Bifacial panel field testing?

[bcroe]

Although bifacial panels are currently odd ducks, several forces
are converging that might make them more mainstream

If bifacial panels are real, i would expect one to be used in a situation like this to replace
2 panels. An alternate to tracking. Building the support without shadows would be a challenge.
Bruce Roe

[DanKegel]

Yeah, that's an example of where they'd be good-ish.

There's only one vendor selling 300 watt - class 60 cell modules at the moment, and they're freaky amorphous-monocrystal hybrids.
List price isn't overly high
( http://www.civicsolar.com/product/su...al-solar-panel $412 vs.
http://www.civicsolar.com/product/lg...ht-solar-panel $414 vs.
http://www.civicsolar.com/product/lg...ht-solar-panel $440 vs
http://www.civicsolar.com/product/lg...ht-solar-panel $450 )
but one has to imagine that LG could offer something similar by just coming out with a version of
the LG305 with a clear back panel, and SolarWorld says they're introducing one in 3-6 months:

http://www.solarworld.de/en/group/press/press-releases/press-release/single-press-releases/article/innovations-on-the-intersolar/

So let's check back in a year; once more vendors offer bifacial, people will probably figure out where they're worth using, and start using them there.

[SunEagle]

Yeah, that's an example of where they'd be good-ish.

There's only one vendor selling 300 watt - class 60 cell modules at the moment, and they're freaky amorphous-monocrystal hybrids.
List price isn't overly high
( http://www.civicsolar.com/product/su...al-solar-panel $412 vs.
http://www.civicsolar.com/product/lg...ht-solar-panel $414 vs.
http://www.civicsolar.com/product/lg...ht-solar-panel $440 vs
http://www.civicsolar.com/product/lg...ht-solar-panel $450 )
but one has to imagine that LG could offer something similar by just coming out with a version of
the LG305 with a clear back panel, and SolarWorld says they're introducing one in 3-6 months:

http://www.solarworld.de/en/group/press/press-releases/press-release/single-press-releases/article/innovations-on-the-intersolar/

So let's check back in a year; once more vendors offer bifacial, people will probably figure out where they're worth using, and start using them there.

I didn't see any special instruction concerning the installation of those Sunpreme panels to get the claimed 20% increase in production. Just a 60mm minimum height above the roof but not a maximum.

Also does the color and type of roof material beneath those panels make a difference in the production increase?

[DanKegel]

I didn't see any special instruction concerning the installation of those Sunpreme panels to get the claimed 20% increase in production. Just a 60mm minimum height above the roof but not a maximum.

Also does the color and type of roof material beneath those panels make a difference in the production increase?

All that matters is how much light hits the backside, I think. Getting it up high, and having a general background color close to white, both help.

http://www.sunpreme.com/products/gxb-series/ just says "high albedo", but other vendors provide a little more info on how their panels behave vs. height and albedo.

Table 3 in http://www.prismsolar.com/pdf/Design_guide.pdf is the best example I've seen. (Higher was definitely better; height can compensate a bit for lower albedo.)

The graph at the bottom of http://www.panasonic.com/business/pe...ion_Manual.pdf for
Sanyo's HIT bifacial modules shows a 15% boost at a mounting height of one meter with a 55% reflective surface on the ground (concrete).

And finally, http://libres.uncg.edu/ir/asu/f/Scia...012_Thesis.pdf carefully measured flush mount of bifacial and found that it was not a good idea

[SunEagle]

All that matters is how much light hits the backside, I think. Getting it up high, and having a general background color close to white, both help.

http://www.sunpreme.com/products/gxb-series/ just says "high albedo", but other vendors provide a little more info on how their panels behave vs. height and albedo.

Table 3 in http://www.prismsolar.com/pdf/Design_guide.pdf is the best example I've seen. (Higher was definitely better; height can compensate a bit for lower albedo.)

The graph at the bottom of http://www.panasonic.com/business/pe...ion_Manual.pdf for
Sanyo's HIT bifacial modules shows a 15% boost at a mounting height of one meter with a 55% reflective surface on the ground (concrete).

