Thanks for all the suggestions, I'm reading them and seeing if any fit.
A difficulty of course that the probes little gnd loops don't seem to act usefully as an antenna. If I shield the loop completely with my hand/body the amplitude of the spikes on that channel actually increase. It almost always seems as if the mains lead is acting as the antenna, and the probes as the gnds, and it's more difficult to position or wave the mains lead around.
Quick quoted answers, but they'll do!
"Hmmm, what do your power providers use for HW heating control?"AFAIK there isn't any encoding on UK mains, in most areas the mains is very reliable. I think there's an economy provider or 2 encoded on to Radio4's 198kHz, together with time/date. For time/date signals we can use MSF 60kHz, R4 198kHz, DCF 77.5kHz, TDF 160.2kHz, GPS.
"That might be electronic lighting ballasts used in sodium lamps, either in your building on on the street."Yes, I was thinking it might be the nearest(~9 meters) outside lamp especially as I've seen it switch to some dim mode, but for some reason this nearest lamp hasn't lit up at all during the last 2 nights, it was on for night1, but no difference to the level of spikes whether it's lit or not.
"thin static looking bars slowly scroll through some of our CATV channels"Yes, I can see it as intermittent thin horizontal lines on a composite-type video source, but luckily nothing at all shows on RGB.
"I can only recommend trying a quality power bar with noise filter"I forgot to try a filtered mains bar, I'll try that tonight.
On night 3 after only 12-15 minutes waiting I managed to capture the start of the spikes as shown. A clean switch on, with the density of the spikes slowly reducing over time, eventually the gaps between the spikes can reach 3 or 4 seconds as in 1st post,1st pic. this is giving me a possible idea of what it might be, which I'll test when I can.