Possible danger with solar power

[bernardgreen]

No matter how many fail safe ( define "safe" ) devices are fitted to the panels the basic problem ( or hazard ) is that a solar cell when exposed to light is a source of electrical power. If the cell is disconnected and has no load there is no power produced but the off load voltage is still present.

For a single cell this voltage will be low enough that it is not going to present a risk of electric shock but if shorted the current may be high enough to create an ignition source.

For an array of series connected cells in a panel the off load voltage in most roof top arrays will exceed the "safe" voltage ( safe = no risk of harmful shock )

The SolarEdge power optimizer is a DC/DC converter which is connected by installers to each PV module or embedded by module manufacturers, replacing the traditional solar junction box. The SolarEdge power optimizers increase energy output from PV systems by constantly tracking the maximum power point (MPPT) of each module individually. Furthermore, the power optimizers monitor the performance of each module and communicate performance data to the SolarEdge monitoring portal for enhanced, cost-effective module-level maintenance. Each power optimizer is equipped with the unique SafeDC™ feature which automatically shuts down modules' DC voltage whenever the inverter or grid power is shut down.

""automatically shuts down modules' DC voltage"" but the panel is still producing voltage.

How does the module perform when the roof beneath it is on fire. Does it still "fail safe" when subjected to high temperatures and the voltage from the solar panel is still present.

 

[SimonH2]

The point is, for safe isolation a solid state switch with "it automatically shuts down" logic doesn't cut it. That's enough for "prevent backfeeding the local grid" protection since the local grid just isn't going to be small enough a load for the "shut down" system to energise it. For isolation for working safely on the connected electrical system, I believe mechanical contacts with a specific opening distance are required - though I suppose you could follow the lead of the electrical supply industry and robustly earth the supposedly switched off circuit (ie physically short the inverter output (L&N) to earth), just don't forget to remove the link before switching the power back on

But back to the fire issue - I can't remember the link, or the thread it was in (it was on a forum here). IIRC they concluded that shorting the DC cables wasn't practical or safe, and blocking the sunlight from the panels wasn't either.

 

[bernardgreen]

One concern is that when attending a fire at night the hazard of solar panels may not be apparent and having extinguished the fire, damped everything down and shut off all services the fire brigade may rightly assume that there is no source of re-ignition. At sun rise solar panels start producing power and if the control gear / invertors etc have been fire or water damaged then anything could happen. If the solar panels are out of sight on an outbuilding then the night time crew may not be aware of them and the hazard they will creat as sun rise. Good record keeping as to who has solar panels fitted would reduce this risk but does everyone who fits panels notify the local fire service. ?

 

[a546345]

My point was that rather than having standard series connected panels putting out an uncontrolled 350V to 500V DC on the end of a cable, with the optimisers you have a short length of cable on the back of the panel with a potential 40V into the device, and although the output of this device is still series connected to the others, you only get a 1V output until the optimisers are "switched on" by the inverter. This is surely safer than without that control?

Appreciate the point that this doesn't give mechanical isolation.

 

[JohnW2]

For an array of series connected cells in a panel the off load voltage in most roof top arrays will exceed the "safe" voltage ( safe = no risk of harmful shock )

That may well be the case, but the point I was making was that it doesn't have to be like that - individual panels could theoretically have as many "interrupters" of the series connections built in as was necessary.

Kind Regards, John

 

[SimonH2]

... with the optimisers you have a short length of cable on the back of the panel with a potential 40V into the device ... This is surely safer than without that control?

It would seem so, but again, what happens when they've been involved in a fire ?

Also, I believe there is some evidence that they increase radiated interference. In effect you've a switched mode power converter on the back of the panel feeding into a cable. There is definite evidence of some "non-optimised" installations creating interference and that's with just one switched mode converter. Add in a load of extra converters and you increase the scope for interference.
Yes, there should be filtering, but c.f. the recent discussions regarding small "imported" power supplies - with price being a significant factor it's not hard to imagine there being some "not very good" units around. Also, given the way power line comms stuff is tested in a way so as to avoid it putting any signal down the mains cable, these units could be EMC tested with a very short cable to a resistive load - very different to long leads feeding an inverter.

 

[JohnW2]

... with the optimisers you have a short length of cable on the back of the panel with a potential 40V into the device ... This is surely safer than without that control?

It would seem so, but again, what happens when they've been involved in a fire ?

There can obviously be no guarantee of what will happen when any of the things we're discussing find themselves in the middle of an inferno - but, as stated above, it's surely better to have such measures than not to have them, isn't it (i.e. they will probably work as intended in a fair proportion of cases)?

Kind Regards, John

 

[JohnW2]

But back to the fire issue - I can't remember the link, or the thread it was in (it was on a forum here). IIRC they concluded that shorting the DC cables wasn't practical or safe, and blocking the sunlight from the panels wasn't either.

Fortunately, one doesn't need to 'remember', since the forum has a search facility - how about this thread?

Kind Regards, John

 

[SimonH2]

That's the one I was thinking of.

 

[bernardgreen]

Not solar but alternative energy source. Micro hydro electric.

Generator belt driven from an antique overshot wheel in what had once been a hammer mill. The sluice feeding the wheel was closed to stop the wheel while work was done to the overhead open wire supply. Either heavy rain or a blocked spill way meant the water eventually over flowed the sluice and the wheel began to turn and the generator started generating.

Fortunately the worker heard the wheel creaking up to speed and got clear of the open wires in time.

Micro power or not the voltage 230 v was enough to be fatal

 

[JohnW2]

...The sluice feeding the wheel was closed to stop the wheel while work was done to the overhead open wire supply. Either heavy rain or a blocked spill way meant the water eventually over flowed the sluice and the wheel began to turn and the generator started generating.

No (electrical) isolator(s) in sight (or used)?

Kind Regards, John

 

[bernardgreen]

Possibly the owner was a bit wet behind the ears.


Seriously though shutting of the water to the wheel seems to be considered as adequate for mechanical safety in most water mills. I cannot recall ever seeing an axle pin that could be inserted to prevent the shaft turning. Probably because if the wheel was loaded all buckets full and the shaft pinned then something would sheer and be more hazardous.

 

[JohnW2]

Seriously though shutting of the water to the wheel seems to be considered as adequate for mechanical safety in most water mills. I cannot recall ever seeing an axle pin that could be inserted to prevent the shaft turning. Probably because if the wheel was loaded all buckets full and the shaft pinned then something would sheer and be more hazardous.

Yes, possibly adequate for 'mechanical safety' - but I would still personally want to see 'electrical safety' (i.e. isolators) separately addressed, wouldn't you? Would you be happy to work on the cables originating from any other sort of (maybe distant) generator on the basis that the fuel supply to the generator had been switched off, without any electrical isolation?

Kind Regards, John

 

[endecotp]

how about this thread?

Thanks. In case anyone else is interested, this link

http://site.ul.com/global/documents...rials/fireservice/PV-FF_SafetyFinalReport.pdf

Seems to have changed slightly to this:

http://www.ul.com/global/documents/...rials/fireservice/PV-FF_SafetyFinalReport.pdf

 

[Blackhand]

Have been doing some minor changes to my electrics and switched off the double pole isolator on the CU. But then a thought, solar is connected to the live bus bar via a 20amp MCB and thus makes the bus bar live again but not connected to the incoming supply. Another reason for checking for dead methinks!

I hope that mcb isnt on a shared rcd bank.