Still learning!

Well it seems that when the system was installed our electrician mis-connected the external (to the GSHP) house heating circuit pump (P1 in IVT speak, this is the usual circulating pump that all systems have) across the terminals for the ground loop circulation pump (P3), with the effect that it operated (& hence the radiator circuit water circuated) only when the GSHP was running instead of continuously. This explains all the oddities of the temperature movements before and after the accumulator tank, which have now gone away as I have re-wired it! (so much for more complex theories). Also the lounge (fan convector) now works much better. The transition is clear towards the left of the lower image above, after which temperatures hardly vary through the accumulator tank and the lounge temperature (yellow) rises to normal. One does wonder whether this might have been affecting the system last winter when we had high return temperature alarms, though of course the effect will be larger now when the GSHP is only on occasionally.

FTR this was an accidental observation after I deliberately switched P3 to continuous to see more clearly what the ground loop temperatures are doing (the average outgoing had come above the average return - but this is an averaging artifact). They do in fact behave normally top above - purple outgoing, blue return - note the ~10 minute delay for the fluid to circulate round the loop and the return temperature to drop).