Posted by Chuck Olson on August 8, 2007, 12:09 am
Normally I would say the cleaner the glass panels the better. But it has
been suggested (by the people that installed my array) that I don't have to
be terribly fastidious about keeping the glass clean. He said that just
squirting the dust off with a garden hose every couple of months is just
fine, and sometimes the panels work better after this less-than-exhaustive
cleaning than when they were new.
I've been thinking if I could attach an extra long automobile windshield
wiper blade to a long pole, like a Gopher Pole, that will reach all the way
down my 3-up array, the grime removal would be much better than just
squirting them with a forceful stream. But is it worth the trouble to put
all this together?
I dug into my data records to compare May, June and July of 06 against those
same months in 07, and found my system produced 2585 KWh this year versus
2528 KWh last year. This was despite a heavy coating of grime that appeared
in June 07 due to demolition of the house next door and all the lot
leveling, compacting and excavation for the new construction.. I cleaned it
in early July using the garden hose, and the comparison showed only about a
2.2% difference in the KWh totals for the 3 months between 06 and 07, with
07 being better than 06, which kind of surprised me.
What's your experience?
Posted by R.H. Allen on August 8, 2007, 3:21 pm
Chuck Olson wrote:
Over the years I've talked to an awful lot of PV system owners. Of those
I have specifically discussed this with, I have yet to find one who
closely monitors energy production *and* thinks regular cleaning is
worthwhile. This includes some who operate systems in very dusty areas
and were utterly convinced that regular cleaning was a must until they
saw how small the impact on energy production was. Even in the academic
literature it's difficult to find systems that lose more than a few
percent to dirt. (On the other hand, with dust applied in a laboratory
under simulated "natural" conditions it's possible to demonstrate
I find that people are generally pretty surprised by how little their
energy production is affected by what appears to be an impenetrable
layer of dust. However, dust scatters light pretty effectively, and
since scattered light can't be focused it tends to mess with vision. A
surface that reflects 10% of incident light without scattering will look
quite transparent -- think home window glass (PV glass is much more
transparent than that). On the other hand, a surface that scatters 10%
of incident light will look fairly opaque -- perhaps something more like
what you'd see if you smeared petroleum jelly on the aforementioned
As for the energy production going *up*, I haven't heard of any
experimental or even simulated verification of such a thing, but in
theory the light reflected out from the inside of the module will also
be scattered by the dust. Some of it will be scattered back inside the
module and get another opportunity to be absorbed by a solar cell. Since
some 15% of the light that enters the module hits a gridline,
interconnect strip, or the exposed backing material between solar cells,
it wouldn't surprise me if dust can act as a sort of light trap.
Determining the optimum amount of dust would be nigh impossible, though,
as the magnitude and direction of scattering depends very much on the
size and shape of the dust particle.
Posted by Bill Kaszeta / Photovoltaic Re on August 9, 2007, 4:00 am
Much depends on the tilt of the PV array.
I have seen a system with horizontal modules (some architects just do not
understand solar) in which there was substantial dirt covering the cells in
one corner. This causes much more loss than the % of the module covered
because it limits the current in the series cells to that of the shadowed cell.
With parallel module strings and bypass diodes, the overall loss may
Photovoltaic Resources Int'l
Tempe Arizona USA
Posted by R.H. Allen on August 9, 2007, 1:31 pm
Bill Kaszeta / Photovoltaic Resources wrote:
Good point -- I didn't really think about uneven dirt distribution. Much
more substantial sorts of dirt, like bird excrement, might have a
similar effect and should probably be washed off if the amount is
Posted by never on August 9, 2007, 1:53 pm
Is the summery of this very interesting discussion now at the point:
better clean than dirty?