Weather compensation boiler controls?

[Doctor Drivel]

On my system (an ATAG boiler)

[David]

I doubt it would suit an old house with poor insulation.

We have a Vaillant boiler with a Honeywell CM927 in our old house and it works very well and the temperature is kept very steady.

[Roger Chapman]

That depends on what view the control designer has of an ideal system. It should be possible to design a system that never overshoots but the cost would be in taking longer in zeroing in on the set temperature. The more money you throw at the design the more accurate the control and the quicker the response but, other things being equal, no overshoot equals slower.

Weather compensation is probably the cheap way of introducing extra heating when it is cold outside but but in theory at least it should be perfectly possible to get first class temperature control by looking no further than the rate of change of the internal temperature.

[Onetap]

Yeah but, for one building, the rate of heat loss ( through the fabric and by ventilation) depends mostly on the outside air temperature. A change in To will take some time to register on the internal room temperature sensor, by causing a drop in the room temperature. It will then take some time for the heating to respond and bring the internal temperature back to the set point. Weather Compensation anticipates this and turns the heating up or down to match the rate of heat loss. It's more critical with high mass systems, especially UFH in screed, which can take many hours to respond to an increase in the flow temperature.

My previous system had cascaded control loops, looking at both To and Ti. It worked very well, eventually ;-). However I cocked up the I and D control parameters so the control valve was continually moving. It wore the gland out & started leaking quite soon.

[Owain]

and 1.5 insulate hot water cylinder if not already foamed.

Owain

[Doctor Drivel]

Tripe. You know nothing of self adaptive controls.

< I will will to snip the rest as it is all wrong >

[Roger Chapman]

It wouldn't matter if I didn't. If you want to have enough damping to prevent the possibility of any overshoot it will always take longer to actually reach the set temperature than the same system with less damping and hence the possibility at least of some overshoot.

Oh well, par for the course. Dribble dismissing out of hand anything he fails to understand. Stand by for the plant pot phase of the Dribble response to reason.

[Doctor Drivel]

Tripe. You know nothing of self adaptive controls.

[Roger Chapman]

That reply took you a long time to formulate given the lack of reason behind it.

It may be over 40 years since I studied control engineering but I don't expect the basics or the maths to have changed in the interim. The bit you have edited out without acknowledgement - that a system damped heavily enough to prevent overshoot will take longer to reach the set temperature than the same system with less damping - is a basic truth however clever the system is in predicting future behaviour on the basis of previous observations.

[Roger Mills]

That possibly needs some clarification - as illustrated by this Wiki article on damping:

Critically Damped system will reach the Steady State setpoint in the shortest time without overshoot. A less damped system will *pass* *through* the setpoint sooner, but will continue to oscillate for longer.

So a lot depends on how you define "reach the set temperature"! There is a sense in which you and Drivel are both right.

[Roger Chapman]

I had 'first reach' in a draft but edited it out. I didn't say anything about about remaining at set point as in order to overshoot the temperature first has to reach set temperature. If I had known about what was in the Wiki I might just have left it in but I doubt it.

[Doctor Drivel]

> A Critically Damped system will reach the Steady State setpoint in the

You have to take into account that the man is mad.