Snow removal

[bcroe]

I'll be curious to know if anyone can think of a good deicer for this application. Something which:

• Isn't corrosive
• Doesn't kill the landscaping
• Is OK to let run into ponds, streams, and storm sewers
• Doesn't degrade roofing material or sealants
• Is an effective deicer

The best candidates I could think of are sugar water (which turns out to be a fairly poor deicer) and ethanol (which I'm afraid might do unhelpful things to asphalt shingles). Many other common deicers are easy to rule out as corrosive (salt) or bad for plants or aquatic animals (glycols).

I wouldn't try anything that isn't a great insulator. That may eliminate most. I spilled a bit of
sugar water (Coke) on an old keyboard, and I finally had to throw it out. Think I'll just be waiting
for a better combination of temp & sun.

I believe your panels are larger than any of mine. Bruce Roe

[pleppik]

It might be time to think of using some type of deicing solution.

I'll be curious to know if anyone can think of a good deicer for this application. Something which:

• Isn't corrosive
• Doesn't kill the landscaping
• Is OK to let run into ponds, streams, and storm sewers
• Doesn't degrade roofing material or sealants
• Is an effective deicer

The best candidates I could think of are sugar water (which turns out to be a fairly poor deicer) and ethanol (which I'm afraid might do unhelpful things to asphalt shingles). Many other common deicers are easy to rule out as corrosive (salt) or bad for plants or aquatic animals (glycols).

[pleppik]

That panel has 60 cells 6" square, or 15 square feet, or about 1.66 sq yard, or about 1.4 sq meter,
was a little off.

I was guessing it was a lot smaller. I think the perspective in the photo threw me off.

[SunEagle]

That panel has 60 cells 6" square, or 15 square feet, or about 1.66 sq yard, or about 1.4 sq meter,
was a little off.




Yea there is a level that will melt snow, and a level that will destroy. Wonder which is higher? Bruce

It might be time to think of using some type of deicing solution.

[bcroe]

That sounds high to me. Direct sunlight perpendicular to the surface it's hitting is
roughly 1KW/m2 at sea level, and your panel looks like it's a lot smaller than a square meter.

That panel has 60 cells 6" square, or 15 square feet, or about 1.66 sq yard, or about 1.4 sq meter,
was a little off.

I'm guessing that if the DC power supply didn't work at 741W, this won't work either.
As Scotty (from Star Trek) might say, "I canna change the laws o physics! Ye need more power!"

Yea there is a level that will melt snow, and a level that will destroy. Wonder which is higher? Bruce

[pleppik]

My guesstimate is a peak of 1600 W of sun could hit the panel, way more than I wanted
to put through. and practically way more than I could develop for an array of some size.

That sounds high to me. Direct sunlight perpendicular to the surface it's hitting is roughly 1KW/m2 at sea level, and your panel looks like it's a lot smaller than a square meter.

I did have one idea about the power. If some strings of panels were clear and putting
out significant power, I could short out half of a snow covered string. The active strings
would dump all their current into the other half of the covered string. After is was clear,
short out the other half instead. The other snow covered panels wouldn't take current,
with the voltage pulled down to around half; they might even make a minor contribution.
Could run that pretty easily If I braved that terrible - XX wind chill. Could end up with a
bunch of burned out solar panels. Bruce Roe

I'm guessing that if the DC power supply didn't work at 741W, this won't work either. As Scotty (from Star Trek) might say, "I canna change the laws o physics! Ye need more power!"

[bcroe]

The 741W is in the same ballpark of how much heat would go into warming the panel if it was placed in full direct sunlight (with no snow). I would have thought that would be enough to at least break the bond between the snow and the panel if the heat was distributed evenly across the surface.

It's possible that 5F is just too cold for this method to work. You need to heat the panel by nearly 30 degrees to get it above freezing, and solar panels are not designed to retain heat (the opposite, in fact).

My guesstimate is a peak of 1600 W of sun could hit the panel, way more than I wanted
to put through. and practically way more than I could develop for an array of some size.

