Solar Panels and Shedding Snow

ChrisOlson · December 2012

Re: Solar Panels and Shedding Snow

BTDT already. Vertical panels get tore right off the roof when we get high sustained winds from the north. They have to be below the peak or they won't survive when the wind gets compressed coming over the peak with 5 foot tall panels standing straight up. The panels themselves are not structurally strong enough to take it. They'll catch the wind at the top coming over that peak at 50-60 mph and dam the wind up behind them and build tremendous pressure behind those panels.

Lost three panels to wind 5 years ago that were mounted on a manual tracker rack on a 8" diameter x 20 foot Schedule 40 pipe with 6 feet of the pipe set in 12 yards of concrete. Tore the bolts and washers right thru the aluminum frames, and they were set at about 75 degree tilt at the time. The wind keeps working on them until they crack and then the panel is gone. After I lost those three panels I gave up on that setup and put them where they are now so they're better protected.

There's a reason I got the bottom of the panels off the roof by a foot, and that's to let the wind slide underneath so it doesn't build pressure behind the panels, between the panels and the roof. If I set them panels vertical it would blow the panels right out the first noreaster we get. The top of those panels would catch the turbulent blast coming over the peak of the roof and they'd all be gone.
--
Chris

Photowhit · December 2012

Re: Solar Panels and Shedding Snow

I think Wayne is talking about mounting the panels on the side of the building, rather than the roof...

You guys north of the border take care of your selves!, sounds like you know what your doing, Have a safe winter.

ChrisOlson · December 2012

Re: Solar Panels and Shedding Snow

Photowhit wrote: »

I think Wayne is talking about mounting the panels on the side of the building, rather than the roof...

Oh yeah. I re-read that and I had missed that. That would be fine for winter. But I don't think I'd like a bunch of struts sticking out from the wall supporting 14 solar panels for in the summer. And those 250 watt panels are too darned heavy to be taking off the roof or moving them all the time.
--
Chris

niel · December 2012

Re: Solar Panels and Shedding Snow

ChrisOlson wrote: »

Oh yeah. I re-read that and I had missed that. That would be fine for winter. But I don't think I'd like a bunch of struts sticking out from the wall supporting 14 solar panels for in the summer. And those 250 watt panels are too darned heavy to be taking off the roof or moving them all the time.
--
Chris

chris,
as i said there sometimes isn't any easy answers and i can understand not removing those pvs for every winter. if the drifts off of the roof are the problem maybe heat-tape the roof before it gets to the pvs?

tony,
i love your typo and i hope you don't mind my having fun with it.
"Climbing on a steve roof with snow on it is a death wish!"

yes, it would be pretty bad for steve. just how many steves does it take to make a roof anyway?:p

ChrisOlson · December 2012

Re: Solar Panels and Shedding Snow

Actually though, I like Wayne's idea of mounting the panels on the wall on the south side of the building. When I get my replacement panels I'm going to try it. We got more than enough for summer anyway, so those wouldn't need to be tilted up, and I can experiment with that vertical mounting on the wall to see if I get more out of them in winter.

Another thing I kind of wondered about, and maybe some folks have tried this - the slopes on our house roof face east and west. We don't really need the peak ouput at solar noon in summer. It would be nice to have more steady output all day. I wonder how it would work to put some panels on the east slope of the roof and some on the west slope? And don't make them adjustable for tilt. Use those Unirac rails. If I put them down near the bottom of the roof I can reach them pretty easy in the winter with a snow rake too.

I realize they wouldn't produce exceptional output on the "perfect" day in winter. But we usually only get ambient light conditions from the overcast anyway in winter. I wonder how that would work in summer?
--
Chris

ggunn · December 2012

Re: Solar Panels and Shedding Snow

ChrisOlson wrote: »

Another thing I kind of wondered about, and maybe some folks have tried this - the slopes on our house roof face east and west. We don't really need the peak ouput at solar noon in summer. It would be nice to have more steady output all day. I wonder how it would work to put some panels on the east slope of the roof and some on the west slope?

