Backtracking system provide less yield than without backtracking. Is it reasonable?

[bkn]

Hello, guys

I would like to ask you why backtracking give less yield than the one without backtracking in the simulation report of PVsyst 6.7.8.

I selected "no shadings" in "Near shadings" box in "Optional" part.

See below the attached pictures. My question are below with bold characters.

[Monofacial case. Left:without backtracking / right: with backtracking]

 

[bifacial case. Left:without backtracking / right: with backtracking]

 

Some doubts cased by the statements below:

1. In PVsyst help says

"It should be noted that the Backtracking doesn't increase the total irradiance received. It only improves the electrical loss effects of the shadings."

-> But, the report (the pictures attached) say the opposite: the system without backtracking has higher "Global incident on collection plane". Am I mistaken some part here?

2. Backtracking's advantage

"Though the angle of the panels is not optimal, the loss from the off-angle is typically less than the loss that would result from shading the panels, added John Williamson, director of engineering at Array Technologies."

Link: https://www.solarpowerworldonline.com/2015/07/how-does-solar-backtracking-make-projects-more-productive/

-> "Global incident collector plane" is higher w/o backtracking. Although, it has higher shadings loss such as "irradiance loss" and "electrical loss" compared to the one with backtracking, total energy yield present that the backtracking system is not effective strategy to improve the energy yield. Is it reasonable conclusion according to the simulation report? Am I missing something here? Simply, I cannot understand why it happen like this.

Could anyone can guide me to figure out this phenomenon, please?

Thanks in advance

[dtarin]

I selected "no shadings" in "Near shadings" box in "Optional" part.

 

Did you model with the unlimited sheds method in the Orientation menu? The electrical effect loss is not shown in the backtracking version's waterfall. I would double check you have modeled shadings properly for both cases. Even when electrical loss is zero, it still displays in the waterfall in v6.78. Either using the unlimited sheds method, or constructing a shade scene and using the "according to module strings" option (more accurate method).

 

Some doubts cased by the statements below:

1. In PVsyst help says

"It should be noted that the Backtracking doesn't increase the total irradiance received. It only improves the electrical loss effects of the shadings."

-> But, the report (the pictures attached) say the opposite: the system without backtracking has higher "Global incident on collection plane". Am I mistaken some part here?

 

These two statements are in agreement and say the same thing. Backtracking reduces POA and lowers shading losses. The reports confirm this. "Total irradiance received" = global incident in collector plane.

 

2. Backtracking's advantage

"Though the angle of the panels is not optimal, the loss from the off-angle is typically less than the loss that would result from shading the panels, added John Williamson, director of engineering at Array Technologies."

Link: https://www.solarpowerworldonline.com/2015/07/how-does-solar-backtracking-make-projects-more-productive/

-> "Global incident collector plane" is higher w/o backtracking. Although, it has higher shadings loss such as "irradiance loss" and "electrical loss" compared to the one with backtracking, total energy yield present that the backtracking system is not effective strategy to improve the energy yield. Is it reasonable conclusion according to the simulation report? Am I missing something here? Simply, I cannot understand why it happen like this.

Could anyone can guide me to figure out this phenomenon, please?

Thanks in advance

 

Make sure you have properly modeled these with regards to shading. If you have a very low GCR, it is possible your shading losses without backtracking are such that the POA gain is higher than the shading loss incurred. Location may be a factor, you have very high irradiance; might be something unique to your location.

Edited March 13, 2019 by dtarin

[bkn]

Hello, Solarguru

 

Did you model with the unlimited sheds method in the Orientation menu? The electrical effect loss is not shown in the backtracking version's waterfall. I would double check you have modeled shadings properly for both cases. Even when electrical loss is zero, it still displays in the waterfall in v6.78. Either using the unlimited sheds method, or constructing a shade scene and using the "according to module strings" option (more accurate method).

 

In the "Orientation" menu, I selected "Horiz. axis unlimited trackers" in "Field type". Due to this, I cannot opt for neither "Linear shading" nor "According to module strings" in "Near shadings" box. (As soon as I draw a construction in "Construction/Perspective", it changes the orientation setup in "Field type" automatically.)

And I assume that the electrical loss get influenced only by "backtracking" option in my case since I did not change nothing but "backtracking" in orientation tab.

 

These two statements are in agreement and say the same thing. Backtracking reduces POA but lowers shading losses. The reports confirm this. "Total irradiance received" = global incident in collector plane.

I understood "It should be noted that the Backtracking doesn't increase the total irradiance received." can be translated into "Backtracking reduces POA" confirmed by the simulation report.

 

Make sure you have properly modeled these with regards to shading. If you have a very low GCR, it is possible your shading losses without backtracking are such that the POA gain is higher than the shading loss incurred. Location may be a factor, you have very high irradiance; might be something unique to your location.

 

Any suggestion to correct the shading setup in my simulation? ("Horiz. axis unlimited trackers"). The GCR value applied to all the reports was 41.7%.

I tested the annual energy yield comparsions by the following and found out some interesting arguments:

Unit: MWh/yr

GCR /without backtracking/ with backtracking/ Diff.

