Smart meter failed in unusual way, twice.

[baimbrit]

A customer had a smart meter fail in such a way that it was passing on phase through to both of the output lugs meaning that both lugs supplied 120V to ground but there was 0 volts between them. The power utility replaced the meter and all was well but the problem reoccurred a few days later. Any ideas what might cause this? 120/240 volt, single phase residential meter. Aclara brand.

 

[Lucien Nunes]

Was the leg that seemed to be 'crossed over' actually powering anything, or did you just measure the voltage on the smart meter output with a multimeter? If one connection on the input side of the meter fails open-circuit, e.g. at the internal connection between the lug and the current sensor, testing the output with a multimeter will show both output lugs as hot, because a small current will pass from the intact side, through the meter's own internal power supply and measurement electronics across to the other. Obviously there would not be enough current to power any actual equipment downstream.

 

[baimbrit]

Was the leg that seemed to be 'crossed over' actually powering anything, or did you just measure the voltage on the smart meter output with a multimeter? If one connection on the input side of the meter fails open-circuit, e.g. at the internal connection between the lug and the current sensor, testing the output with a multimeter will show both output lugs as hot, because a small current will pass from the intact side, through the meter's own internal power supply and measurement electronics across to the other. Obviously there would not be enough current to power any actual equipment downstream.

Thanks for the reply. Both legs are powering things in the home. I just received a call from the customer (2 hour drive away) who said the power utility just found a jaw of the meter base overheating and melting the meter. I haven't seen a meter fail this way before but one for the memory bank.

 

[Lucien Nunes]

I'm not familiar with the internal layout of typical US smart meters to know what could physically change or move to cause this, but it does sound strange. I wonder whether the heat from the high-resistance lug caused internal insulation to soften and fail, resulting in a short-circuit between the two hot legs. That then blew one of the two open circuit, leaving the other powering both outputs.

Here in the UK, meters have somewhat different construction as they do not plug into a socket, and they only have the one hot and neutral unless 3-phase, so not something we could experience.

 

[pc1966]

I sounds a very strange failure mode to allow L1-L2 to be linked by the fault and yet not blow up due to the supply energy that is possible here. Usually the sort of creepage / clearance needed for mains devices would make that very hard to do, and a simple short would, I suspect, take out both line's upstream fuses instead of allowing one to limp along.

It would be interesting to see inside one of those meters but I presume they are the electric company's property and would be removed when swapped?