Shading: Complex Arrays at the String Level

Hello Paul,

I am trying to figure out how to model complex arrays in the 3D shading calculator. Let's pretend I have an array with 48 strings equally divided into two sub arrays.

First of all - if I only create active surfaces for one string per subarray in the 3D calculator, are those shading values used for the shading in the rest of the strings in their respective subarrays? I have created the table below to illustrate how I think it currently works. Please correct it if I am wrong.

Subarray

---

String

---

Shading from 3D Calculator
1

---

1

---

Subarray 1, string 1
1

---

2

---

Subarray 1, string 1
1

---

3

---

Subarray 1, string 1
...

---

...

---

...
2

---

1

---

Subarray 2, string 1
2

---

2

---

Subarray 2, string 1
2

---

3

---

Subarray 2, string 1
...

---

...

---

...


2nd - if I create two active surfaces for two different strings per subarray, how are the calculated shading values distributed to the remaining strings?

Subarray

---

String

---

Shading from 3D Calculator
1

---

1

---

Subarray 1, string 1
1

---

2

---

Subarray 1, string 2
1

---

3

---

?
...

---

...

---

...
2

---

1

---

Subarray 2, string 1
2

---

2

---

Subarray 2, string 2
2

---

3

---

?
...

---

...

---

...

Regards,
David

Paul - can you give me some insight on this?

Thanks,
David

Hi David,

The Subarray property of an active surface in the 3D shade calculator determines the subarray in SAM to which the shading losses are assigned. If two or more active surfaces in the shade calculator have the same subarray number, the active surfaces are treated as a single subarray. SAM only allows up to 4 subarrays, so you should only use up to 4 subarrays in the shade calculator.

Strings in the shade calculator only come in to play in SAM when you use the "Database lookup" option in SAM. If you assign the same subarray number to three active surfaces, and assign a different string number to each surface, SAM will calculate a DC shading loss using an empirical lookup model to estimate the effect of partial shading on the subarray's DC output. That is valid for crystalline silicon modules with 3 bypass diodes, and for subarrays with up to 8 strings.

For your examples above with the "Database lookup" option enabled in SAM, the first case would model the effect of partial shading for each subarray as a single string. The second case would account for the array having more than one string.

You can read more about the "Database lookup" model in Deline et. al. (2015) "Partial-Shading Assessment of Photovoltaic Installations via Module-Level Monitoring," which you can download from the Performance Model documentation page:

sam.nrel.gov/performance

If you use one of the other options "Average of strings," "Minimum of strings," etc., then SAM will calculate the shading loss on the subarray in each time step by averaging, taking the minimum, etc. of the string shading losses instead of using the database lookup method.

Finally, we are reworking the user interface and revising the Help system to better explain all of this for the next version of SAM, which we hope to release later this month.

Best regards,
Paul.

Paul,

The link for the document you specified above is broken, as it points here: sam.nrel.gov/performance

Which is, of course, right back to the page it is on. Could you please provide a valid link to the document?

Thanks!
David

Thanks,
David

Oops. Thanks for pointing out the broken link. I fixed that, and also changed the heading to "Partial Shading Model."

Best regards,
Paul.