Mounting PV panels vertically

[craig110]

I live in an area of the country that is "blessed" by a lot of snow each winter (New England) and, besides not looking forward to an ongoing chore, my roof configuration makes cleaning snow off PV panels either quite difficult or hazardous.  I don't like either of those choices.  Combining three notions gives me a possible solution to this problem: 1) solar panels get warm in sunlight, 2) snow is a good reflector, and, 3) snow doesn't accumulate easily on vertical structures.  What I'm thinking of doing is mounting my panels vertically -- yes, aimed directly at the horizon -- and put aluminum panels flat in front of them to reflect sunlight up to the panels.  (Putting the reflective panels in a semicircle around the PV, of course, to continually reflect onto the PVs no matter which direction the sun is at.)  I live at about 42.5 degrees N, so in the summer, the panel would be "wrong" by about 45 degrees but would have plenty of light reflected up by the panel.  In the winter when the sun is very low, the panel would be less "wrong" (which is good for catching those feeble winter rays) and also get light reflected off the reflectors and/or snow.

I've tested this on the ground with one of my panels and the panel put out about the same amount of power whether aimed directly at the sun or oriented vertically with reflective panels on the ground in front of it.

One other advantage seems to be that since the reflective panel wouldn't be a perfect mirror, the reflected light is dispersed which tends to minimize the impact of a small shadow.

Has anyone tried this configuration?  If so, how did it work out?  Yes, I realize this is not as efficient as continually cleaning the snow off the panels and periodically reorienting them, but I'm trying to strike a balance between low maintenance and decent electricity production.  (And no, mounting them anywhere but the roof is not possible.)

Thanks!

Craig

[ghurd]

Hi Craig,

This is all related to my experiences in the Great Lakes snow belt, about 41 degrees. 60 miles west of Cleveland.

Mount them 10 degrees steeper than the latitude. Even 25 degrees steeper would be fine if you are going change the tilt in mid spring and fall. And that is quite close to the best winter angle anyway.

Here, the snow slides off in sheets around mid morning, or a corner is exposed enough to heat the panels so the snow melts if any significant sunlight shows.

A couple things to keep in mind.

-Even 60 degrees is going to be a minor loss in winter at your location.

-The AL plate is going to be covered in snow, just like the PV would have been.

-The sugestion takes care of itself, like when you go to work while its snowing.

-Snow is going to be a good reflector, like the AL plate.

-I know of a verticle 80W 12V PV with a 3A fuse not blow the fuse, for 5+ years and counting. Snow reflecting even. That really says something to me, like he is getting less than 64% of what he paid for. All year. Best case. Worse in the summer.

-Around here, I recomend low mounting, and keeping a ladder and long handled broom handy. They don't have to be clean, just 1/2 clean.

G-

[veewee77]

Don't take this wrong, but think about your idea for a minute.  You are having trouble with snow collecting on your PV panels, so you want to mount them vertically and place a flat shiny aluminum something to reflect the light up to the panel. How are you going to keep the snow from collecting on the aluminum thing and defeating the whole purpose?  Seems to me it would be easier to sweep the panels than to try to get high enough to sweep those flat things laying down in front of the panels. . .

Doug

[craig110]

Hi Doug,

The thing is that snow is an excellent reflector, so there is no need to remove the snow from the aluminum sheets.  Think of the aluminum panels as "snow replacement" to get the reflectiveness when there is no snow on the roof.

Craig

[Tom in NH]

Actually, I think it's worth a try. Snow would make an excellent reflector in its own right. I suspect a reflector of any sort would yield some gain, but the reflected light would strike the PVs at a sharp angle, making me wonder how much gain there is to be had. It is quite amazing how much the output falls off as the sun angle moves from a perfect 90 degree angle. --tom

[craig110]

Hi Ghurd,

Thanks for the comments.  I agree that mounting them vertically isn't the most optimal for energy production.  Mounting them low and keeping the ladder and broom handy won't work, sad to say, given the layout of my roof.

Craig

[TomW]

Folks;

With a whopping 4 years of experience in winter and snow on panels I can attest to the fact that, given some sun and time the snow will slide right off if the panels are set to the latitude + 10 degrees. It only takes a tiny bit of current flowing in the panels to break the bond of the snow to the glass then gravity takes over especially if any tiny area is exposed.

