|
|
re: cooling of pv modules (was calculating windload...)
3 jan 2003
anthony matonak wrote:
>> >if the main goal is to maximize electricity production and minimize
>> >cost then it seems to me that a home made tracker would give you the
>> >most bang for the buck.
>>
>> but the rebate program only pays for stc watts.
>
>if the goal is to maximize rebates...
we might think of the goals as providing a good tax-free investment for
the client and a good profit for the contractor. after collecting the
max $25k rebate, we might ask "will adding another $1 to the cost raise
the value to the client by more than $1?" if the answer is yes, we might
do it. if the $1 added cost increases the client's tax-free return ie
electrical production by more than $0.10 per year, it seems worth doing.
if the selling price for a 5 kw system is $31,250 (any less, and we run
into the 80% rule) and the client pays $6,250 and it makes electricity
worth $1,700 per year, that's a 27% tax-free roi for the client. if the
panels cost $17,325 and the rest of the system costs $5k, the contractor
makes $31,250-17325-5k = $8,925. if a shiny tarp and its support cost
$150 and they increase the value of the electricity produced by 10% to
$1,870, a rational client would be willing to pay $1870/0.27 = $6,926
for that system, $676 more than the system without the tarp. if that
happens, the contractor invests another $150 and collects $526 more...
if the client is buying a $42k system with a 10% tax-free return, the
10% increase in production is worth $170/0.1 = $1700 to the client, and
the contractor makes an additional $1550. if we only raise the price by
$1k, the client gets a 100x$1870/($43k-$25k) = 10.4% tax-free return.
these returns seem surer than stock market returns...
>then you would want the cheapest parts and least expensive install
>regardless of performance, reliability or anything else.
we can put a low price on reliability, if things rarely need replacing
and the house also has a grid supply.
>water cooling will not increase the stc watts either. in this sense,
>neither the water cooling or trackers will increase your rebate.
true, but it looks like cooling increases the client's value and
the contractor's profit.
>> 32'x14' = 448 ft^2 is a lot to move, with serious wind loading and
>> a serious foundation. i just saw a wattsun tracker with a much smaller
>> array mounted on a 6" steel pipe in a 6' deep x 3' diameter solid
>> reinforced concrete cylinder in the ground.
http://www.solar-trackers.com/
>i'm thinking the traxle design is one that lends itself to being home made.
>since it uses a two or three leg structure instead of a single pole in the
>ground, i would think it could be constructed with smaller steel and lesser
>foundations. i'm sure there are a great many other tracker designs out there
>as well. some of them are likely easier to fabricate from junk yard parts
>than others.
sounds promising for some people, but i'd rather do something simple and
easily reproducible with off-the-shelf parts than crawl around junkyards
seeking 50 one-of-a-kind solutions.
>> nrel says a fixed array with a 40 degree (latitude) tilt in phila gets
>> a yearly average of 4.5 kwh/m^2-day of sun. a 1-axis (n-s) tracker gets
>> 5.7, ie 27% more, comparable to a tarp and cooling, by my estimates.
>
>http://rredc.nrel.gov/solar/codes_algs/pvwatts/
>this indicates that a fixed 1 kw array at 34 degree (latitude) tilt for
>los angeles would produce about 1852 kwh/year vs 2343 kwh/year for a
>1-axis tracker or a bit more than 26% more. alternatively, folks in
>albuquerque, nm might expect 34% more by using a tracker.
sounds promising for people in la or albuquerque...
>if i recall correctly you were saying something like 11% more from the
>use of water cooling. it seems to me that i would get more energy out
>of an array by adding a tracker than by adding water cooling.
probably so.
>since the main thing is to get the best kwh/dollar here then i would think
>it would be very difficult to compare the two without estimates as to
>the costs of each (tracker, water cooling).
ok. i estimate water cooling costs $300, including labor. what's the cost
of a 5 kw tracker?
>> >it seems to me that a home made heat sink from scrap metal and parts,
>> >perhaps even with some kind of fan or blower, might be slightly less
>> >costly and a bit more reliable in the long term than water, pumps and
>> >poly film.
possibly. poly film is only a part of the picture if we use the waste heat.
>> ...try numbers. i can imagine "cheaper" (tho not much) but not better,
>> since water can move 4000x more heat than air, and its film conductance
>> is about 60 btu/h-f-ft^2 vs 0.67 for air, and it evaporates.
whoops. the still air film conductance is 1.5.
>i'm not sure if i'm reading your numbers and calculations correctly but
>let's see if i can work it out. the film conductance doesn't seem to be
>a big deal because we can change the surface area of our panels by adding
>a heat sink.
that costs money, and it's difficult to get panel heat into the heat sink.
>if water conducts heat 90 times better than air through a surface then we
>simply need 90 times the surface area to get the same heat transfer.
well, 40x better, and we may lose heat from both sides of a heat sink.
so, each 14 ft^2 sharp panel needs a 280 ft^2 dry heat sink... jeez.
what would it weigh? water also evaporates...
>certainly more surface area is going to mean more cost but with scrap metal
>it may not be a large cost.
all the parts need to be good thermal contact with each other. the parts near
the panel need a larger thermal conductance and a good thermal conductance to
the panel itself. would they somehow attach to the back of the panel? we don't
want to block the sun.