It would work as long as you didn't split a string between the east and west facing subarrays. It wouldn't max out the production of both subarrays unless you had an inverter with two MPPT channels, though.

BB. · December 2012

Re: Solar Panels and Shedding Snow

"Dave Sparks" likes longer hours at less peak current for recharging batteries--Longer life for the batteries. That is also why he likes trackers so much.

-Bill

icarus · December 2012

Re: Solar Panels and Shedding Snow

Never go on a Roof with Steve!

Meanwhile, I have my array split between wall mounts and roof mounts. I would do all wall mounts,, except I don't have enough wall area. The wall mounts are hinged, and the whole system works great. (although the system is so small, it is no headache)

Tony

inetdog · December 2012

Re: Solar Panels and Shedding Snow

ggunn wrote: »

It wouldn't max out the production of both subarrays unless you had an inverter with two MPPT channels, though.

Yes and no.....

If the two subarrays are composed of the same type of panel and either have no shading or have 100% identical shading then, regardless of the amount of light hitting them, the Vmp values should be very close if not identical for both arrays. If that is the case, a single MPPT controller will find and track that Vmp value and all that will change will be the relative amount of current contributed by each array.

If, on the other hand, one string had one or two shaded panels, its Vmp would be much lower (by one or two times the per-panel Vmp) than for the same string with all panels unshaded and two different MPPT channels would absolutely be required.

Photowhit · December 2012

Re: Solar Panels and Shedding Snow

VMP shouldn't be effected by shading, in fact in summer VMP might go up if the panel cools, shading would effect the current, which an MPPT controller should handle fine.

inetdog · December 2012

Re: Solar Panels and Shedding Snow

Vmp of a panel will be reduced if one section of the panel is "cut out" by its bypass diode. If the available Isc of a shaded panel section is reduced below the current that the rest of the panel is able to produce, the current will flow through the bypass diode instead and that section will not contribute to the panel voltage (and it will actually reduce the voltage even more, by the diode forward drop.
The same effect will also be seen in a string of panels where either whole panels or sections of panels are shaded.

Voc might not go down on a shaded panel because no current is being demanded of the shaded section, but Vmp will definitely be reduced.

Any panel temperature effect will in fact cause a Vmp difference, and I did not consider that. But during the times that the subarray temperatures are significantly different, only one of them will be producing significant power, and therefore the MPPT will follow that subarray. Any power from the other subarray will be less than optimal, but not enough to worry about.

Cariboocoot · December 2012

Re: Solar Panels and Shedding Snow

Two words: shunt resistance.

inetdog · December 2012

Re: Solar Panels and Shedding Snow

Cariboocoot wrote: »

Two words: shunt resistance.

Two different words: negligible effect
Four letters: IMHO

More seriously, how do you expect shunt resistance to play a role in determining the panel Vmp at any output which is large enough to care about, or the Voc at any light level high enough to care about?

Cariboocoot · December 2012

Re: Solar Panels and Shedding Snow

Sorry; thought we were talking about a large array divided into two halves facing in opposite directions, not a single shaded panel out of many. My mistake, I guess.

Photowhit · December 2012

Re: Solar Panels and Shedding Snow

IMHO if you have enough light to have shadows, you will have enough light to meet VMP, If your talking about the original snow shading, then perhaps you could, but short of something laying on the panels...I'll stand pat.

inetdog · December 2012

Re: Solar Panels and Shedding Snow

Cariboocoot wrote: »

Sorry; thought we were talking about a large array divided into two halves facing in opposite directions, not a single shaded panel out of many. My mistake, I guess.

We are. That is exactly the situation. I do not see why the light level or the angle of light to the panel will have any noticeable effect on the Vmp of that array. Series resistance rather than shunt resistance would have an effect, but I think that the current source characteristics of a PV panel would make the Vmp voltage level almost independent of the load current. I may have to think about that some more though.
I am basing my reasoning on the expectation that the reason the current reaches a maximum at Isc is not the internal resistance of the panel but rather the amount of light which is converted to electrical energy, and that as the voltage approaches Voc, fewer photons actually get converted to separated charges. Imp is strongly dependent on the light level, but Vmp should not be nearly as dependent.

inetdog · December 2012

Re: Solar Panels and Shedding Snow

Photowhit wrote: »

IMHO if you have enough light to have shadows, you will have enough light to meet VMP, If your talking about the original snow shading, then perhaps you could, but short of something laying on the panels...I'll stand pat.