I did have one idea about the power. If some strings of panels were clear and putting
out significant power, I could short out half of a snow covered string. The active strings
would dump all their current into the other half of the covered string. After is was clear,
short out the other half instead. The other snow covered panels wouldn't take current,
with the voltage pulled down to around half; they might even make a minor contribution.
Could run that pretty easily If I braved that terrible - XX wind chill. Could end up with a
bunch of burned out solar panels. Bruce Roe

[pleppik]

Finally I actually ran an experiment. Its around 5 F, and I put up this panel. Its
the typical 250 W 60 cell panel. Some snow is on it from the erection process.

.....

OK, I raised the power to 46.3V @ 16A, about 741 W. After 10 minutes the voltage dropped
to 45.9V, but the snow didn't move. That is about all the current this supply can deliver, and
about all I want to put through a panel. My conclusion: this is not the right approach to snow.

The 741W is in the same ballpark of how much heat would go into warming the panel if it was placed in full direct sunlight (with no snow). I would have thought that would be enough to at least break the bond between the snow and the panel if the heat was distributed evenly across the surface.

It's possible that 5F is just too cold for this method to work. You need to heat the panel by nearly 30 degrees to get it above freezing, and solar panels are not designed to retain heat (the opposite, in fact).

[bcroe]

Well thanks for trying.
I now have three arrays. One is a pole mount adjustable for tilt, one is mounted off a wall and is adjustable for angle and the last array is on a shallow pitch on a 2nd floor roof. Both of the adjustable arrays do shed snow but on occasion I need to clean them off or wait for a sunny day. The roof mounted array is definitely a snow catcher and if the conditions are right it could be weeks for it to clear off. Part of the issue is that the snow will start sliding slowly down the panels exposing the top of the panel. The snow then builds up at the base of the panels on the roof and tends to freeze. which slows down the snow sliding down. I have considered installing some flashing at the base of the panels to direct the snow from the edge of the panels to the edge of the roof and possibly installing some heat tape under the flashing. I expect it doesn't need to be on for long but if the sun and outdoor temps are high enough the heat tape may be what makes the snow move. The down side is that the flashing at the base of the panels will screw up summertime convective cooling as it will impede air flow at the base.

Its a little like that here. 2013/14 was my 1st "panel winter", maintaining neg 20F throughout.
Very uncomfortable but great solar when the sun came out. Snow remained mostly powdery.
This year got some snow in warmer weather, sun started to melt it off. It slid down the panels
about 20% and then froze solid. Maintaining 0 F. About all I can do is wait for a combination
of sun and warmer weather to get it off. Maybe sprinkle it with black ash some sunny day?
But despite that its putting out around 30%, largely thanks to the vertical and clear E & W
panels.

Keep us informed on the heat tape experiment. Makes me wonder if there should be a gap
between the upper & lower panels, so the snow wouldn't have to slide so far to fall to the
ground. Will be considering that. Bruce Roe

[peakbagger]

Well thanks for trying.

I now have three arrays. One is a pole mount adjustable for tilt, one is mounted off a wall and is adjustable for angle and the last array is on a shallow pitch on a 2nd floor roof. Both of the adjustable arrays do shed snow but on occasion I need to clean them off or wait for a sunny day. The roof mounted array is definitely a snow catcher and if the conditions are right it could be weeks for it to clear off. Part of the issue is that the snow will start sliding slowly down the panels exposing the top of the panel. The snow then builds up at the base of the panels on the roof and tends to freeze. which slows down the snow sliding down. I have considered installing some flashing at the base of the panels to direct the snow from the edge of the panels to the edge of the roof and possibly installing some heat tape under the flashing. I expect it doesn't need to be on for long but if the sun and outdoor temps are high enough the heat tape may be what makes the snow move. The down side is that the flashing at the base of the panels will screw up summertime convective cooling as it will impede air flow at the base.

[bcroe]

Finally I actually ran an experiment. Its around 5 F, and I put up this panel. Its
the typical 250 W 60 cell panel. Some snow is on it from the erection process.

The 1 KW variable DC power supply built a while back is used to supply the warming power. The
panel is so thin, I expect the time constant is short.

