Vertical mounting bi-facial panels

[azdave]

Once bi-facial solar panels cost the same per foot as fencing that will be a great idea!

Seriously, I'm sure there are cases where a solar fence might make sense when conventional won't do but in my opinion, bi-facials panels are a solution looking for a problem.

[J.P.M.]

Yea: SAM. That'll allow user input for the albedo which is fixed in PVWatts at 0.2, and in any case is not only incident on the backside of a bifacial but incident on both sides of a bifacial panel in any and all orientations and from any and all directions, and in any case for more reasons than is possible to talk about here, is nearly impossible to measure much less predict. If by stray backside radiation you are referring to albedo, that component of radiation is always present on a surface of any orientation.

Aside from SAM you can also use a model that can can do a reasonable job on a bifacial array. I'd also suggest a 45 deg. tilt for New England or closer to array location latitude.

Doing as you suggest is probably not an unreasonable 1st approx. for the specific case of E-W facing vertical bifacials but less so for other orientations. Another factor: Most models won't work for array or module tilt angles > 90 degrees, and most go completely haywire when combined with tilts > 90 deg. and azimuths >270 but <90degrees (i.e., the backside of a single face panel tilted in the regular fashion at less than 90 deg. and oriented toward the equator.

I wrote a model about 30 years ago or more that will, for example, model irradiance on a surface of any orientation which used, along with direct irradiance, diffuse fractions of irradiance, and an albedo estimate which used view factor geometry. See a decent heat transfer text that gives radiation heat transfer a reasonable treatment for more info if you're interested. One such tome is by one of my mentors: Kreith, "Principles of Heat Transfer" (ISBN # 0-06-043774), or see Rosenow & Hartnett: "Handbook of Heat Transfer" (ISBN # 0-07-053576-0).The model I did at the time was too complicated for anything less than a mainframe computer back in the day, but the big problem was verification by measurement because the cosine correction factor for most every pyranometer then (and still), as well as the reflection characteristics of most glazing materials at what are usually low incidence angles makes measurement accuracy either too poor or too irrelevant to measure.

Anyway, PVWatts does a fair job for it's intended purpose and it can be bastardized and made to do things it's writers probably never foresaw, such as what you suggest. For the specialized case of a vertical E-W facing array, doing 2 single face runs is probably not terribly far off in terms of output compared to a double faced array, but much less so for other orientations. One example: Even for the subject orientation(s), the cell temp. of the bifacial will be higher than 2 white backed arrays and will need to be accounted for.

All that aside, that's one of the ugliest fences I've ever seen. It's so ugly it wouldn't be fit to enclose barfatorium.

[bob-n]

I saw this article in Solar Builder proposing ground-mounting bi-facial panels vertically. They say it doubles as a fence. The idea fascinates me.

Vertical Reach.jpg
I wondered if it was a good idea in terms of sun utilization and did some quick runs of PVWatts to try to answer this question.

#1: New England location, South facing, 30 degree from horizontal, all else default. 5194 kWh/year.
#2: New England location, East facing, 90 degree from horizontal, all else default. 2832 kWh/year, to model one side of a vertical panel.
#3: New England location, West facing, 90 degree from horizontal, all else default. 2826 kWh/year, to model the other side of a vertical panel.

No model is perfect. This one is far from perfect. For starters, it ignores stray backside radition like ground reflections. But despite the limitations, do you think this is a fair way to compare a standard ground mount (#1) to a horizontal ground mount (the sum of #2 and #3)?

Do you have a better idea for comparing this approach using bi-facial panels to a traditional ground-mount using bi-facial panels?