meteorological Year, the hottest days
157  89    61          still have coolish nights, especially if one slides
176  90    72          a 2 or 3 day window over the data, consistent with
181  93    71          storing coolth somewhere in the structure. I think
188  91    72          this is now done in commercial buildings, eg in
190  92    75          elevator shafts, but they probably chill the air
210  90    74          mechanically rather than just ventilating with cooler
226  93    73          night air. A good passive cooling system would clamp
254  86    66          daytime house temps to the night lows, for starters.

If the solar closet is divided in two halves, say on either side of a
walk-out basement, it seems to me that a house above could have excellent
temperature control, even if it had little thermal mass. One half of the
closet could store coolth and the other store heat over multiple days
in the spring and fall, in which case there would be a cool source and
a warm source available to drive the house temp actively in either direction,
as in Norman Saunders houses, vs typical solar houses which have at most one
driver. Norman says a well-designed house is as likely to need cooling as
heating on any particular day, even in wintertime...  Another option is
ground-coupled cooling using the basement floor, with a vapor barrier
underneath, and a small dehumidifier near a floor drain. Air movement might
come from the stack effect and/or using the sunspace as a solar chimney.

Another option is to use reflecting pond water for cooling, eg with a
fan-coil unit or duct heat exchanger or interior fountain.

This data is for a typical year, a mild year, with no temperatures over 100
or less than 0. We hope to get 30 year's worth of hourly data off the NOAA/
NREL Samson CD-ROM soon. Meanwhile, we have these 8760 lines of data to start
with :-)

Date: Mon, 18 Mar 1996 19:53:38 -0800
To: nick@ece.vill.edu (Nick Pine)
From: eswanson@atlcom.net (R. Eric Swanson)
Subject: RE: TMY2 Data for Philadelphia

TMY2 Solar Data for Philadelphia, months 1-12
*********************************************
    1st col = Day of year, 1-365
    2nd col = Hour of day
    3rd col = Global horizontal radiation during preceeding hour, Whr/m2
    4th col = Direct normal radiation during preceeding hour, Whr/m2
    5th col = Diffuse horizontal radiation during preceeding hour, Whr/m2
    6th col = Temperature, deg F
    7th col = Dew point Temperature, deg F
    8th col = Relative Humidity, %
******************************************************
  1   1     0     0     0  34  31  88
  1   2     0     0     0  34  31  89
  1   3     0     0     0  34  29  82
  1   4     0     0     0  32  20  61
  1   5     0     0     0  28  16  60
  1   6     0     0     0  25   8  48
  1   7     0     0     0  22  10  60
  1   8    25   143    12  16  -1  46
  1   9   131   588    31  14  -5  42
  1  10   279   779    43  15  -5  40
  1  11   264   258   161  15  -5  40
  1  12   209    70   178  15  -4  42
  1  13   345   408   161  17  -5  37
  1  14   315   425   142  17  -3  41
  1  15   203   335    97  16  -2  43
  1  16   141   511    45  15  -4  42
  1  17    32   171    17  12  -7  42
  1  18     0     0     0  11  -8  41
  1  19     0     0     0  11  -8  41
  1  20     0     0     0  11  -6  46
  1  21     0     0     0  11  -5  48
  1  22     0     0     0  11  -4  50
  1  23     0     0     0  12  -3  50
  1  24     0     0     0  13  -1  53
  2   1     0     0     0  14   1  56
  2   2     0     0     0  16   1  51
  2   3     0     0     0  16   2  53
  2   4     0     0     0  17   2  51
  2   5     0     0     0  17   2  51
  2   6     0     0     0  17   2  51
  2   7     0     0     0  17   2  51
  2   8    19    67    13  17   3  54
  2   9   119   450    42  19   4  51
  2  10   259   659    60  21   4  47
  2  11   270   245   172  24   6  46
  2  12   183    71   151  23   5  45
  2  13   253   140   190  25   7  46
  2  14   364   647   100  27   4  37
  2  15   281   686    62  27   5  38
  2  16   139   493    45  26   3  37
  2  17    29   110    19  25   3  38
  2  18     0     0     0  24   4  42
  2  19     0     0     0  21   4  47
  2  20     0     0     0  21   4  47
  2  21     0     0     0  21   4  47
  2  22     0     0     0  21   4  47
  2  23     0     0     0  20   4  49
  2  24     0     0     0  19   4  51
  3   1     0     0     0  19   4  52
  3   2     0     0     0  19   5  54
  ...

Suppose this new house is built into the hill h to the north, with a shallow
monopitched roof sloping up to the south, like this:

(Connect the two "r" dots at the top         r
to make the roof.)                           gi                   rttttt
                                          .  gi         .         it   t h
This house might need 50K Btu/day for        gi         .         ittttt
20 days for additional summer cooling   .  s gi         ..........i  h
beyond what is passively available on      s gi                   i
summer nights, ie 1 million Btu.      ? 24'  gi                 h i
                                             gi               h   i
Suppose there is a plywood/EPDM tank t       giii...........h.....i
sitting on the ground as shown above,      s giBi         h       i
full of ice...                    .        c giBi       h         i
                                             giBi     h           i
                                 pppppppppppppiii...h.............i

The tank might be 8' wide x 16' long x 6' deep, built like the basement tanks,
with pressure-treated 2 x 4's and exterior plywood and 4" of foam on the
outside. This would give it a surface area of about 500 ft^2, so over 6 months
it might lose about 180days*24hr*(62-32)*500 ft^2/R20 or 3.2 million Btu of
the 8*16*6*62*144 = 6.8 millon Btu stored in the ice, leaving 3.6 million Btu
available for cooling the house (so the tank could probably be smaller, eg 12'
or 8' long), by letting some cool air flow down into the house, or using a
chilled water loop.

One might somehow tuck a root cellar in or near the north wall, and it would
be a good place to put or make a refrigerator, and it could dehumidify
the house.

The box would also make a nice gravity-feed 6000 gallon rainwater supply.

The roof might be covered with two pieces of EPDM rubber with a standing seam
in the middle, with 9 1" PVC draindown pipes on 4' centers with holes in them
running from north to south up the roof, to act as sprinklers in winter, to
make ice and warm the roof when it's -10 F outside. On a cold cloudless night
with no wind, the roof may be 10-20 F cooler than the outdoor air, because of
radiation to the night sky. the A small submersible pump and a header at the
bottom of the tank would move the water, which might supercool on the roof
and drain back down the roof and the tank cover into the tank and freeze on
the tank ice surface.

The roof and tank would have about 1000 ft^2 of surface area, so at an
average nightime outdoor temp of 22 F, freezing the tank would take roughly
6.8 million Btu/(32-22F)/1000 ft^2*R0.7 = 500 hours, ie 21 days. A 100 watt
pump running 500 hours would use $5 worth of electricity per year.

Nick