+ power is applied to the panel + output wire, etc. About the time the applied voltage approaches
the Voc, current flows in the reverse direction. This is probably the same thing that happens when
a panel is left open circuit in the sun. I applied 42.6V @ 5.9A, about 251 W. Voltage is fairly
critical, as the cell diodes approaches forward conduction. I regulated the current, expecting the
voltage to decrease a bit as the panel warmed up.

After 20 minutes, the voltage dropped to 42.2V @ 5.9A. However the snow remained firmly
in place.

Power was increased to 44.6V @ 11A, about 490 W. After 10 minutes the voltage dropped
to about 44.3V, maybe someone can estimate how much the temperature increased. Guess
I need the infared thermometer. But the snow didn't move.

OK, I raised the power to 46.3V @ 16A, about 741 W. After 10 minutes the voltage dropped
to 45.9V, but the snow didn't move. That is about all the current this supply can deliver, and
about all I want to put through a panel. My conclusion: this is not the right approach to snow.

That leaves tilting the panels either for the season, or maybe over vertical before a storm. Its
already obvious that the closer to vertical a panel is, the less snow accumulates. Or scraping.
Bruce Roe

[bcroe]

With one below zero winter behind. I have had more thoughts on this.
I didn't have any trouble with snow sticking, maybe because of the very
low temps out at my ground mound. And it was noted far less snow
was on the near vertical panels than the 45 degree elevated panels,
sometimes almost none.

The heated (backed) experiment may be run winter 14/15. But before
doing that large scale, I think I'll rebuild the mount to make changing the
elevation angle a very simple thing. That means some kind of pivot near
the center to take the weight, either a bearing, big bolt, or a hinge on
each side (8' apart). And activators could be placed to do the work.
Reuse the original concrete foundation, different above ground structure.

The thought is, easily set the panels vertical for the snow, or even a
predicted heavy hail storm (one damaging cars just missed here). Maybe
set them a few degrees over vertical overnight for a storm, then back.
Bruce Roe

[Sharich]

icing

While a steep angle my help minimize the snow accumulation on panels, once there tilting more may not help. As the snow falls, especially if the temp is slightly above freezing and drops as the snow accumulates, it ends up frozen to the panels. This happens often and when it does I need to physically break it loose. I usually either push up from the bottom and then let it slide, when that does little, I lift the broom and somewhat gently let it fall on the snow, about a foot or 2 up from the lowest cleared area depending on the snow conditions and adhesion and pull downward. If the sun is shinning I only do strips, diagonally and the sun warms the panel rather quickly and it slides off on it's own in short order. A little experience soon educates you on how much needs to be pulled off and what the sun will remove in a short time. If the sun is bright and the temp in the upper 20's or higher, I often just push the snow off the bottom 2' or so across the 30' width of the array. Then the suns warms the surface and the panels self clear in a reasonable time.

That sounds good, but at my age I'd have to be careful. I did see RV antifreeze that may be worth looking at, one brand has ingredients that are non corrosive.

[maple flats]

While a steep angle my help minimize the snow accumulation on panels, once there tilting more may not help. As the snow falls, especially if the temp is slightly above freezing and drops as the snow accumulates, it ends up frozen to the panels. This happens often and when it does I need to physically break it loose. I usually either push up from the bottom and then let it slide, when that does little, I lift the broom and somewhat gently let it fall on the snow, about a foot or 2 up from the lowest cleared area depending on the snow conditions and adhesion and pull downward. If the sun is shinning I only do strips, diagonally and the sun warms the panel rather quickly and it slides off on it's own in short order. A little experience soon educates you on how much needs to be pulled off and what the sun will remove in a short time. If the sun is bright and the temp in the upper 20's or higher, I often just push the snow off the bottom 2' or so across the 30' width of the array. Then the suns warms the surface and the panels self clear in a reasonable time.

[J.P.M.]

I thought about heat tape around the panels, but haven't heard of anyone suggesting this.

I once suggested treating panel glazing like windshields w/ embedded electrical defrosters. That idea was DOA.