commercial SPD

[padawan]

what is the frequency for testing commercial SPD's, (office building) what is the cost to service them ?

 

[timhoward]

Mindful of the famous old quote "Better to remain silent and be thought a fool than to speak out and remove all doubt"....and ignoring it.....
What leads you to think that
a) you can test an SPD in the first place?
b) they need routinely servicing?
In both domestic and commercial environments I've only ever fitted ones that tell you if they are broken by changing from green to red. Some have contacts for external monitoring systems to report their failure.

So to the extent of my knowledge it boils down to fit, occasionally glance at, and replace if it's been activated.
Am I wrong?

 

[timhoward]

Well to answer myself I just watched an interesting video about testing the varistor in an SPD so was wrong about a)...
But testing this isn't required by BS7671 as far as I know.

 

[pc1966]

I have not seen any specific test cycle, etc.

As above you can get them with remote monitor contacts to warn if the thermal cutout has gone inside (or someone has pulled a module), so alarm generated, etc. For high-value sites or ones with a high incidence of lightning damage in the past that would make a lot of sense.

I would suggest that the visual fail flags are checked at least twice a year when the RCDs get self-tested. You, yes you! Stop laughing at the back!

During an EICR you can check them with an IR tester to some degree, though I believe some newer MFT have a more specific test menu to ease that. Basically they ought to be tens of Meg at 250V and should show tens/hundred kOhm at 500V.

 

[snowhead]

Everything you've ever needed to know about SPD including Testing and frequency of testing which is,,,,,,,,,,,Periodically..

BEAMA Guide to Surge Protection Devices (SPDs) – Selection, Application and Theory

www.beama.org.uk

 

[timhoward]

This was quite an interesting video:

 

[stevethesparks]

More test equipment required :-(

 

[pc1966]

More test equipment required :-(

Many IR testers already show you the voltage they have achieved during the test. For a good circuit giving many Mohm then it is usually a bit above the 250/500/1000V that you have set your meter to, but if you have a soggy circuit then it might well be down in the tens or hundred or so volts.

If you test a varistor (MOV) style of SPD then it starts conducting at a bit above the usual peak AC voltage. So with the max UK being 253V RMS that is 253 * sqrt(2) = 358V peak and you don't want the SPD toasting on normal sort of volts so they might be designed to start conduction (~1mA) in the 400-500V region, and are conducting heavily (in the 10kA region) at the 1-1.5kV sort of point.

Gas discharge tubes (GDT) are a bit different, they show nothing typically until near peak volts (often 1kV-ish) then flash in to conduction down to ~100V as they dissipate the energy, but usually they are kept for Type 1 SPD, or sometimes the N-E protection side of a dual MOV L-N and GDT N-E device.

Deploying GDT on L-N is trickier as you don't want follow-through where the normal AC power keeps the thing burning, so some units have a low-breakdown MOV and GDT in series, etc. But N-E side has little normal AC volts so no follow-through arcing problem.

Anyway, back to testing - if you run a IR test on a SPD and watch the volts you will see it limit in the 400-500V sort of region, that is why they show tens of Meg if tested at 250V but usually well under 1M if tested at 500V. Attempting to test at 1kV would prove it (clamping to ~500V-ish) but if the SPD is missing/broken/not connected OK you need to be sure the rest of the installation is safe to test at 1kV which is not really the case for a nominal 230V system.

The SPD will be fine, so lifting its cable to test it at 1kV will show it working (or not) but usually you don't want to modify a system during testing if at all possible. Here the cheaper SPD that require a MCB for the L supply have a bit of an advantage!

 

[timhoward]

Interesting stuff @pc1966. What sort of voltage causes them to pass the point of no return?
I'd like to assume that 1000V will never push an SPD over the edge and require it's replacement (unless it was faulty anyway). Is that safe to assume?

 

[pc1966]

Interesting stuff @pc1966. What sort of voltage causes them to pass the point of no return?
I'd like to assume that 1000V will never push an SPD over the edge and require it's replacement (unless it was faulty anyway). Is that safe to assume?

It is not really the voltage as such, it is the temperature the SPD reaches.

For pulses that is in effect the energy (analogous to the I2t limit for cables, etc). In real life something like 10kA surge being clamped at 1kV is around 10MW of power, but for the 20us or so equivalent surge duration the energy is of the order of 200J.

But for static testing it is long-term dissipation. I doubt a MFT delivers more than 1-2mA at best, so worst case you are looking at 1W of power and for a device rated in the high tens / hundreds of Joules that is going to take tens of seconds / minutes (at which point heat dissipation starts to make it easier and it is not an adiabatic situation).