And finally, http://libres.uncg.edu/ir/asu/f/Scia...012_Thesis.pdf carefully measured flush mount of bifacial and found that it was not a good idea

Ah. I was only looking at the Sunpreme "installation manual". That Prismsolar Design guide and data from Panasonic makes it more clear. Thanks.

[J.P.M.]

I didn't see any special instruction concerning the installation of those Sunpreme panels to get the claimed 20% increase in production. Just a 60mm minimum height above the roof but not a maximum.

Also does the color and type of roof material beneath those panels make a difference in the production increase?

1.) I ran some #'s on ESTIMATED increased P.O.A irradiation on a 2 sided surface tilted at 30 degrees (in the usual sense - facing south) vs. a 1 sided surface at the same orientation. The algorithm is my own, and pukes out what I believe are reasonable #'s for any azimuth, and also tilt angles greater than 90 degrees. The #'s from my stuff agree quite well w/SAM for normal orientations and 1 sided panels. To the best of my knowledge, Sam does not do 2 sided (bifacial) panels at this time.

2.) At location 92026 using TMY3 Miramar data and zero albedo, the increase in yearlong total P.O.A.irradiation on a 2 sided vs. a 1 sided surface was about 2.1%.

3.) The ESTIMATED increase in yearly output with zero albedo for the same conditions for a 2 sided array vs. a 1 sided array with Sunpower 327 panel eff., optical and temp. characteristics was about 1.9%. I'd think the diff. in output vs. irradiance might be due to at least 2 things: increased panel temps. due to slight irradiance increase, and the (assumed) nature of the backside irradiance being mostly diffuse in nature and thus at a generally higher incidence angle for transmission calculation purposes and thus lower transmitted values.

4.) As an approx. way to estimate what happens with increased (>zero) albedo:

Every 1% increase in albedo increased both the 1 and 2 sided conditions annual irradiation by approx. .0006, and increased system output by ~~ .0004 to .0005.

So, for example, at 30% albedo and 30 deg. tilt, the P.O.A. irradiance for both 1 and 2 sided conditions increased, very approx. by : (30) X (.0006) = 1.8% over the zero albedo condition.

The output increase for both 1 and 2 sided conditions increased by ~~ : (30) X (.00045) ~ = 1.4% over the zero albedo condition.

To be clear, the above conditions are not the same as deliberately, and as part of a design, having diffuse or specular reflectors directly in front of/behind/around a surface or array. That's different and not considered here.

5.) As the tilt angle changes, the above estimates change. As the albedo changes, the estimates also change. There is also a synergistic interaction between albedo changes and tilt angle variation. Example: the albedo increase on an array mounted parallel to a roof with perfect specular reflectors (conditions of 100% albedo with a view factor of ~ 1) mounted under an array composed of 2 sided panels won't get much, if any, increase in either irradiance or output. An array composed of 2 sided panels oriented vertical and "south/north" facing in a snow covered environment will do better. How much better is left as an exercise for the student.

6.) I'm not claiming any more than that the above is an estimate based on my own studies. I appreciate comment, objective and constructive criticism and discussion. However, I'm not going to get dragged into a non productive and ignorantly pedantic bunch of "why daddy" type of interminable questions that go nowhere. I'm aware that my # seem to vary quite a bit and run contrary to some mfg. sales literature. To those who cherry pick sales lit. or agenda based published stuff to serve as justification for their opinions, I'd suggest following the money. I'd also suggest a serious literature search will find a lot of data and information that's a lot closer to what I've found and learned over an engineering career than what 2 sided panel mfgs. seem to imply.

Take what you may want/need of the above. Scrap the rest.

[bcroe]

1.) I ran some #'s

So, for example, at 30% albedo and 30 deg. tilt, the P.O.A. irradiance for both 1 and 2 sided conditions increased, very approx. by : (30) X (.0006) = 1.8% over the zero albedo condition.

The output increase for both 1 and 2 sided conditions increased by ~~ : (30) X (.00045) ~ = 1.4% over the zero albedo condition.

To be clear, the above conditions are not the same as deliberately, and as part of a design, having diffuse or specular reflectors directly in front of/behind/around a surface or array. That's different and not considered here.

