re: parabolic tracking collectors
24 mar 2001
anthony matonak wrote:
>sheila havard wrote:
>> i have a lot more time than money so i am trying to design and build
>> a tracking parabolic collector to heat water.
sheila's posting never appeared on my newsfeed.
>hmm. first place to start thinking. do you need a tracking parabolic
>collector to heat water?
well, mirrors are cheaper than flat plate collectors, and a smaller
collector loses less heat, and a higher collecting temp decreases the
heat storage volume required, but "tracking" seems like overkill.
>> 1) are there any free (or cheap) designs available?
i've posted some untested "12' non-cubes" here...
>> 2) is it practical to store heat from the summer in water to be used
>> in the winter to take me past extended cloudy periods?
>i don't know about this. my gut feeling is 'no, it's not practical' simply
>because trying to keep heat for even a few weeks, let alone 6 months, is
>a difficult proposition and would require massive structures.
massive, but maybe not all that expensive. you might be happy to provide
hot water for a cloudy month, vs a whole winter: 30dx50kbtu/d = 1.5 mbtu,
like 1.5m/50 = 30k pounds or 3750 gallons or 469 ft^3 or an 8' cube or a
12' diameter x 4' deep swimming pool (a plastic film liner inside 2 welded-
wire fence cylinders with some bags of leaves between them and more bags
floating on top) full of 180 f water cooling to 130.
"keeping heat" over time is less of a problem than providing hot water
for showers. a 12' swimming pool completely surrounded by r24 insulation
has rc = 30kxr24/377ft^2 = 1910 hours or 80 days, so it would cool from
180 to 120 f after -80ln((120-30)/(180-30)) = 41 30 f days.
rc = 12^3x64xr48/864ft^2 = 6144 hours or 256 days for a 12' r48 cube, so
it would cool to 120 f in -256ln((120-30)/(180-30)) = 154 days.
for truly seasonal heat storage with 20x less volume and no loss of heat
over time, you might look into a desiccant system, eg a licl still and
an evacuated groundwater heat exchanger.
>...mylar is more reflective than the aluminum flashing...
true... 90% vs 30%? and it's a lot more specular, ie focusing.
>> 4)how deadly accurate does the parabolic curve have to be?
you might make the smallest line segment 1/4 of the target size.
>> 5) is there any particular benefit to a wide shallow parabola
>> or a deeper, narrower one...
a deepish downwards-facing linear parabolic reflector optimized
for december with a slightly larger fixed target (vs tracking)
could also provide shelter for humans or horses or lawn mowers.