My panels are on a low pole with tracker and specifically set up so I can broom them off but it is really seldom needed. Once the sun gets on them they will shed the snow fairly soon.

However, if I had to mount them fixed in an inaccessible location, I would consider a simple automotive [or truck] windshield wiper rigged to swipe the lower section with a remote switch. Then, if it snows, you can swipe off the lower section which will soon allow some current to flow and allow the snow to slide off. I would also ensure there is room below to let the snow fall clear of the panels when it does.

I actually think I lose more % of power in summer from dust, bird crap and other debris than snow in winter so it is handy if you can hose them off occasionally but with good water pressure you may be able to do that from the ground.

Cheers.

TomW

[veewee77]

well, snow may be a good reflector, but it will be VERY diffuse! You will only get a very small percentage to actually fall on the panels. That much probably makes it through the snow anyway. . . Another option would be to make a wiper assembly, mounted to one corner of the panels and tied to a rope to ground level and when the snow builds up, just go pull the rope and sweep them off. Even partial sweeping will let enough light(heat) in to melt the rest.

I wouldn't mess with the reflective snow thing. . .

JMO - YMMV

Doug

[GaryGary]

Hi,

I have a little program that that calculates radiation on a collector given your location, and the panels tilt and azimuth.  I pasted in a couple of runs for your reflector idea.  The first is the panel facing south and tilted at 42.5 degrees (your lat) -- this would be the normal recommendation for best year round collection.  The 2nd run is for a panel facing south and titled at 90 degrees (vertical).  The last column gives the energy incident on the panel for one day in each month.  So, if you compare the last column in each of the two runs you will see what kind of hit you take for the vertical panel with NO REFLECTOR.  Adding a reflector will help.  I think I have some material on how to estimate the gain that a reflector will give you -- if I can find it, I'll put it up tomorrow.  

If you want to play around with other angles, you can download the program that does the incident radiation calc from my site here:

http://www.builditsolar.com/Tools/RadOnCol/radoncol.htm

Without the reflector, its clear that you take a big hit in mid-summer for the vertical mount, because the sun is up very high (42.5 + 23.5 = 66 degree at noon in mid summer), so the angle between the sun and the panel at noon is (90-66) = 24 degrees, where you would like it to be 90 degrees.  A horizontal reflector in mid summer  at noon would reflect rays up such that they also hit the panel at 24 degree angle -- this does not seem ideal -- maybe the reflector needs to move seasonally.  Maybe tilting the reflector up some in the summer would help -- you could experiment with a mirror and flashlight :-)  

Just a note -- in the summer, horizontal panels do quite well -- how about vertical in the winter and low tilt or horz in the summer?  You could use the horizontal reflector to boost up the winter input, and forget the reflector in the summer when you get more sun anyway?  Or, if you have good snow coverage, use the snow as a reflector in the winter with panels vertical, and change the tilt for summer to something like latitude -15 degrees (ie nearly flat)?

This is the panel tilted at 42.5 degrees:

-----------------------------------------

Month by Month Summary of Sun on Collector

(100% sunny weather)

  Collector Area:        1.0 (sqft)

  Collector Azimuth:     0.0 (deg) measured from South

  Collector Tilt:       42.5 (deg) measured from horiz

  Latitude:             42.5 (deg)

  Altitude above SL:     0.0 (ft) Above Sea Level

Date ----  Sun ----------------  Collector -----------------------

Month Day  Direct   Di-   Total    Direct      Difuse     Total

           Normal   fuse

   1   21   2074     120   2194       1631        105       1736

   2   21   2506     150   2656       1953        130       2083

   3   21   2879     204   3084       2124        178       2302

   4   21   3094     301   3395       2043        262       2305

   5   21   3185     386   3571       1919        336       2254

   6   21   3226     432   3657       1846        375       2220

   7   21   3086     419   3505       1858        364       2222

   8   21   2912     354   3265       1936        307       2243

   9   21   2676     249   2924       1986        216       2202

  10   21   2336     169   2506       1837        147       1984

  11   21   2000     126   2125       1576        109       1685

  12   21   1811     104   1915       1431         90       1521

Sum        31783    3015  34799      22138       2619      24757

Radiation in BTU/day

This is the panel tilted at 90 degrees

---------------------------------------

Month by Month Summary of Sun on Collector

(100% sunny weather)

