Partitioning for the Electrical calculation with String Inverters

[Georgy PM]

Hello PVSYST team,

I have a couple of question regarding the partitioning according to your guide (for the new version 7.4, June 2023). In the summary for common cases, you indicate how to make the partition in the shading scene when one has 2 or more rows of modules in parallel.

I want to confirm with you, how many partitions should I apply when the strings are not in parallel vertically, but horizontally and connected to independent MPPT of a string inverter (1500Vdc system)? Here some of my cases:

1.       Half cut module in portrait, 20 to 28 modules horizontally in series, 2 or 3 rows per rack. Each row is connected to an independent MPPT of the string inverter. According to the guide, I understand that I should apply 4 partitions (if 2V, 6 partitions if 3V), is it correct?

2.       Same configuration before, but with modules in landscape instead. Following the guide, I should apply 2 partitions (if 2V, 3 partitions if 3V). Is it ok?

3.       For the same configurations, but with Bifacial First Solar modules instead (6 modules in series, horizontally cabled). How should it be done the partitioning?

4.       When having the modules cabled in U and 2 string in parallel (Half Cut, portrait), I understand that I should apply 2 partitions (if 2V, 3 partitions if 3V). This also applies when assigning the strings to independents MPPT?

5.       When having half cut modules in landscape, 3x partitions (6 in this case) should be applied for modules connected in U. When they are cabled horizontally, x partitions should be applied (2 in this case). This also applies when assigning the strings to independents MPPT?

When having racks with both U and horizontals strings assigned to the same inverter, I get the partition shown below. However, this seems to me that I should expect less shading losses when cabling in U, than when cabling horizontally. But electrically does not seem logical to me because when the bottom modules are shaded in the U case, the current of all the string is affected. I the second case, only the shaded row is affected. Hence, I would expect more shading losses when cabling in U.

In addition, note that when partitioning a module in 2 or more, the following warning message is shown below. The message should be ignored in that case?

6.       When 2 or more rows of strings are cabled in parallel (6 modules in series), what should be the partitioning for the Bifacial First Solar modules? 

Guide link : https://www.pvsyst.com/help/shadings_partitioninstrings.htm 

Many thanks in advance for your help,

Best regards,

Georgy

[Michele Oliosi]

1) That is correct. You can also adapt the number of partitions in length (X). If there are two strings side by side then you can put 2 partitions in X.

2) No in fact what matters here is that the strings in parallel on a given MPPT are on the same level, so probably will experience the same shading. In this case you should put 2 partitions in Y for 1 V, 4 partitions for 2 V, 6 for 3V, as for 1). In the help this case is named "1L". As above you can put 2 partitions in X.

3) First solar modules are somewhat resilient to shading due to the submodule structure. It is recommended to use "Linear shadings" which doesn't need any partitioning.

4) Yes by assigning strings to different MPPTs there is no change. Since there are two strings in X you can put 2 partitions in X.

5) When cabled horizontally, this is like case 2) i.e. 4 partitions.
MPPT 1 in your drawing: in general mutual shading affects 1/6th of the submodules. If the voltage range of the inverter allows for it, the string can just function with 5/6 of the voltage (and power) by bypassing the shaded submodules. MPPT 3 in the drawing: in general mutual shading affects 1/3 of the submodules. This situation is less advantageous and produces a bit more losses. Note that MPPT 2 will be generally unshaded.
You can ignore the warning, you are right we should update that.

6) same as 3)