AMD 3 and "non-combustible" CUs...

[JohnD]

These MEM boards have a hinged and latched steel door, giving access to the main switch and the fronts of the MCBs sufficient to see them and operate the switches, and to see the labels or circuit identification list which is usually on the inside of the door.

Behind the door is a screwed-on steel panel which can be removed for maintenance, in the photos it has been removed, and gives access to the connections on both ends of the MCBs, and the earth and neutral bars. There is a separate cover over the supply connections of the main switch.

I think the pre-fire photos will have been taken during maintenance, when the screwed-on cover had been removed. It would be dangerous, and highly irregular, for the db to be in service without the cover.

My guess is that the overheating must have started, or worsened, fairly recently or there would surely be scorching or smoke marks in the casing that would be noticed during the periodic inspection.

View media item 1574
Something like this, though older ones may have a smaller central door with a metal latch.

 

[stillp]

I wonder when their last inspection was?

 

[JohnD]

The fire report mentions an inspection in 2010, so perhaps that was the last one.

 

[stillp]

"National Trust Director General Dame Helen Ghosh said it was "concerning" a faulty fuse board at Clandon Park was signed off safe ahead of the fire."
Hard to imagine how an EICR would have detected that.

 

[JohnW2]

Hard to imagine how an EICR would have detected that.

Unless it was already getting hot. Maybe EICRs should include thermal imagining? In some ways it could be said to make more sense to undertake a test which can reveal an already-present (but not otherwise easily detectable) serious problem - rather than tests which identify the potential for a problem which, in reality, may never arise!

Kind Regards, John

 

[RF Lighting]

I thermal image all DBs and supply gear on EICRs

 

[JohnW2]

I thermal image all DBs and supply gear on EICRs

Yes, you've mentioned that, and it seems to make very good sense - as I said, it's arguably far more important to detect a serious problem that already exists than to undertake tests which only "may" have impact on safety issues at some point in the future. I was therefore pondering whether thermal imaging should perhaps be required as part of an EICR - do you think that will ever come?

Kind Regards, John

 

[bernardgreen]

There could be a case for fitting heat sensors inside distribution board cases to provide early warning of overheating and flagging the need for a full inspection.

 

[JohnW2]

There could be a case for fitting heat sensors inside distribution board cases to provide early warning of overheating and flagging the need for a full inspection.

Given the need to cater for a wide range of ambient temperatures etc., I'm not sure that such sensors would be able to detect a problem before it was getting quite close to becoming a fire - but I did recently suggest in another thread that a requirement for smoke/heat detectors within a CU/DB might have been more sensible than the requirement for it to be 'non-combustible'!

Kind Regards, John

 

[RF Lighting]

I thermal image all DBs and supply gear on EICRs

Yes, you've mentioned that, and it seems to make very good sense - as I said, it's arguably far more important to detect a serious problem that already exists than to undertake tests which only "may" have impact on safety issues at some point in the future. I was therefore pondering whether thermal imaging should perhaps be required as part of an EICR - do you think that will ever come?

Kind Regards, John

It's not inconceivable, especially as there are thermal imagers for you phone costing as little as £150 these days.

http://www.tester.co.uk/flir-one-personal-thermal-imager?gclid=CPTY06mxuMoCFSEIwwodG5oBfQ

 

[ban-all-sheds]

Given the need to cater for a wide range of ambient temperatures etc., I'm not sure that such sensors would be able to detect a problem before it was getting quite close to becoming a fire

My understanding is that nothing should be at more than 70°C, no matter what the cause of heating. That's a long way from close to being a fire, unless there's a small heat source which is rapidly rising in temperature - then a detector in the enclosure might not pick it up.

 

[ban-all-sheds]

The report says """He went to the fuse board to investigate the loss of power and on opening the cupboard discovered that there is a fire inside. """.

Assuming (I know..) that they meant "opening the door of the DB", then clearly it can't have been too hot to touch. My immediate reaction was to wonder WTH there weren't any CO2 or water mist extinguishers he could have grabbed and tried to do something about the fire.

 

[JohnW2]

My understanding is that nothing should be at more than 70°C, no matter what the cause of heating. That's a long way from close to being a fire, unless there's a small heat source which is rapidly rising in temperature - then a detector in the enclosure might not pick it up.

I agree that there should not normally be anything hotter than 70°C, but my thought was that if, say, just one small terminal was at a much higher temperature than that, it could take a long time before the 'general temperature' within the enclosure rose to above, say, 70°C. Of course, it would depend upon how many sensors and where they were located - and, equally obviously, any sort of sensing would be better than none.

Kind Regards, John

 

[raykf]

Looks as though these guys think early thermal abnormality detection is viable:
http://thermarestor.co.uk/pdf/brochure.pdf

 

[JohnW2]

Looks as though these guys think early thermal abnormality detection is viable:
http://thermarestor.co.uk/pdf/brochure.pdf

I guess they would! However, that's a bit different from what I was talking about (just monitoring the 'general temperature' within the CU enclosure), since it actually senses the temperature of the outgoing terminals of each MCB, and hence is far more likely to detect localised temperature rises. However, it does not seem to also sense heating occurring at the incoming (busbar) terminals, and it wouldn't be directly usable with RCBOs.

However, the concept is, IMO, a good one. Prevention of fires is obviously far better than trying to contain fires that have already started.

Kind Regards, John