@derek-m single heat pump, just run on weather comp, no stats, no setback.

The Carel platform that my heat pump is built on allows full tuning of the PID settings, along with full control of the compressor ramp speeds. It’s close to maintaining setpoint, but I’ve still got some work to do. I’d like to deal with the initial overshoot next, I think I’ve got the oscillation damped quite well.

When it’s around freezing outside, it runs happily for a long period of time, it’s dealing with the overshoot when it’s cycling that’s the next area of tuning that needs my attention.

I suspect that it will not be possible to accurately control the LWT during milder weather conditions, simply due to the fact that the heat pump will be starting to operate outside its control envelope. i.e. it cannot turn-down any further.

As far as controlling IAT is concerned this should not be a major issue, though it could take some time to optimise for maximum efficiency.

Out of interest what are your PID settings?

It is my suspicion too, that in milder weather, we are simply oversized. I didn't take into consideration the possibility that the ACH value used in the heat loss spreadsheet (and by heating engineer) could be considerably higher than we actually achieve.

The fact that it was tracking the requested setpoint to within 0.1 degree during the last 2 weeks of freezing weather leads me to think my PID settings aren't a million miles off.

At the moment they are: P = 5 deg. C, I = 50 secs, D = 0 secs.

The controller also allows me to open up the overshoot window, which gives headroom for compressor rampdown - currently set to 5 degrees. I can't remember the rampdown settings though, it's specified in Hz/sec inside the Power+ inverter settings.

To comment on the original question in the thread title - I was away for a lot of the 2 week freeze, but the residents reported stable internal temperatures (21-22) 24/7 and there were no issues with defrosting, even with our limited system volume.

Whole house electric consumption was running around 50kWh/day - all electric house now.

The Carel platform that my heat pump is built on allows full tuning of the PID settings, along with full control of the compressor ramp speeds. It’s close to maintaining setpoint, but I’ve still got some work to do. I’d like to deal with the initial overshoot next, I think I’ve got the oscillation damped quite well.

When it’s around freezing outside, it runs happily for a long period of time, it’s dealing with the overshoot when it’s cycling that’s the next area of tuning that needs my attention.

I suspect that it will not be possible to accurately control the LWT during milder weather conditions, simply due to the fact that the heat pump will be starting to operate outside its control envelope. i.e. it cannot turn-down any further.

As far as controlling IAT is concerned this should not be a major issue, though it could take some time to optimise for maximum efficiency.

Out of interest what are your PID settings?

Although there's no such thing as 'too hot' in this house - Wife and Sister-in-law (who live there) are happy as long as it's >22

Interesting, is this to get the house temp right or just the flow temp.  Intuitively would have thought that the latter should be quite straightforward, but maybe compressor/fan responses complicate things.

The Carel platform that my heat pump is built on allows full tuning of the PID settings, along with full control of the compressor ramp speeds. It’s close to maintaining setpoint, but I’ve still got some work to do. I’d like to deal with the initial overshoot next, I think I’ve got the oscillation damped quite well.

When it’s around freezing outside, it runs happily for a long period of time, it’s dealing with the overshoot when it’s cycling that’s the next area of tuning that needs my attention.

What kind of control system are you using, is it a single loop or a cascade (master - slave)?

On the subject of control strategy, some units can be set to control on LWT or RWT…. 

Mine used to be switchable between the two but a recent software update removed that option. You could always strap your buffer temp sensor to the return line I guess.

Interesting.

It is of course in principle equally valid (although presumably achieved in practice by turning the 'wick' up  or down to change LWT, or by modulating flow rate - which would not achieve the required end result).  However it has an inherent delay in its response, which is undesirable in control loops, so I'm not sure why manufacturers would choose this option!

Posted by: @derek-m

↑

The heat pump merely tries to supply sufficient energy to meet the demand.

 

@jamespa I think I am right in saying is of this view, and I do get this point: the heat pump responds to the feedback contained in the RWT. Indeed i was looking into this earlier today, in the context of cycling, trying to discern what triggers the cycle:

On first glance it makes sense that it is the LWT going over a certain point above the set LWT, but in fact it may be the RWT reaching the set LWT. Comments welcome.

Yes.  My view is simple, to think about how much energy the heat pump supplies, first think about how much energy is demanded by the emitters.  

In more detail:

The heat pump knows: RWT, OAT, LWT and (maybe) IAT.  It also has a target LWT determined by the WC curve (or a fixed number).

The starting point for the control system is clearly to achieve RWT = target RWT, which happens by supplying just enough energy to make up the difference between LWT and RWT, ie the energy lost by the emitters at the chosen LWT.  That's a very simple feedback loop and, unless there is an auto adapt mode which has been turned on, the only one necessary to achieve what the heat pump promises to do. 

If there is an auto adapt mode that presumably operates by adjusting the target LWT based on measurements (and history) of measured IAT/OAT.  However that still results in a target LWT which the simple feedback look can then operate on with the same result.  

The upshot of this is that the energy supplied by the heat pump is determined by the energy lost from the emitters.  This is in turn determined by IAT, OAT and LWT.  

Some heat pumps also change flow rate but that has only a very small influence on the energy lost by the emitters at any given LWT.  I'm not actually sure whether there is any advantage in modulating flow rate in this way, leaving it fixed actually straightens out the WC curve a bit (for radiators), and since most heat pumps feature linear wc curves that would seem to be an advantage.  It may help with the transfer in the heat pumps heat exchanger though.

Various things will trigger the heat pump to shut down (cycle) including LWT > target LWT plus permitted hysteresis, (very likely) RWT >  target LWT minus permitted hysteresis, and various fault conditions which leads the pump to conclude that its control loop cannot achieve what is being requested of it or its in danger of being damaged.

Mine used to be switchable between the two but a recent software update removed that option. You could always strap your buffer temp sensor to the return line I guess.