[OpenTRV-dev] Aggressive energy saving

[Damon Hart-Davis]

Hi,

Having not yet seen a PIR that I like that works at 2.2V, and given that the LDR 1C fallback in the dark seems to work quite well for me, I'm contemplating a more aggressive strategy for the simple frost/warm or frost/warm/bake mode (no schedule) TRV design.

1) In this case I would take very low light levels as a cue to set back the target temperature 1C regardless.  I don't know whether to combine it with, ie add any setback to that from, (2) below.  I may additionally regard a dark-to-light transition as indication of occupancy if a sustained 'on' level is maintained (eg not just someone flicking a light on briefly to check a room).

2) I would also take operation of the single button as very definite indication of occupancy, and aim to set back the 'warm' target temperature by 1C for every couple of hours since the button was last operated up to a maximum of 8h (say).  (None of these setbacks can drive the target below the safety 'frost setting.)  So if a TRV was forgotten about that would give (say) 30% savings over time (given a total 4C setback), and would test what occupants' actual temperature tolerance threshold is, dynamically, in more normal use.

3) If a PIR is attached, it maybe resets the occupancy timer in the same way as a button press.  Or I may, like the LDR, want to see some sustained continuous indication of activity to eliminate noise, eg from a room being briefly entered but not really 'occupied'.

4) if a schedule is set, every upward change in set-point in the schedule is treated as if an indication of occupancy.

Rgds

Damon