Mixing DC String Sizes

[Carultch]

As we all know, it is bad design practice to put different string sizes in parallel on the same single MPPT inverter or same MPPT of a system in general. Some manufacturer manuals may use absolute language against it (using words like "must", "have to", and "required"), some may just use suggestive language against it (using words like "should" or "for best performance").

But is it actually a NEC violation or safety issue to have the occasional string that is short a module or two? Directly connected strings of modules, with no optimizers.

The reason I ask, is to determine as a 3rd party reviewer, whether it is mandatory to make the contractor fix errors like this before starting the system, or whether it just means loss of performance that the owner can choose to accept.

 

[GoldDigger]

As we all know, it is bad design practice to put different string sizes in parallel on the same single MPPT inverter or same MPPT of a system in general. Some manufacturer manuals may use absolute language against it (using words like "must", "have to", and "required"), some may just use suggestive language against it (using words like "should" or "for best performance").

But is it actually a NEC violation or safety issue to have the occasional string that is short a module or two? Directly connected strings of modules, with no optimizers.

The reason I ask, is to determine as a 3rd party reviewer, whether it is mandatory to make the contractor fix errors like this before starting the system, or whether it just means loss of performance that the owner can choose to accept.

As long as the strings are protected by the mandatory (for three or more strings) series fuses, there is no safety hazard. The only NEC violation would be violating the equipment manufacturer's instructions.

But the performance hit will be substantial, up to the entire output of the short string.

Oddly enough, though, if the two strings are aimed in different directions (e.g. one West and one East) so that their maximum power production hours do not overlap there may not even be much of a performance penalty.

The general rule of thumb for paralleling either panels or strings is that if the difference in Vmp values is less than 5% the power loss will be minimal.

 

[ggunn]

The reason I ask, is to determine as a 3rd party reviewer, whether it is mandatory to make the contractor fix errors like this before starting the system, or whether it just means loss of performance that the owner can choose to accept.

Well, seeing as when you connect a 12 modules string in parallel with a 10 module string on the same MPPT input it will perform pretty much like two 10 module strings, I think you'd be doing the customer a disservice by not making the installer move one panel to make two 11 panel strings.

 

[GoldDigger]

Well, seeing as when you connect a 12 modules string in parallel with a 10 module string on the same MPPT input it will perform pretty much like two 10 module strings, I think you'd be doing the customer a disservice by not making the installer move one panel to make two 11 panel strings.

Actually, depending on the MPPT algorithm used, there is also a good chance that it will behave like one 12 module string.
I agree that it should be corrected except possibly if the two strings have different orientations.

 

[jaggedben]

I have a case in point right now that raises a related question for me...

Do different string lengths have any effect on potential induced degradation (PID)?
Specifically, say, for Sunpower modules from a few years ago?

 

[kangaroogod]

Out of curiosity, if there were 3 strings with 11,10,12 that met at a combiner, how would the output be affected?

Sent from my SM-N920V using Tapatalk

 

[GoldDigger]

Out of curiosity, if there were 3 strings with 11,10,12 that met at a combiner, how would the output be affected?

Sent from my SM-N920V using Tapatalk

The MPPT mechanism of the inverter or charge controller that they are feeding will not be able to optimize the output of all three strings at the same time.
Exact details depend on the MPPT algorithm used, but most likely the 11 panel string will be running at close to optimal load while the 10 panel string will be operating with current below Imp and voltage above Vmp while the 12 panel string will be operating at current above Imp and voltage below Vmp.

In the worst case situation, if the operating point is close to the Vmp of the 12 panel string, the voltage may be above Voc of the ten panel string leaving it producing no power at all.