re: solar barn update
28 aug 2004
gary wrote re:
>>>i have a simulation that includes the house (modeled as equivalent to
>>>6k lb of water), and the garage door collectors (with no storage). i
>>>use the tmy2 weather data for billings to drive the simulation...i let
>>>the garage door collector put heat into the house during the day until
>>>the house temperature gets up to 75f. once the house has reached 75f,
>>>i dump any additional heat from the garage door collector -- this
>>>happens fairly often in the shoulder months.
>> it might be useful to let some thermal mass heat to a higher temp,
>> after the house is warm enough.
>the easiest place in the house that i could use ceiling mass is the
>garage. the ceiling is high, and there are no aesthetic issues.
>this might eliminate the dumping of excess heat collected in the shoulder
>months by allowing it to be used in the evening.
or reduce it, or postpone it until warmer times of the year.
>the garage is insulated to r19, and two of the walls plus the
>ceiling adjoin the house -- so, not much of the ceiling mass heat
>would be lost to outside.
you might add more ceiling insulation...
>is there any easy way to bypass the ceiling mass on those days when
>you would like all the collected garage heat to go into the house
if you circulate air between the house and the garage with a big fan,
the garage ceiling won't warm to much more than 75 f until you turn off
the fan, but that still wastes some heat, if you don't need garage heat...
it's better to put thermal mass in the house, in the ceiling if possible.
how about some bottle gabions? :-)
>>>at night i let the house temperature cool to 60f before turning the
>>>furnace on -- this means i am getting 6000lb * 1btu/lb-f *(75f - 60f)
>>>= 90k btu out of house as thermal storage?
>> sure, altho it would last longer if you could keep the house at
>> a constant 60 f all night. you can do this with ceiling mass...
>>>this is good for about 6 hrs -- 30f + (75f - 30f)e^(-6/16) = 61f?
>> yes, with no internal heat gain. what's your monthly electrical use?
>i was a bit surprised how much electricity we are using -- its about
>1000kwh/mo during the winter, or 3.4 million btu/month, or 114k btu per day.
i've read that an average us home uses 10k kwh/year.
>>>(i plan to check the 6k house mass (which i have no real justification
>>>for :-) by measuring actual cool down rate and comparing it to
>>>predicted with 6k mass and 400 btu/f-hr loss rate -- this should work?)
>> ...including internal electrical use. why 400 btu/h-f? you might check that
>> with a kwh meter and an electric heater and thermometers on a mild day.
>> and try to estimate the house mass from the materials, eg 1/2 btu/f for
>> a square foot of 1/2" drywall.
>the 400 btu/h-f is the calculated heat loss based on r values etc.
>so, my thought on using this to estimate equivalent house mass is:
> (400 btu/hr-f)(thouse-tamb) = w*dthouse*cwater
> ie the loss out of the house will cause house mass to
> cool by dthouse
> so, if over a 1 hr period, thouse went from (say) 70 to 65f,
> and tamb is 30f
> w = 400*(67.5f - 30f)/(5f)*1 = 3000lb of water equivalent??
looks ok. you might estimate the house thermal conductance this way: shade
all the windows on the outside, then read the kwh meter on the wall, then
turn off the propane furnace when the house air temp is 70 f, then turn on
a few electric space heaters with thermostats set for 70 f, then record the
indoor and outdoor temps for a week (the indoor temp should be constant),
then read the kwh meter again.
>i'll have a go at the mass calc:
> count all wall and ceiling sheet rock?
> count wood flooring?
> count stuff under carpets?
> count wood ceiling?
> count first inch(?) of stone fireplace?
>only count things within reasonable distance of where the solar heated
>air will be introduced into the house?
i'd count everything, but realize that the deeper fireplace stone will have
a longer time constant, ie it will only contribute (sink) heat after a few
days with a lower (higher) house air temp.
with an accurate conductance estimate, you might estimate capacitance by
letting the house cool for a few hours at night with no indoor electrical
use. if g = 400 and c = 6k, rc = c/g = 15 hours. in 3 hours, when it's 30 f
outdoors, the house would cool from 70 f to 30+(70-30)e^(-3/15) = 62.7.
if it cooled from 70 to (say) 59, rc = -3/ln((50-30)/(70-30)) = 9.33 h,
so c = c/gxg = rcxg = 9.33x400 = 3732 btu/f.