You are welcome to stand, but I prefer to sit and discuss it. :-)

A shaded panel will produce an open circuit voltage which is verrrrry close to Voc of a panel in full light. We agree on that, I am sure.
A shaded panel will therefore be able to produce Vmp as long as the current flowing through it is not too high. Still in agreement?
A shaded panel which can produce an Imp of 1 amp, put in series with other panels in the string which can produce an Imp of 10 amps will not be able to maintain Vmp across its terminals.
That is the point that I am trying to make.

The MPPT CC or GTI will be trying to draw more current to get more power, as long as the drop in voltage combined with increase in current does not instead cause the power to fall.
When the MPPT starts to draw current above Isc of the shaded panel, the voltage contribution from that panel will nosedive and drop almost vertically to -(NxFD) where N is the number of bypass diodes and FD is the forward voltage drop at current. The effect is that for that string, the MPPT controller will see such a large voltage drop as it passes through Imp/shaded that with the other string still present and delivering current it will not try to reduce the voltage any further to get to the Vmp-Imp point of the shaded string.

A single partially shaded string can actually be a problem all by itself for an MPPT circuit that does not actually do a full scan over a wide range instead of just following the short range slope of the power curve.

Cariboocoot · December 2012

Re: Solar Panels and Shedding Snow

Vmp is a variable. As far as panels are concerned they produce current first, Imp @ 0 Volts up to the rated Vmp, Voltage second. If the Voltage difference between two paralleled arrays is too great, the lower V array becomes a drain on the higher rather than contributing current to the total. This is usually seen with very high Voltage arrays such as GTI's use rather than on battery-based systems. The exception to this is when the array have two different types of panels which, although they add up to the equivalent Vmp, have different shunt resistance. The same effect can be achieved with similar panels under very different levels of illumination.

ChrisOlson · December 2012

Re: Solar Panels and Shedding Snow

OK, well you guys started spewing a bunch of Solar Speak that I don't even understand

So if I put 8 panels on the east slope our house roof, and 8 on the west slope, can I hook them all to one Classic? Or do I need two controllers for that? If I could do that, and mount them on those Unirac rails I wouldn't have any problems with wind load or snow load on them. And they'd be easy to clean off.

I'm not exceedingly happy with the setup I have right now.
--
Chris

Cariboocoot · December 2012

Re: Solar Panels and Shedding Snow

Chris; my advice is that if you're going to split the array in two use two separate controllers. Although there have been previous discussions about this with people arguing against it, Solar Guppy did find some time ago that two different (or "too different") arrays feeding one MPPT input can confuse it resulting in less than ideal power point for either array. Note also that some GTI's have two separte MPPT "front ends" specifically to deal with this.

BTW, my panels are in one row and at the ridge to avoid snow build-up behind them. Having a 37 degree roof slope doesn't hurt either. Sounds like you're getting snow so heavy it just builds up by the foot until it piles up over the roof anyway. Just put your mind to it; you'll figure out something that will work best for you. Panels mounted vertically back-to-back at the ridge perhaps.

ChrisOlson · December 2012

Re: Solar Panels and Shedding Snow

Thanks. I like the idea of putting some panels on the vertical wall of the shop - I think four panels would be good there. Those wouldn't do much in the summer but maybe I could get more from them in the winter.

The reason I never put panels on the house roof is because it's pointing the wrong way (east and west). But if it works to split them up that would be the best place because the wind can't get at them and snow will build up but it's really easy to reach from the ground with a snow rake to pull it off. I just didn't know how it would affect output of them in the summer.

