DC isolators and RC62

I think this possibly a Contractual and Designer issue.

If it is an issue at all. I'd be minded to do the same, keep all strings on their own isolator, it's clearly a medium sized install (rather than a 4kW domestic), so being able to isolate individual strings would be beneficial at times in the future. You could also have the problem of too much current for one/two/three  isolators, so string based isolation would be my choice. 

lyledunn said:

do we really need to have an isolator on each string

I guess some means of isolation will be needed (to comply with the EaWR for instance as well as BS 7671) - if the inverter has a d.c. isolation built in (to break the current flow) then unplugging connectors to remove voltage from the inverter (e.g. if the inverter itself needed replacement) would be an option. Otherwise I don't see you can do any less than one isolator per MPPT.

  - Andy.

No experience with inverters this big. But the isolators do provide a way to reboot the inverter if it decides to stop working for any reason.

Turn off all the AC and DC isolators. Wait a few minutes. Turn all the isolators back on again. Without the DC isolators, you'd have to leave it overnight.

Does the inverter have a built in DC Isolator and pluggable connectors out to the strings. If yes, then I'd question the need to fit additional isolators.

Well the unit has two “DC Switches” as labelled in the product bumf. Whether they are “disconnectors” or “isolating switches” in 7671 speak, I am not sure. It might be advantageous for several reasons to have the DC isolators (in coal face speak), but that needs to be weighed against the desire to mitigate against fire as per insurance provider recommendations. Not something that can be dismissed lightly!


By the way, it never ceases to amaze me how good some YouTube presenters are. Whilst top of my list is Joe Robinson, there is a fella called Gordon who also has a very special talent in presenting!

I can't help wondering why isolators are such a fire risk (if they really are - the statistics given percentages of which component that started a fire presumably should be weighted against the probability of a fire starting at all). Moving contacts will always be a weak point so if there is a fire it is perhaps natural that it's more likely to occur at a weak point).

But on the assumption that isolators are a problem - then why? Is it the moving contacts or is it a re-run of the consumer unit problem where we've gradually moved to terminals that are less than ideal (especially in situations where there's very significant daily thermal cycling), or is it just that it's just about the only point on the d.c. side where screw terminals are used (most other connectors are crimped) so there's a greater risk of human error during installation (and no loop tests that might catch poor terminations like we have on the a.c. side)?

  -  Andy.

Thanks for the feedback.  I'll let Joe know! -  I have made a video on this topic https://youtu.be/2BBCclJuM4I?si=TbZdh106nul-7gvz.  In practice rarely see DC isolators on commercial solar projects. 

It's probably more an installer issue than a fundamental design of the isolator. A DC isolator can be set up to handle different ranges of voltage and current, but inserting or removing shorting links to increase creepage and clearance. Secondly, they often end up being installed outside we often see poor practice in maintaining IP ratings.

The COP is clear that the DC switch disconnector can be built into the inverter. Therefore, there is no regulatory imperative to have individual switch disconnectors for individual string connections. You can imagine the difficulty a contractor would have if he did not have a robust design strategy that considered and controlled the risk from the devices, if a damaging fire was traced to their use. 
I acknowledge good reasons for their use and I accept that there is little substantive evidence to properly evaluate risk, but I would caution contractors and clients to avoid situations as indicated in the photo unless the insurer is consulted.

I suspect that in part it's also that we've lost the practical historical knowledge of DC isolation and it's problems (DC traction is essentially long gone; Blackpool as an exception?), plus the solar panel source is a new set of phenomena, and we now source parts from 'everywhere' with varying cost, quality and local knowledge standards.