>>>i can add about 180 ft^2 more collector area to the south wall of the
>>>house, but it seems like there is not a lot of point to doing this
>>>without adding some new thermal storage that would allow the collected
>>>heat to be held over for the night.
>> more heat dumping...
>not sure what you mean here --
just agreeing with you. without more storage, you'd be dumping more heat.
>...10/1 is quite a bit -- the test collector with the "bad" one gallon
>jugs is only about 2.3/1.
>if i were to use 8 ft high vertical pipes in my 4ft by 8ft test
>collector for storage, in order to get 320ft^2 of surface area and
>have about 40 gallons of storage, i would need to use about 200 3/4
>inch dia pipes -- thats a whole lot of pipes :-). smaller numbers of
>larger dia pipe end up having too large a ratio of internal volume to
how about larger numbers of larger diameter pipe? we need lots of surface for
overnight heat storage with a high heat transfer rate, and lots of volume for
cloudy-day heat storage with a lower heat transfer rate.
>is there a better type of container to use?
i tend to like ceiling mass for overnight heat and larger containers for
cloudy days. if the cloudy-day store provides no heat on an average day,
it doesn't need much glazing to stay hot.
>i guess it must also depend on the air velocity through the storage
>area as well as the area ratio? does 10/1 go with a low velocity air
yes. but more air velocity tends to mean more electrical fan power.
>> maybe it's time to try to convince your wife to let you push pieces of
>> foamboard into the windows in wintertime.
>we are planning to work on the window loss (and also insulation,
>sealing, ...), but some of that is already factored in the 400 btu/hr-f.
>> or do a blower door test/
>i am working on getting the local utility to do a "free" blower door test.
>> or do a blower door test/
>>improvement program with a large exhaust fan in one window and your new
>>velocity stick in a slot in another window in the same room, for starters.
>>as you seal up the house, the slot air velocity should increase.
>i had a go at this.
it looks like you did something different... i suggested turning on the fan
and closing all the house windows tightly except one, which would be closed
almost all the way, except for a 1/2" slot, with the velocity stick in the
slot to measure the air velocity in the slot. you might stuff up the rest
of the slot with a rag. this doesn't give you an accurate air-leakage
measurement, but it allows you to go round the house and feel for incoming
drafts and fix them and watch the slot air velocity rise as you fix them.
in the beginning, you might have the fan and the slot in the same room.
then move the slot to a point far away from the fan, and open and close
doors to various rooms to find and fix important air leaks.
>i have a whole house fan, and used this to exhaust air -- i measured
>velocity in at the window, and the velocity out at the fan (window
>closed and open) -- results:
>fan ------------------- window ------------------ dstatic pres
>low 550fpm 2670cfm closed na na --
>low 650fpm 3180cfm open 410fpm 2150cfm --
>high 800fpm 3920cfm closed na na 0.13 inwater
>high 1050fpm 5145cfm open 550fpm 2890cfm 0.02??
>the dstatic pres is the difference in pressure from inside the house
>to outside with the fan running -- measured with a water manometer --
>not very accurate (but repeatable).
the water manometer might help later. or a $50 magnehelic 0-0.25" meter.
the velocity stick numbers bounce a bit. averaging 5 samples might help.
>the fan is 30 in dia, and window opening 24 x 31.5 inches.
i suppose you used the stick to find its cfm.