Mono

83,30%/ 29,6 / 29,5 / 0,1

41,70%/ 34,8 / 34,6 / 0,2

20,80%/ 36,3 / 36,4 / -0,1

13,90% /36,8 / 36,9 / -0,1

Bifacial

83,30%/ 34,5 / 32,2 / 2,3

41,70%/ 38 / 37,1 / 0,9

20,80%/ 39,2 / 38,9 / 0,3

13,90%/ 39,5 / 39,4 / 0,1

By this result, backtracking application depends on at least GCR and geo-dependent climate whether apply or not. In my case, bifacial with backtracking wouldn't be effective as other places, would it?

And regarding to the location-dependency, I will check out further.

Thanks for your help, Solarguru

[dtarin]

In NA, backtracking with bifacial yields higher than non backtracking I believe in most cases I've seen (c-Si module). I did a test on a site of mine in Southern US; backtracking was 1% higher in total energy than non backtracking. Near shading loss on a 39% GCR without backtracking was ~6%.

In seeing your high irradiance, I picked a different location (Chile) with similar yearly GHI, and this trend was also present for this region. 39% GCR, one module in landscape, using the 3D shade scene for calculations and not unlimited sheds. This trend would be the same for monofacial. I would suggest building out your model in the 3D shade scene and ensure your inputs are correct and the same for both models (pitch, etc).

[bkn]

Hello, Solarguru

 

Near shading loss on a 39% GCR without backtracking was ~6%.

 

1. What you meant by "~6%"corresponds to "Near shadings: irradiance loss", didn't you? What I found out a reply from PVsyst team regarding "Near shadings:irradiance loss" is the following:

"The shading is accounted for the beam component (the shade what you see), but also on the diffuse and albedo components, which are integrals of the shading factor over all the directions "seen"by the PV module."

So to clarify, with backtracking can decrease "Near shading:irradiance loss" and do not cause "Shadings: electrical loss". Is it reasonable think in this way?

2. Taking into account your suggestion, the selection of "using the 3D scene for calculations" - which I should turn on "Near Shadings" in "Optional tab", the PVsyst doesn't allow me to construct 3D scene with "Horiz. axis, unlimited sheds." in "Field type" in "Orientation"box. The thing is that I am doing a simulation with bifacial case then I cannot use another "Field Type" such as "Tracking horizontal axis E-W"and "Tracking tilted or horiz. N-S axis" which ALLOW me to construct 3D scene in "Near shadings" box. Can I ask you how and what setup you used to do a bifacial simulation with 3D shade scene?

 

3. I, manually, calculated the loss with your waterfall.Simply from "455193 MWh"(Backtracking, at the second stage of waterfall), total loss turn out -20.8% (-0.1-6.6+0.4-1.5-1.2-0.4-1.5-2.3-7.6 = -20.8). 455193 * (1-0.208) = 360512 MWh which is not equal to "366997 MWh" indicated value. Why does it happen like this?

Thanks for your help, solarguru

[dtarin]

1. Yes. Yes.

2. Select tracking tilted or horiz. N-S axis option under orientation. Under system, select bifacial system above the pan file and select use unlimited trackers for 2D-model. Enter your inputs. Define system (inverters, strings, etc). Go to near shadings, construct a tracker block which matches your orientation settings (tilt if any, tracking range, backtracking/no backtracking). Ensure the tracker block matches your design in terms of module orientation, width, pitch, and representative length of 1 or more strings. Define your partition. Save, exit. Select according to module strings under near shadings. Go back to the bifacial system settings and ensure your inputs match the shade scene: pitch, width, tracking range, etc.

3. The waterfall should be calculated sequentially. In addition, the numbers are rounded (only showing one decimal place). You are not likely able to reproduce the result based on the waterfall. When calculated sequentially I get 367,533.5 (backtracking report), which is due to the rounding of losses in the waterfall.

[bkn]

Hello, solarguru

2. Doubt here. What would be the major difference between "Horiz. axis, unlimited trackers" vs "Tracking tilted or horiz. N-S axis" when using bifacial modules? (I already posted one question in the forum today). When to use each filed type for simulation? I've always used "Unlimited" one in fix and tracker when doing bifacial simulation without being critically considering the difference and now I would like to fully understand about it. Could you explain to me, please?

3. You are right, I totally got mistaken the way of calculation of loss.

Thanks for your great help, it actually help a lot to improve my usage of PVsyst.

Have a great day!

[dtarin]

The bifacial calculation presently only uses the unlimited sheds method (in the bifacial settings), but I believe this will change in the future. The use of a 3D shade scene is for shading loss calculations, and is more accurate than the unlimited sheds method (whether it's bifacial or not). It also allows you to include nearby shading object such as trees or poles if present on your site.

[oyautc]

Hello Andre;

I did two analyzes at the same location, same bifacial panels with the tracker system. Only the inverter was different. However; There are 2 different Global irradiation and sky-diffuse on rear side values in the analysis. Why the Global irradiation and sky-diffuse values are different?

Thanks.

[bkn]

Hello, solarguru

Thanks for your comment, I will try to draw 3D shading scene to check out backtracking issue again.

Have a good day!