[RE-wrenches] Lightning Strike Fries Array

[Gary Willett]

Wayne, et al:

I also recommend that you perform a Megger test on the DC Source Circuit 
conductors before recommissioning the system. There could be wiring 
faults caused by the lightning strike that may go undetected.

It's also possible that an undetected ground fault might have occurred 
during initial construction. The presence of a longitudinal imbalance on 
the PV conductors could make the equipment more susceptible to lightning 
induced voltage.

I was contracted to troubleshoot and determine the extent of damage for 
a lightning strike on a 21 module (3-strings of 7) ground mount system 
located about 180' from the garage where the string inverter was located.

The lightning irreparably fried the 7 modules in the middle landscape 
oriented string (furthest from the N-S support posts) and the string 
inverter. The other two strings were unaffected.

When I did a Megger test on the PV conductors for all 3 strings, I found 
that the home run wiring for the lightning damaged middle string had a 
4.5 k-Ohm ground fault on the PV+ conductor to ground. The ground fault 
resistance was too high to blow the inverter's 1A GFDI fuse.

If left uncorrected, the ground fault would have seriously affected the 
performance of this string. I measured the performance of this string 
with my Solmetric PVA1000 analyzer after the damaged modules were 
replaced. The ground fault caused a 44% reduction in PV production as 
compared to the other two strings.

It was not possible to replace just the faulty conductor without 
damaging the other wires in the buried conduit. The solution was to 
install the replacement inverter at the ground mounted array, then reuse 
the remaining good wiring for the inverter's PV output circuit.

Regards,

Gary Willett PE
Icarus Solar Services / Engineering
Irving TX USA


On 5/21/15 7:35 PM, Richard L Ratico wrote:
> Wayne,
>
> Never had an entire array fried. My experience consists of a total of less than
> 10 modules out of a total of maybe 50 at two different sites. One site had the
> much more common (these days) modules with leads and diodes potted in an
> epoxy-like material. We use a Dremel tool with a burr bit to access the diodes.
> Pretty labor intensive, but modules of the right size were no longer available.
>
> If the junction boxes have removeable covers, that's half the battle won. You
> can test the individual diodes and cell strings. Diodes are inexpensive and not
> too hard to replace. I think we got them through Mouser.
>
> Dick Ratico
> Solarwind Electric
>
>
>
>
>
> --- You wrote:
> Hi Wayne,
>
> Several years back I had all sixteen SolarWorld modules on a tracker get
> zapped.  The two other trackers adjacent to this one had no damaged
> modules.  The same strike blew out the phone box at the end of the road
> (some distance away).
>
> The SolarWorld jboxes were all sealed with no way to get to the diodes.
> The homeowners insurance covered the loss and I replaced the modules with
> something similar (SW had since stopped manufacturing 175s).  It would have
> been nice to have had access to the diodes.
>
>