  Collector Area:        1.0 (sqft)

  Collector Azimuth:     0.0 (deg) measured from South

  Collector Tilt:       90.0 (deg) measured from horiz

  Latitude:             42.5 (deg)

  Altitude above SL:     0.0 (ft) Above Sea Level

Date ----  Sun ----------------  Collector -----------------------

Month Day  Direct   Di-   Total    Direct      Difuse     Total

           Normal   fuse

   1   21   2074     120   2194       1619         60       1679

   2   21   2506     150   2656       1663         75       1738

   3   21   2879     204   3084       1435        102       1537

   4   21   3094     301   3395        943        151       1094

   5   21   3185     386   3571        607        193        800

   6   21   3226     432   3657        476        216        691

   7   21   3086     419   3505        575        210        785

   8   21   2912     354   3265        883        177       1060

   9   21   2676     249   2924       1328        124       1452

  10   21   2336     169   2506       1572         85       1657

  11   21   2000     126   2125       1568         63       1631

  12   21   1811     104   1915       1497         52       1549

Sum        31783    3015  34799      14166       1508      15674

Radiation in BTU/day

Gary

[craig110]

Hi Gary,

Thanks for the pointer to your program and for looking for the additional info on the reflectors.  I reran the program to see what the panel would do if it was mounted vertically (you had compared 42.5 degree tilt with a 90 degree tilt which would be aimed straight up) and was quite surprised.  The three sets of numbers are

42.5 degree tilt: 24,757

aimed straight up: 15,674

aimed at horizon (i.e. mounted vertically): 20,804

"Only" losing an annual 16% by mounting it vertically (before factoring in the reflector, of course) was a lot less than I anticipated.  But, the impact during the middle of the winter was pretty bad.  I'll have to do some more ground tests with the reflectors, but combining these numbers with everyone's input that the snow on the panels will melt easily, I have some more thinking to do.

Craig

[RP]

Personally I think the windshield wiper idea deserves more thought.  A couple ideas:

1.  Two horizontally aligned threaded rods.  One each across the top and bottom of the panel with a brush affair connected to nuts between the two rods.  A small dc gear motor to drive it across from either left to right across the panel (both rods would need to be connected by chain or toothed belts on one end).  Add an end limit switch at each end with diodes across the contacts.  You only need two wires running into the house.  After a snow you flip a switch that reverses polarity across the system and the motor drives the brush to the other end limit.  When it gets there it opens the limit switch and stops until the polarity is reversed again.
   
2.  How difficult would it be to create a motorized way to raise the panel to vertical during snow and then return it to normal afterwards.  I wonder if the linear motor mechanism from an old C-band satellite dish could do this.
   

[frogvalley]

My solution and recommendation is to use standoff type mounts with your racks, making the bottom of the panels themselves 8 to 12 inches or more off the roof the snow will shed itself well. Go with the best angle for winter or all around power and let the snow fall where it may. Although this year we haven't seen any real snow here in eastern WV, I have had snow melting off my panels for years with out ever sweeping.

What follows is a rough translation of what went through my head to arrive at this conclusion. Well the stuff thats legal and related to solar energy that is.

Craig,

First, the extra difficulty and cost in making and mounting the reflector panels may not justify the gains in energy. Dollars per kw. Think long term and add the labor and material costs into your figures and you may find it takes 10 extra years to get your investment back. I could be wrong.

Next, the wind load on your gear and the ratings on your roof need to be factored in to your calculations. The force on twice as much surface area as panels alone is tough to figure, but it will require a really beefy rack or mount at least. This live load is hard to calculate for many reasons, least of which is the turbulence you will get around such a design, face on wind on some panels, edge on on others, wow my brain hurts thinking about the engineering. I could be wrong, but I have helped remove the broken bits of other peoples arrays.

Thirdly, now you have solar panels and reflectors, or snow collectors shall we say? Double or more the dead load for the equipment alone(panels and racks, and reflectors and racks, plus double the weight of snow collected by the panels alone. The reflectors, by their design and required angle, will not shed any snow. A significant amount of weight to figure in for New England.

