Funnily enough I had something a bit similar today (though lower voltage).
During a CU change there were 3 sets of T+E in a B32 device. I wanted to check it was two for an RFC and one for a Spur and of course check end-to-end readings anyway. I got lucky first time and first two brown wires were showing continuity. As did the blue's and the CPC. I then moved a clip to the 3rd cable to double check and my Metrel tester made the dull sound when it refuses to test, showing 29v L-CPC.
As this was in a cellar, I was a bit naughty and had the board and lighting circuit live while testing other circuits.
When I traced the cable concerned I found it ran very parallel to the live lighting circuit for a long distance, and ended up at a socket with a very long extension lead plugged in. Unplugging the extension lead was enough to make the tester happy and prove what I'd worked out anyway by then, the cable wasn't connected to anything else.