>if i take the case of fan on high with window closed, and try to
>correct up to the standard blower door delta p of 50 pascals
>(0.2inches water), and if the leak rate is linear with pressure(?), then:
>estimated blower door flow at 50 pascals = (0.2/0.13)(3920cfm) = 6030cfm
>if i then try to use the "standard" correction to estimate actual air
>changes from blower door tests:
> natural air changes = 6030cfm *60 /(vol * h*c*s)
> c = "climate correction" = 18.5 for mt
> h = house height correction = 0.8 for 2 floors
> s = isolated house correction = 0.9
> vol = house volume -- aprox 35k ft^3
> nac = 6030*60/(35000*18.5*0.8*0.9) = 0.78 changes per hour
> they put this in the moderate category.
>the heat loss on a 30f day might be:
>q = (35000 ft^3)(0.78chng/hr)(70f -30f) (0.066lb/ft^3) (0.24btu/lb-f)
> = 17k btu/hr
>this seems too high to believe? maybe the procedure is just not
>accurate enough, or i am calculating something incorrectly?
the water manometer may be a problem...
>>if 1910 btu falls on 1 ft^2 of r1 south glazing with 90% solar transmission
>>over 6 hours on a clear 27 f jan day and the average collector temp is 100 f,
>>the net gain would be something like 0.9x1910-6h(100-27)/r1 = 1281 btu...
>this implies an efficiency of 1281/1910 = 67% -- is it realistic to
>think that you can get 67% of sunshine into actually heating the water?
maybe not. then again, if the average collector temp is 100 f,
the water is probably cooler.
>are there other losses that add up in turning the hot air from
>the collector into stored hot water?
loss through glazing, loss through walls, air leakage, a temp rise that
increases with solar power and decreases with cfm, a temp drop from air
to water at the container surface...
>anyway, maybe i should aim for more than 13 lb of mass per sqft of
more is better?
>>>i am also thinking about using the same system for domestic water
>>>preheat, since this would allow the collector to "earn its keep"
>>>during the summer (at the cost of some winter house heating performance).
>> you might preheat dhw in 5 btu/h-f-ft fin-tube pipe.
>i'll take a look at this -- this is as described in the solar closet paper?
yes, altho it might be easier to run fin tube under a roof ridge.
>i like the idea of ceiling mass, but its difficult for my house.
>a lot of the ceiling is "cathedral" type.
fill that space with lots of big pipes. paint them bright red. a mobile. art.
>it would be nice to have a math model for the whole collector/storage
>unit, so that i could play around with channel sizes, area ratios,
>fans, ... on paper, and then try a few on the actual test setup.
back of an envelope?
>one thing i am wondering about is, if i use a fan (or two), how much
>baffling, ducting etc do i need to get fairly uniform velocity
>distribution across the storage containers?
that seems figure-outable. equal length "ducts" with equal cross sections?
>i take it that there are people who have had good experience with
>ceiling mass houses?
the (italian) horazio barra system is described on pages 169-171 and 181
of baruch givoni's climate considerations book (wiley, 1998.) the basic
reference is barra, o. a., g. artese, l. franceschi, r. k. joels and a.
nicoletti. 1987. "the barra thermosyphon air system: residential and
agricultural applications in italy, uk, and in the sahara." international
conference of building energy management. lausanne, switzerland.
barra's are said to be a lot more comfortable than other passive solar houses,
with more uniform north-south temp distribution. his "spancrete" ceiling slabs
in single and multistory buildings let hot air-heater air thermosyphon through
the slab tunnels from south to north, where it exits and travels back north
through the bulk of the room to the air heater inlet near the floor. no fans,
and no selective surface beneath, but the hot air store lots of heat in the
slabs. lots of successful systems were built in europe, but barra seems fairly
unknown in the us.
>is it correct to say that a successful ceiling mass system system
>would have these features?
>1) the solar heated air is delivered to the ceiling area?
hot air rises and pools under the ceiling (in my book.)
>2) there is a good heat transfer path from the heated air to the mass?
a 10:1 glazing surface to ceiling surface ratio seems good.
>3) the rate of transfer from the mass to the room is regulated by
>choosing material with the right emissivity?
that, and maybe a slow ceiling fan attached to a room temp thermostat.
steve baer hates fans. he uses movable aluminum louvers to modulate
ceiling radiation, as in a satellite "deep space cooler."