Posted by: @cathoderay

↑

The heat pump logic says it is warm outside, so the house needs less energy (only enough to balance the heat loss at that OAT), but in fact the house needs more.

That's exactly what my analysis above would predict.  After the rise in oat  the bit of the house you live in continues, for a while, to lose energy to the middle of the walls at a higher rate than would be the case in the steady state at the new oat.  This continues until a new thermal gradient is established which falls continuously from inside to outside, rather than having a minimum in-between.   The heat pump supplies energy corresponding to the steady state at the higher OAT and thus there is a deficit.  This is a direct consequence of the fabric thermal mass (heat capacity in scientific terms).  I would expect the converse to happen if there is a sudden fall in oat.  

My boiler, which had a control loop based on iat, exhibited different, but no less annoying, deviations from perfect control.  The problem for the designers of heating control systems is, i imagine, that they don't know the characteristics of the house (which will vary widely) and can't rely on installers or homeowners to dial them in.  That's where controllers which learn about the house over a period of time have the edge!

A similar pattern happened again last night and into this morning, low overnight OAT, followed by rapid rise in OAT fooling the heat pump into thinking the house was warm, IAT must be fine when in fact it was anything but, it was actually falling:

I have now written some extra code in my auto-adapt script to detect this and if detected, then add a boost to the WCC end points. I will report back once it has had a chance to run, and can be seen to be working. 

I look forward to seeing the results of your script.

I fiddle with our WC curve increasing/decreasing between 0.55 and 0.75 according to the extent of wind chill. Strong winds and driving rain doesn’t get adequately compensated by WC alone, even on Vaillant’s active mode.

Posted by: @judith

↑

I look forward to seeing the results of your script.

Thanks for your interest. It hasn't really been tested yet. The recent modification to the code was to deal with a rather specific circumstance, a rapid rise in OAT after a cold spell dialling down output prematurely, and that hasn't happened since adding the modification (I suppose a bit like it never rains when you take your umbrella!). What it doesn't (yet) do is deal with wind chill, which we have had rather a lot of recently. The script does do its thing, but it does it assuming no wind chill, meaning it doesn't (yet) do enough auto-adaption.