Forthly, with double the amount of stuff on your roof, there is just about double the amount of things that can go wrong. Screws come loose, bees build nests, birds poop, squirrels chew, cats chase squirrels across reflectors behind the panels, wind causes vibrations, well, which vibrates things. Again, I could be wrong, but I have helped remove the broken bits of other peoples arrays.

Lastly, and perhaps most importantly, unless your panel manufacturer says its OK, most panels do not like to have reflectors. At some point on a clear day you will get too much reflection and overheat the panels no matter how diffuse the reflected light appears to be, the angle and temp will be just right some day. They do not like to have reflectors and are not designed in most cases to handle them. It will shorten their life span and/or decrease their output.

Mark-Please feel free to refute,rebuke,rebuff or just plain dis-regard.

IMHO

[ghurd]

Hi Craig

Try running the numbers with the panel mounted at 55 degrees.

(I'm curious, but software challenged to put it mildly... so I'll skip the frustration)

Really, if the snow sticks at 55 degree angle, there wasn't much power available that day anyway.

I have installed quite a few systems, somewhere in the triple digits, all fixed & non-adjustable. Only one has any snow problems due to strange winds from a strange roof at different heights and angles. The single panel needed installed in the magic spot where all the snow piled up.

BTW, Reverse my location. Cleveland is 60 miles west of me. It snows here.

G-

[GaryGary]

Hi,

Actually, 90 degrees is vertical in this program.

I'm pretty sure about this, since I wrote the program :-)

I guess I could have worded the input form more clearly -- sorry about the confusion.

Horizontal panels in general do well on a year round basis, but don't do so well in the winter -- take a look and the Dec and Jan output.  I still kind of like the vertical in the winter, and horizontal or low tilt in the summer.  Is this feasible?  

I'll have a look around for that reflector data this morning.

Gary

[GaryGary]

Hi Craig,

I posted the info on reflectors earlier, but it seems to have have been eaten.

Briefly, if you put a horizontal reflector in front of a vertical collector panel, you can expect to get a gain in the 30% to 50% area for the winter months.  The reflector can slope at little up or a little down (eg 5 degrees) without much effect.  The reflector should ideally extend out twice as far as the PV panel is high, but a reflector that only goes out as far as the panel is high still gives a worthwhile gain.  

The refelctor could be snow.  

new snow reflectance = 80 to 90%

old snow reflectance = 40 to 70%

Polished alum        = 75 to 95%

The reflector info comes out of The Passive Solar Energy Book, Mazria -- this is available used on Amazon for next to nothing -- pretty good book.

I calculated the year long gain for a panel that is mounted vertically in the winter with a reflector, and then tilted back to a 30 deg tilt in the summer without a reflector.  This gives you high summer gains because of the low tilt, and the reflector with vertical panels gives high winter gains.  Comparing that to a panel tilted at 42 degrees all year without a reflector, the vertical winter, 30 degree summer has about a 12% gain in year long radiation.  In addition, the gain each month is more uniform.  Since PV panels cost about $50 per sqft, if you can do a reflector arrangement for less than $6 per sqft it will be more cost effective than just making the PV panel larger.  Snow is pretty cheap :-)

---

Ghurd was interested in the numbers for a 55 degree tilt:

Month by Month Summary of Sun on Collector

(100% sunny weather)

  Collector Area:        1.0 (sqft)

  Collector Azimuth:     0.0 (deg) measured from South

  Collector Tilt:       55.0 (deg) measured from horiz

  Latitude:             42.5 (deg)

  Altitude above SL:     0.0 (ft) Above Sea Level

Date ----  Sun ----------------  Collector -----------------------

Month Day  Direct   Di-   Total    Direct      Difuse     Total

           Normal   fuse

   1   21   2074     120   2194       1744         95       1839

   2   21   2506     150   2656       2008        118       2126

   3   21   2879     204   3084       2074        161       2235

   4   21   3094     301   3395       1866        237       2103

   5   21   3185     386   3571       1659        304       1963

   6   21   3226     432   3657       1558        340       1898

   7   21   3086     419   3505       1602        330       1932

   8   21   2912     354   3265       1765        278       2044

   9   21   2676     249   2924       1935        196       2130

  10   21   2336     169   2506       1891        133       2024

  11   21   2000     126   2125       1686         99       1785

  12   21   1811     104   1915       1552         82       1634

Sum        31783    3015  34799      21341       2372      23714

Radiation in BTU/day

Only about 4% down from the 42 degree tilt -- pretty good, but if you have very cloudy winters, it might not be too good a plan, in that it cuts the summer gains, and increases the winter gains compared to the 42 degree tilt.