Those are the sort of numbers I would expect. My thinking this is, hardly worth much cost
as an upgrade to a standard system. Unless some new array arrangements are to be used,
bifacial panels need to be a nearly no cost upgrade. I did see some claims of 20% increase
which I don't believe; I have seen outputs of under 10% from diffused light. Bruce Roe

[SunEagle]

Those are the sort of numbers I would expect. My thinking this is, hardly worth much cost
as an upgrade to a standard system. Unless some new array arrangements are to be used,
bifacial panels need to be a nearly no cost upgrade. I did see some claims of 20% increase
which I don't believe; I have seen outputs of under 10% from diffused light. Bruce Roe

I agree. J.P.M. calculations seems more appropriate for the expected % increase of output of those panels.

There is always the chance that using light "magnification" or "concentration" will get that extra 20% but it has to be simple and easy to install.

[J.P.M.]

I agree. J.P.M. calculations seems more appropriate for the expected % increase of output of those panels.

There is always the chance that using light "magnification" or "concentration" will get that extra 20% but it has to be simple and easy to install.

Booster specular reflectors as "wings" or in front of arrays, or 1 axis tracking w/attached reflectors are two possibilities to get measurable increases in output. Usually however, the economics don't work out too well. Same for the many practical and design considerations. If they did, I'd expect more solar farms would have reflectors on the arrays.

Low concentration of solar energy (conc. ratios <~1.25/2 or so) as is possible w/ singe/2 flat reflectors is probably more appropriate for low/med. temp. solar thermal process heat applications. Long, single row thermal arrays tilted at lat. w/ single fixed specular reflec. at the bottom of same dimension as array will have about 10-15% annual output boost. Been there, done that.

[DanKegel]

I doubt reflectors or anything like that is worth it. A 10-15% boost from diffuse lighting for a panel mounted above a high-albedo surface (like snow or concrete) seems about the best one can hope for.

[SunEagle]

I doubt reflectors or anything like that is worth it. A 10-15% boost from diffuse lighting for a panel mounted above a high-albedo surface (like snow or concrete) seems about the best one can hope for.

Which is why most of the research to get more production out of a cell revolves around the type of materials used or the arrangement.

The multi-junction or stacked cells are not easy to make but have shown a tremendous potential to get > 47% (in the lab) compared to the 20's of conventional cells. Maybe some day they will find a better way to build those cells.

[DanKegel]

Nothing stopping heterojunction cells from being bifacial

[SunEagle]

Nothing stopping heterojunction cells from being bifacial

True. Which is why I have not given up on someone finding a better design for bifacial.

[J.P.M.]

Hetero panels may have some future, but an e-w aligned ( facing north and south) hetero bifacial panel with ~50% efficiency still has a P.O.A. irradiance on the north side that is about 2-3% or so of the south side and thus will not produce more than about 2-3% more than the single face panel. Until someone repeals entropy, that's just the way it is.

For most sensible or common orientations, think of them for the most part as north facing panels on the back of south facing panels. They have the worst azimuth at a disastrously and nonsensically bad tilt angle, worse if an array is parallel and close to its mounting surface - like a roof.

Or, another situation of many possible: If aligned n-s (east-west facing) with vert. orientation, depending on latitude, output will be about 80% or so as much output as a south facing single face panel tilted at local latitude for twice as much surface.

In general, solar panels produce the most annual output and operate most cost effectively when orientated to maximize the integrated annual positive energy flux (annual total P.O.A. irradiance). Bifacial panels are a poor idea mostly because they inherently compromise and limit that ability to use panel orientation to maximize that input - basically, limiting flux density to some less than max. value by design. Max. out solar flux on one side and minimize the other side, or get mediocrity by making annual P.O.A irradiance equal for each side.

[DanKegel]

an e-w aligned ( facing north and south) hetero bifacial panel with ~50% efficiency still has a P.O.A. irradiance on the north side that is about 2-3% or so of the south side and thus will not produce more than about 2-3% more than the single face panel. Until someone repeals entropy, that's just the way it is.

Is that 2-3% for a dark background, or a light one?