I discovered the issue I believe, so I thought I'd update all of you who gave me ideas on what to look for (thank you).
And this is something that I stress the importance of all the time... The branch circuit pairs were not properly paired up! All of the L1 conductors went to L1 on the bus and all of the L2 to L2. However, for example, circuit 1 L1 and circuit 2 L2 were on a single 2-pole breaker. And circuit 2 L1 and circuit 1 L2 were on another breaker. Everything worked because it was electrically connected to the right bus bar. However, when there was a previous fault one of the breakers probably didn't trip when it should have as a result of the improper pairings. It was reset a few times and just kept heating up and heating up. This isn't conclusive, but it sure makes sense. It wasn't fun tracking this down with 5 strings and 5 junction boxes on the roof, but once we started continuity testing and finding strange results, it became clear. Jason Szumlanski On Tue, Jan 21, 2020 at 12:10 PM Jason Szumlanski < ja...@floridasolardesigngroup.com> wrote: > (System installed by another contractor...) > > I have been tasked with assisting in the investigation of damage in a > subpanel used to combine inverter output circuits (5 strings of > microinverters). There appears to be no damage to conductors or other > system components. It looks like the damage started at the stabs of the bus > bar where the breakers connect. Picture attached. Here are a few clues: > > - Monitoring indicates that damage occurred overnight a few weeks ago. > That night we had rain as a winter front came through Florida, possibly > lightning. The prior day everything was fine. > - The next morning, only 2 of 5 strings started producing power, but 4 > of 5 strings were reporting data to the Envoy. The two strings that were > reporting but not producing power reported 0Vac and Voc on the DC side of > the microinverters. > - A couple of weeks later, 1 of the 2 strings that was producing power > quit doing so, but continued reporting data. > - Customer discovered damage yesterday. > - About a week before the initial damage apparently manifested itself, > another contractor installed a whole house generator transfer switch on the > line side of the PV interconnection. The generator has never been run > (there is not even a LP fuel source on site yet). I pointed out that the > solar interconnection, which was previously on the supply side, would need > to be moved to the supply side of the generator transfer switch's main > breaker before operation. During the transfer switch installation, the > contractor also switched line 1 and line 2, but that shouldn't > really matter, except for Enphase consumption monitoring, which was messed > up by the swapping of the lines. > > So I'm looking for ideas. I'm wondering if the OCPD would be a likely > place for lightning damage to manifest itself. I can't visually detect any > other damage anywhere else. I can't imagine that the transfer switch > installation would have anything to do with it, but the timing is > interesting. Other than replacing the subpanel and OCPD and firing it back > up (no pun intended), I'm not sure how to approach further investigation. > > Jason Szumlanski > Florida Solar Design Group > > >
_______________________________________________ List sponsored by Redwood Alliance List Address: RE-wrenches@lists.re-wrenches.org Change listserver email address & settings: http://lists.re-wrenches.org/options.cgi/re-wrenches-re-wrenches.org List-Archive: http://www.mail-archive.com/re-wrenches@lists.re-wrenches.org/maillist.html List rules & etiquette: www.re-wrenches.org/etiquette.htm Check out or update participant bios: www.members.re-wrenches.org