Gary

[ghurd]

Thanks Gary!

G-

[craig110]

Thank you, Gary, for the additional info!  I was thinking of putting the reflective panels out at least 2x and possibly even 3x the height of the panels to catch the low angle of the winter rays, and this matches your suggestion.  I'd love to just leave the panels vertical with the reflective panels in place in the summer as that simplifies the mounting.  That should give around 1.2-1.3x normal solar intensity on the panels and might risk overheating it, so I'd likely have to put a temporary coating on the panels to cut their reflectance.  Even with a 30 degree tilt in the summer, come to think of it, I'd have to "tone down" the reflective panels to keep the PVs from overheating.

(I'd want to either remove or entirely mask the "further out" panels during the summer anyway as I'm very close to an airport and wouldn't want to be reflecting lots of light up in the middle of the day that I know wouldn't come anywhere near the PVs.  And yes, before I implement any reflector-based plan I'll be checking with the FAA.)

Craig

[craig110]

Hi Gary,

Oh, I wasn't saying the program was wrong, only that I was using "vertical" differently than how the program was using it.  I've been talking in terms not of which direction the cells are facing, but the orientation of the back of the panel.  Just to confirm that we're talking about the same configuration, "vertical mounting" to me means that the cells are facing the horizon and that the reflector panels are laying flat and are aimed straight up.

Craig

[zyewdall]

Just my experience.  Our off grid home has the PV mounted at about an 80 degree angle (48 degree latitude).  This sheds most of the snow, and allows the remaining to clear pretty quickly when the sun finally comes out.  The rack is adjustable to tilt them to a flatter angle in the summer.  However, in the 15 years we've had the rack up, we've never adjusted it.  Since the array is sized for wintertime and it is so much sunnier in the summer (plus lower lighting loads), we are charged up by 9am even at the winter tilt angle, and there is no real use for the rest of the power.

If you have a grid-tied PV system, or have larger summer loads than winter loads, this might not be the case.

And we have seen almost 14 amps out of modules rated for only 11.9 amps total, so I would say that snow reflection can have an effect.  Edge of cloud effects also boost it a bit.

[craig110]

Thanks for the comments!  Sounds like the winter snow reflection works nicely for you.

Craig

[Clifford]

I would certainly recommend putting a hinged spring loaded (or weighted) mount on your reflector...  so if you get a couple of inches of snow, the reflector will dump and spring back into place.  Won't help with frost and ice, but it may improve efficiency a bit.

In theory you could also use hinged / spring loaded mounts on the panels.

Just thinking about reflections at morning / evening times.  You will want to design the reflectors so that some of the morning/evening sun still falls on the panels.  Perhaps orienting the panels on the side rather than vertically would help as well as making the reflectors somewhat longer than the panels.

Not sure what happens to your power generation if a single row of cells (serially connected) is completely obscured from light by snow, even while the upper cells are still open.

[craig110]

Hi Clifford,

I didn't want raise the "standing tall" or "laying down" ways of mounting the cells vertically (i.e. aimed at the horizon) for fear of muddying the waters with too many options.  I agree, though, that mounting them "laying down" makes the reflector smaller if one is using a single panel.

(For others who are pondering this -- since the reflectors should extend about 2x the height of the panel and cover much of the arc in front of the panels, the area of the reflector ends up being based on the height and the two bottom horizonal end points on the panel.  The area of the reflector for a single panel is minimized on a "laying down" approach.)

Since most people have more than one panel, here is a fun "homework" question for anyone who is interested in this to ponder:  If you have multiple panels lined up in either the "standing tall" or "laying down" approach with a full complement of reflector panels in place that follow the "2x the height" notion, what is the area of the reflectors that you have to add to add one additional panel in line with the others?

Craig