RCD a but slow. Time for a change?

[Electron]

I tested a customers BS4293 MEM 100A, 30mA RCD, which is about 25 years old, after it tripped overnight. The trip time at x 1 was 207.2, 78 and 54.2 ms at 0 deg. and 68.6 ms at 180 deg. I know that the limit is 200 ms, so one reading was slightly over, which is why I repeated the test. At x 5 it tripped at 50 ms at 0 deg. and 41.5 ms at 180 deg. So, again both over the limit of 40 ms. No trip at x 1/2.
Do these results mean I have to change the RCD? Or is there any tollerance? If so, can only get used identical replacement from the internet. Or should I change it to a RCD to BS EN 61008-1, like a Wylex 100A 2 Pole 30mA RCD?

 

[Deleted member 26818]

They might, but it's still not acceptable as they have already failed the test.

It's no use telling people to exercise the device with the test button several times before each time they have an Earth fault.

The device has failed.

Although the device did operate recently, which is why it’s been tested.

 

[Risteard]

Although the device did operate recently, which is why it’s been tested.

It might have operated, but we don't know whether it operated within specifications.

 

[Toneyz]

This is why it should be tested periodically.

 

[stevels4]

Does this help?

No. I always say to customers "Can I assume that you test the RCD's every three months?" They look at me as if they have broken the law!
Does anyone ever check the RCD? I only ever do it when someone pays me!

Mine gets tested every time my other half decides to toast a tea cake and a current gets stuck between the casing and heating element!