Northern Spark said:

Did the spd work?

I think by definition, no.

Suppose that the SPD achieved nothing. It would not trip its 32 A type B MCB. It would not protect the installation. You have not specified whether the smart speaker was on the circuit which tripped. However, the surge would might have led to an over current through the circuits and trip their MCBs.

I have to confess that I do not fully understand why an SPD has OCPD. As I understand it, the SPD can withstand multiple over voltages, but if it shorts (for want of a better word) the MCB will trip. If I have that wrong, please say so.

Smart speaker was on the circuit that tripped.

From reading the Beama guide on SPDs, Could it be you have a type 2 SPD, which can handle common mode surges. But if the surge was a different kind, a differential mode surge, the SPD may have failed to shield your devices ? You may need to get a type 3 SPD, which can deal with differential mode surges, and plug it in where the speaker is as it’s likely classed as sensitive equipment. 

How do you know it did not work, and that without it there may have been far more damage ? Short of having an identical test house with all the same potential victim equipment but no SPD, we can really say almost nothing with any great  confidence. There may have been spikes that the SPD successfully clipped or there may have been none. 

The tripping of the MCBS and the arc damage in the speaker may well relate if during the arc there was in effect a near L-N short.  However that sort of fault would be accompanied by a low voltage and we'd not expect the SPD to help much with that.

Personally I'm in the SPDs are scarcely worth it camp, but it is very hard to prove if they are worth having or not.

Mike.

Some network faults (e.g. broken Ns) can results in sustained overvoltages across single phase loads - say 300V or 400V for many minutes - which aren't the kind of short duration (μs) much higher voltage (kV) surges (from lightning or switching) that SPDs are meant to deal with.

My guess was that the installation was presented by something in the region of 400V that caused the electronics to fail, causing a short that took out the MCBs (MCBs supplying MCBs is never a good idea from a discrimination point of view) - and the SPD didn't get called into action.

   - Andy.

I would concur, given that the neighbours type 3 SPD operated while the OP's customer's device did not have one... I also wouldn't be suprised if the smart speaker's withstand rating was did not coordinate with a type 2 (which was all the way back at the CU so ringing/resonance might have occurred to lessen protection further).

SPDs are really intended to suppress high voltage spikes in the kilovolt range.  If the fault was at lower voltage, perhaps due to a lost neutral, it may not have been enough to trigger the SPD.

There is lot of sensative and electronic equipment in the average home these days compare to 10 - 15 years ago.  I would always advocate a type 2 & 3 combined SPD.  If only a type 2 is in the CU/DB then it can be changed or a type 3 place closer to the appliance.  Eg type 2 in CU and then a type 3 on a 4 way trailing socket for your media wall.  The SPDs are more effective when applied in a cascading manner.

Points to note.

Type 2 SPD can take multiple hits before the flag turn red.  (confusingly enough some are green for good state and some are red)

Type 3 SPD can only take so much abuse without a type 2 upstream

In my opinion the type 2 only clearly what it could and probably did work to within it's parameters.  It is also worth checking the colour flag on the SPD and see what colour for that model means it is still working effectively Green or Red.  In effect SPDs type 2 & 3 are consumable items.

Interesting stuff, thank you all. So it would seem likely it was a 'low' over voltage. I did also wonder if the SPD being in the consumer unit in the house and not at the source of the installation had anything to do with it.

How long before the regs start asking us to fit voltage monitoring relays (like https://uk.rs-online.com/web/p/monitoring-relays/1026131) (and a thumping big contactor) to avoid this problem?

In a sensible world it could probably just be done with a software update on smart meters (as they already have voltage measurement and a contactor built-in) and arguably would be appropriate for the supply industry to bear the responsibility of ensuring that what they deliver is within spec. however...

   - Andy.