I ordered six panels for the four that broke. The shipping was the same amount. And they're cheaper now so I got six for the price of four when I got the other ones. So now I'll have 16 of the things

--
Chris

inetdog · December 2012

Re: Solar Panels and Shedding Snow

ChrisOlson wrote: »

So if I put 8 panels on the east slope our house roof, and 8 on the west slope, can I hook them all to one Classic? Or do I need two controllers for that?
Chris

As long as the two sets of panels are identical types, have the same number of panels per string, and have NO partial shading during the hours when either of them would be producing a significant amount of power, my answer would be that the added power from using two separate MPPT circuits would not justify the cost of the additional CC. If you had more panel power than one CC could handle, it would be a different question. Or if you wanted to have a second CC as a spare so that you could still get power at a reduced capacity if one failed. With your available wind power and generator, that level of redundancy may not be justified.

But all that is a purely theoretical recommendation, not one based on experience.

UPDATE:

I finally found the image I was looking for.

Attachment not found.

This shows a family of voltage versus current and POWER curves for a representative panel under different levels of light. The Max Power Point on each curve is where the near-vertical line intersects the blue power curve.
The MPP is not constant over illumination, but it remains within 5% of the nominal STC/PTC value of Vmp down to 10% of full output and below. So as long as there is enough light that the panel is contributing useful power, the Vmp of the two subarrays will be within the 5% rule of thumb for putting strings in parallel. Hence a single MPPT device (CC in this case) will get so close to the maximum power from both arrays that the extra MPPT input is not worth the cost (unless you have, for some other reason, chosen a CC which has two inputs.)

Where there will be a difference, if at all, will be when the temperatures of the two arrays are very different. But that will only be the case during the short transition between running with the already warm east-facing array and starting to pick up power as the light starts to hit the west-facing array. At that point, the output of the west-facing array will still be so small that again the difference in Vmp will not lower the sum of the outputs by very much.

ChrisOlson · December 2012

Re: Solar Panels and Shedding Snow

inetdog - that is very cool. Thanks! If I could use that mounting setup it would solve a LOT of problems here that I've had mounting solar panels. I've tried everything for mounting to have them facing south and ran into issues with everything I've tried. I've always wished our house was built with a south facing roof slope, but it wasn't. If I can get this work, and don't see why it won't. it would make the solar panels a lot easier to manage for snow and wind load. And hopefully still work fine in spring/summer/fall when we need them the most.

We do not need the peak output at solar noon. It would nicer to have a more steady lower output all day. And on the house roof I have zero problems with shading from trees or anything.

Thanks!
--
Chris

ggunn · December 2012

Re: Solar Panels and Shedding Snow

inetdog wrote: »

Yes and no.....

If the two subarrays are composed of the same type of panel and either have no shading or have 100% identical shading then, regardless of the amount of light hitting them, the Vmp values should be very close if not identical for both arrays. If that is the case, a single MPPT controller will find and track that Vmp value and all that will change will be the relative amount of current contributed by each array.

If, on the other hand, one string had one or two shaded panels, its Vmp would be much lower (by one or two times the per-panel Vmp) than for the same string with all panels unshaded and two different MPPT channels would absolutely be required.

He said the roofs face east and west. With the roofs facing 180 degrees from each other I would expect that for much of the day the ideal MPP tracking point would be different for the two subarrays. Of course, roof slope comes into play; the lower the pitch of the roof the less the difference in insolation between the two surfaces, but since he is living in a high snow area, I would assume that the pitch of the roof is high. For parts of the day, one or the other of the arrays might be completely shaded. He lives at high latitude.

Anyway, qualitatively speaking I think there would be a difference. Quantitatively, I haven't a clue.

inetdog · December 2012

Re: Solar Panels and Shedding Snow

ggunn wrote: »

He said the roofs face east and west. With the roofs facing 180 degrees from each other I would expect that for much of the day the ideal MPP tracking point would be different for the two subarrays.

The ideal MPP will be different, yes, but the curves of Vmp versus insolation strongly indicate that the difference is so small that "not-ideal" will in fact be "good enough".

Cariboocoot · December 2012

Re: Solar Panels and Shedding Snow

inetdog wrote: »

The ideal MPP will be different, yes, but the curves of Vmp versus insolation strongly indicate that the difference is so small that "not-ideal" will in fact be "good enough".

