Trackers in backtracking on positive (south) slopes vs. on flat terrain

[Michele]

Hi,

I'd like to share a strange result that I really didn't expect and don't understand at all...

I ran two simulations with trackers and monofacial modules: one on flat terrain (for simplicity using the mode "unlimited HSAT") and another using a 3D scene with slopes imported in .pvc.

Both in backtracking, keeping all the remaining simulation's parameters the same up to the MPP input stage.

As you can see in the screenshot below, the flat scenario performs better than the sloped one.

Starting from Gpoa, the main difference is due to a HUGE -2.61% of electrical mismatch effect (~half of temperature losses), that decrease the 37.88 GWh (higher than the opposite 37.29 GWh thanks to lower AOI basically) to 34.32 GWh (lower than the opposite 34.72 GWh).

Now, the strange thing is that the topography is towards south (favourable) and almost regular (assimilable ideally to an inclined plane).

The GCR doesn't vary much in the backtracking management tab (52%), so the backtracking should work fine.

The question is: how is it possible (with backtracking, on an almost regular topography) such a underperformance (-2.61% on yearly basis) even if the corresponding shading table has some non-zero values only for very low solar altitudes and/or high azimut (sunrise/sunset). In these cases the irradiation should be low and then its weight in the overall mismatch contribution, no?

Hope the doubt is clear and I'd be very glad if someone can clarify or share their thougths about this.

Regards

Michele

Edited September 1, 2022 by Michele

[dtarin]

What does the module strings table look like? This table confirms the near shading loss. For confirming electrical loss, run the shading analysis in the shade scene and it will graphically show the electrical effect losses, can help get an understanding of why your elec. eff is high. 

Edited September 1, 2022 by dtarin

[dtarin]

Also, simulate your "Flat" scenario according to strings. It looks like you have modeled it with linear shadings, so this isn't a good comparison.

[Michele]

Hi dtarin,

thanks for your reply!

1 hour ago, dtarin said:

Also, simulate your "Flat" scenario according to strings. It looks like you have modeled it with linear shadings, so this isn't a good comparison.

I tried just to be sure but nothing changes...For a 20 MWp plant, simulating horizontal trackers in backtracking with a 3D scene or not (unlimited HSAT) doens't make any appreciable difference in results. This is because the whole shading factor table for beam will be always zero in both cases thanks to the (ideal) perfect backtracking and because the edge effects may be neglected for medium/big plants.

 

1 hour ago, dtarin said:

What does the module strings table look like? This table confirms the near shading loss. For confirming electrical loss, run the shading analysis in the shade scene and it will graphically show the electrical effect losses, can help get an understanding of why your elec. eff is high. 

The module strings table shows greater values.

I was interested in the linear shading table because that's the starting point in any case, trying to figure out how those relatively low geometric shading factors (shaded area/total area) can produce such a high loss, when the irradiation is expected to be low.

Anyway, I'll try to play with the tool you suggest...thanks!

Edited September 1, 2022 by Michele

[dtarin]

6 hours ago, Michele said:

Hi dtarin,

thanks for your reply!

I tried just to be sure but nothing changes...For a 20 MWp plant, simulating horizontal trackers in backtracking with a 3D scene or not (unlimited HSAT) doens't make any appreciable difference in results. This is because the whole shading factor table for beam will be always zero in both cases thanks to the (ideal) perfect backtracking and because the edge effects may be neglected for medium/big plants.

 

I see, yes, if perfectly flat, there will not be. In the image above, I thought your "LOT 1 - Flat" was the "flat" case you were considering. It looks like those modules are at different elevations (if even by a little), so there would be an electrical effect loss. There really are no completely flat sites, and ones that are very close will still have some electrical effect depending on the design. 

Edited September 2, 2022 by dtarin