>certainly a metal heat sink should last many decades...
soytenly.
>while poly film pillows may only last a handful of years.
that's a different system, with usable valuable waste heat.
>the cost of the poly film would need to be multiplied by the number of times
>it would need to be replaced compared to the metal heat sink and to be fair
>some sort of labor cost should also be considered.
that's a different system, vs just trickling water over the panel faces.
>> a 12 watt pump can move 500 gph. a 400 watt blower might move 1000 cfm.
>
>http://mb-soft.com/juca/info/blower.html
>this moves 465 cfm with 115 watts or 4 cfm/watt.
better, but we left out the static pressure spec. a 100 w ceiling fan
can move 25k cfm...
>i would assume a similar fan that uses 400 watts might move 1600 cfm.
ok.
>...one might also need to factor in natural wind or convection currents
>in the air as well.
sure. they also help uncooled and water-cooled systems.
>if we were talking about a fixed array then perhaps a solar chimney might
>even be in order. this could move a lot of air without consuming any
>electricity at all.
got specifics?
>> with a 10 f temp diff, that's 500x8x10/12 = 3,333 btu/h-w for the pump vs
>> 1000x10/400 = 25 for the blower. we need to remove about 0.85x250x448ft^2
>> = 95k btu/h from a 5kw array with no concentration.
>so, let me see if i can figure out the units here.
>500 g/h * 8.34 lbs/g * 10degf / 12w = 3475 lbs * degf / h / w
>one pound raised one deg f takes one btu so that's something like
>3475 btu per hour per watt.
agreed...
>the blower then does 1000 ft^3/min * 10 degf / 400w = 25 ft^3 * degf/min/w
>maybe we should change minutes to hours. so that's now 1500 ft^3 * degf/h/w
>but does it take one btu to raise a cubic foot of air one degf?
no. a btu can raise 55 ft^3 of air 1 f. that's close to 60, so a c cfm
airflow with a d (f) temp diff can move about cd btu/h.
>http://www.kerrheating.com/online_training/homeheating/pdf/heat_transfer.pdf
>shows two simplified equations for water and air.
>water = 500 * g/m * degf = btu/h
>air = 1.08 * ft^3/m * degf = btu/h
ok. i tend to use 1 vs 1.08.
>let's try those calculations using these formulas.
>water, 500 * 8.3 g/m * 10 degf / 12w = 3458 btu/h/w
>air, 1.08 * 1000 ft^3/m * 10 degf / 400w = 27 btu/h/w
>or even, 1.08 * 1600 * 10 / 400 = 43 btu/h/w
ok.
>so, pumping water would move from 80 to 128 times more btu per watt than
>moving air. of course, air can't freeze, leak, corrode equipment, short
>wiring or any of the other wonderful things water is known to do...
evaporation comes to mind :-)
>and a certain amount of air will move all by itself without the blower.
nrel says the yearly average windspeed in phila is 9.5 mph. that would
make the airfilm conductance about 2+9.5/2 = 6.75 btu/h-f-ft^2... that
could reduce the 280 ft^2 heat sink to about 60 ft^2... this still seems
completely impractical, given the careful detail needed to get the heat
out of the panel and the thermal resistance through the back side.
>how happy are people going to be with something that breaks once a year
>or that they just leave broken because it's too much bother to fix?
not very happy. otoh, many people replace auto oil filters more often.
attwood estimates a 700 hour operating lifetime for their v1250 bilge
pump, based on their worst-case testing, pumping crud all day, clogging
with hairs, pumping salt water at high temperatures, running dry... the
engineer i spoke with says they will replace them without question if they
are less than 3 years old. he sent me a couple of samples and invited me
to abuse them until they failed. he also says people tend to get longer
lifetimes in non-bilge applications, eg continuous pond fountains with
cleaner fresh water...
>wouldn't this have some bearing on their perception of your work and
>solar pv in general? it might not be worth the bad impression to get an
>extra 11%. for instance, as a solar installer you might make more profit
>by selling two systems than by selling one system and having the second
>sale fall through because the first guy was unhappy.
expectations matter a lot. i'd be honest and open and offer choices.
>i suppose if you knew the pumps were going to fail that you could rig
>up multiple pumps such that when one pump fails it gets switched out
>automatically and the next one in line gets switched in.
high-rel industrial applications often use two pumps with alternating
service, so a failed pump can be detected and fixed before the other pump
fails... then again, the consequences of a cooling failure are minimal,
10% less output until the pump is replaced. a snow-melting failure would
be more disastrous. we might use 2 pumps, and scale back the cooling if
only one pump is working, preserving its lifetime for snow-melting, until
the failed pump is replaced.
>a manual method might be to use quick disconnect connectors of some kind
>and a box full of replacement pumps sold as part of the original package.
heh :-)
>i'm not sure what a homeowner might think if you hand them a box of 20 pumps
>and say "you'll need these".
sounds pretty funny :-)
>having to go out to the client and repair or replace parts every year or two
>is one method of trying to guarantee business but it can backfire on you by
>convincing people that your equipment is unreliable.
again, expectations matter. are oil filters "unreliable"? how about
house paint, or dishwashers or light bulbs or furnace filters? this
rebate program requires the contractor to give a 2-year warranty...
nick
|
|