Unless, of course, you're in Minnesota or some such in mid Winter and really need to squeeze out every possible Watt hour.

For anyone wondering about this, a couple of things to ponder:

1). If you connect a "12 Volt" panel (Vmp 17.5) in parallel with a "24 Volt" panel (Vmp 35), what is the effective array Vmp under equal illumination and why?

2). If you connect a "12 Volt" panel to a "12 Volt" battery with no charge controller or blocking diode and put the panel in darkness (id est night), what happens and why?

inetdog · December 2012

Re: Solar Panels and Shedding Snow

Cariboocoot wrote: »

Unless, of course, you're in Minnesota or some such in mid Winter and really need to squeeze out every possible Watt hour.

Absolutely right.
My point is that the idea of losing 10% of one subarray's array output (which is probably an overly conservative figure-variations from MPP do not cut into the output very sharply) may sound really bad. But when you consider that at the time this loss is happening that subarray is only delivering one tenth of its max power, you are only looking at a 1% difference between a single MPPT and dual MPPT.
For some people with a hard minimum usage of 2Kwh per day, the difference between getting 1990 wh from the array and 2010 wh from the array can be critical. (Where's that generator anyway?)
But for those people, adding two more panels would benefit them a lot more at lower cost than buying a second MPPT CC.

Cariboocoot · December 2012

Re: Solar Panels and Shedding Snow

Or you might as well buy one or two PWM controllers and give up any potential MPPT benefit.

An MPPT controller being fed from two such different arrays will find a power point, but it will not be the maximum power point for either array. And that's sort of the whole idea behind Maximum Power Point Tracking.

Note that Chris's system is quite large by off-grid standards, so for him the benefit would be greater than it would be for a small system (like mine).

inetdog · December 2012

Re: Solar Panels and Shedding Snow

Cariboocoot wrote: »

Or you might as well buy one or two PWM controllers and give up any potential MPPT benefit.

An MPPT controller being fed from two such different arrays will find a power point, but it will not be the maximum power point for either array. And that's sort of the whole idea behind Maximum Power Point Tracking.

I assume that you are not serious about that suggestion, since a PWM controller with a perfectly matched array configuration will still have greater losses than an MPPT controller that is running a short distance from the MPP point. :-)
Once again, my assertion is that if the two arrays are producing unequal amounts of power, the MPPT algorithm will follow more closely the one with the higher output, and when the two are producing roughly equally their Vmp will be identical.
Keep firmly in mind that what the two subarrays will have in common is just the voltage level at the CC. They will each contribute whatever current their I versus V curve family shows for that voltage and that insolation. A difference of 5% in voltage away from MPP will produce less than a 2% difference in power output.

If the Voc of the two arrays is different, I do not dispute that at some point a large enough difference in Voc and Vmp for two subarrays of different panel specification (even with the same insolation) will result in zero output from one of the two subarrays, ALL of the time. At that point, getting even a PWM controller for the second subarray will be better than paralleling them to an existing MPPT controller.

Some of this is counterintuitive unless you look very carefully at the math, so I would welcome input from anybody who is current running two subarrays into two MPPT CCs or one CC with two MPPT inputs. A comparison of the the actual Vmp from the the two CCs at different times during the day will either support my hypothesis or shoot it down. :-)

Cariboocoot · December 2012

Re: Solar Panels and Shedding Snow

Voc doesn't actually exist on panels that are connected to a controller: no "open circuit" hence no "Voltage open circuit".

I know of no reason to assume an MPPT controller will chose to follow the higher Voltage array, especially since if they are in parallel there is only one array and its Voltage point will be somewhere between 0 and Vmp. The controller has no way of knowing what it will be in advance. It sees an array, tests it for maximum power point, and picks what V and I it thinks is best. The results of this will vary with the particular panels, the particular controller, and the particular installation.

If the Voltage difference between the two panel sets is large enough, one set will actually draw power rather than contribute it.

Since Solar Guppy has already done this real world experiment and achieved the results I've explained here there's no reason to repeat it.

Solar Panels and Shedding Snow

Comments

Categories