Generally, MC 4 connectors from  reputable makes (Stäubli come to mind, there are many other makes) have an environmental rating of IP67 but only when the correct cable diameter is used and the glands are done up as per the instructions. This means they are then fine for short periods of immersion. However,  generally cables and connectors left in the sun and the wind and the rain, will not last as long as the same part under some sort of shelter, even if it is just tucked under the panel, and I'd certainly expect things to be neatly tied up and not flapping about in the open or able to be trodden on or tripped over.

Apart from mis- assembly , as alluded above, the other problem  is if they have been mated while damp, as then that sealing works against you and the ingredients for corrosion are sealed in, instead of out.

What sort of problems are you seeing - it is quite possible that the connectors are not the real problem, but rather like seeing mismatched screw heads on door handles can be a flag to check the rest of the woodwork, a sloppy approach to what you can see, may indicate the same (lack of) attention to detail in other areas less obvious.

Mike.

Hi Mike, 

Thanks for your reply. I have attached photos. 
Basically, inverter is picking up loads of issues. 

I am starting to wonder if the installations were in fact correct or not? The cables are submerged in water. 

Is this correct installation in the first place as per BS7671? 

Appreciate your expert review and feedback. 

Thanks,

Kiran

Proper PV cables are pretty robust - Elland for example say their PV cables are suitable for AD8 - immersion - and I'd guess other reputable brands will be very similar - so the cables themselves should be OK on that score.

The panels do look to be at a very shallow angle though - and with likely shadowing from that concrete(?) upstand shown in the last picture, I wouldn't be at all surprised if the performance was pretty disappointing during the winter months when the sun is low in the sky.

   - Andy.

Its not as bad as it might be !
That roof top weed pond reminds me why truly flat roofs are not supposed to exist according to building control. ;-)  If there are joints sitting under that day after day, I'd be looking to lift them up out of it, though if the cable is unbroken it should be fine - cables underground are often in ducts that are full of water for a large chunk of the year, but you can't see it.
The use of concrete blocks suggests the metal frames are sat on the roof and retained only by gravity and friction - which from a roof integrity perspective is the best,  but means that the low angle is probably more about keeping out of the wind than a tropical location with near overhead sunshine.

Are the flat strips cable guard that is not really attached to the cables ?

I'm assuming the panels are wired in long series strings and the inverter electronics is elsewhere - what are the reported issues?

Mike.

mapj1 said:

If there are joints sitting under that day after day, I'd be looking to lift them up out of it, though if the cable is unbroken it should be fine

Unless I'm much mistaken, looking at that last photo, several of the connectors have been laid on top of buckets!  I'm not sure which section of BS7671 that corresponds to 

not sure which section of BS7671 that corresponds to

133.5 ?

  - Andy.

Hi Andy 

To be honest, all panels stops functioning when the full sun hits and works partially in the winter. 

We are not meeting the targets set by the Council. 

Thanks,

Kiran

Hi Andy 

To be honest, all panels stops functioning when the full sun hits and works partially in the winter. 

We are not meeting the targets set by the Council. 

Thanks,

Kiran

To be honest, all panels stops functioning when the full sun hits and works partially in the winter. 

Definitely something very wrong there then. But hard to tell from this distance what the cause might be - could be insulation faults on the d.c. side or an undersized inverter shutting down when it's overloaded, or any one of a dozen possibilities.

   - Andy.

panels stops functioning when the full sun hits and works partially in the winter. 

I will suggest  that alone is a good enough reason to get the installers back to explain themselves.

The electronics will hopefully produce a log of error codes that can be read to work out why in more detail, but that may need kit you do not have.

At the risk of the first class idiot question, with respect to the panel tilt, which way is south?

Mike

having realised purpose of the buckets, connectors are not designed for that sort of long term immersion. Cable basket on some more of those concrete blocks may be worth considering,  (https://www.tlc-direct.co.uk/Technical/DataSheets/Univolt/Univolt_PVC_Tray_Guide.pdf )but ideally the roof needs to slope towards an unblocked drain!!

KIRAN.M said:

To be honest, all panels stops functioning when the full sun hits and works partially in the winter.

You need someone who can check the error logs on the inverter(s).

My suspicion is an over-voltage trip on the AC side.  If the wiring is under-sized or too long anywhere between the inverter and the local transformer, it can pull the local voltage so high that the inverter shuts down.

You do realise that if you're going to repair the roof, then you're going to have to take all the panels off? Maybe that would be a good time to tidy up the wiring.

That is why I am trying to figure out if the installations and cabling can be left open to elements in this way laying on the flat roof or not? Regardless of how robust the cables are made, my intention is to ask this group if the best practices for cabling are followed or not. 
If any aspect of regulation and safety aspect missed. Not to mention the trip hazard. 

And if that is the case, the over-voltage trip level is something that within limits can be adjusted by the installer, as a software parameter.

The installers for Solar PV have closed their business. Its the developers still in business. Got 4 roof top installations, 2 towards South, 1 East and 1 SE (installations behind lift shafts so in shading basically). Entire roofs and the falls are not designed correctly and that is why we have ponding. 

I agree Simon and that is the best case. Hoping for NHBC to approve that. They are looking into the roof situation. 

And they asked us under what wiring regulations and standards that the panels does not meet the standards? I came across BS 7671 but I do not have the technical knowledge. I was told this is the best place to get some clarity. 

And I was told because of the installation and orientation, a wrong inverter was installed. It is Afore string inverter 10KW but the panels are set for 8 KW

inverter 10KW but the panels are set for 8 KW

That doesn't sound disastrous. The larger the inverter the greater the losses (as an absolute figure) so it's common to size the inverter to the peak rating of the panels, or even a little less(!) (my system has 2.14kWp of panels but only a 2kW inverter)  - but inverters still work perfectly well with reduced output from the panels (as happens 90+%of the time anyway as the sky isn't perfectly clear or the sun isn't at its zenith) - so 8kW of panels feeding a 10kW inverter should still work reasonably well - even if the losses a little higher than the ideal and so the final output is a percent or two lower.

   - Andy.

And they asked us under what wiring regulations and standards that the panels does not meet the standards?

It's going to difficult to prove a defect that way I think. BS 7671 isn't a "thou shall do exactly this to comply" sort of document - rather it talks in general terms of things needing to be adequate for the circumstances. I can see the PV installers (if they were still about) taking the view that if the roof itself had been correctly built, there wouldn't have been any standing water and so both the connectors and the cables would be perfectly adequate, at least from a water ingress point of view. Routing and/or trip hazard is even more vague in BS 7671 - laying cables unclipped on a suitable surface where they're unlikely to be disturbed is certainly recognised as OK - whether your roof classes as "unlikely to be disturbed" is probably a matter of engineering judgement and the exact circumstances. Best practice, adequate, just-about-comply aren't always the same thing. Costs always come into play too - many customers given a choice between a state of the art cable management system or a sign on the access door that says no unauthorized access and a work policy that says all other trades when on the roof need to stay clear of the PV installation, will pick the less costly option. Ideally designers/installers should discuss all these sort of details with the end customer, but I suspect that most of the time they'll just assume that what they did on the previous jobs will likely be acceptable.

   - Andy.