Evohome TPI cycles

[asimion]

Hi guys,

I'm new to this forum so please excuse (what you may find) potential trivial questions.

I run a system composed of a Viesmann Vitopend 222 combi-boiler, controlled by an Evohome system (simple setup with 1 BDR91 controlling the boiler and only 1 HR92-as, for the moment, I'm trying to understand how this system works).

What I observed by long monitoring is that the boiler is continuously cycled every 10min (in the proportional band). Except some initial overshoot when the system was forced to keep the boiler off for apr. 1h there is no single observation of any cycle skipping. It regularly fires every 10min. This is because the HR92 will almost never close completely (in fact the minimum valve position I have observed is apr. 20%) so it continuously ask for heat. My first question is if this behavior is normal or there is something wrong with my particular setup.

If this behavior is normal, do you think it can have an impact on the life time of the boiler itself? I don't think it is easy for the boiler being fired so often. On top, in my particular case, there is quite a long overrun time of the circulating pump(~10min) . Overall result is that the pump will run continuously.


I thank you very much for the support and I wish you to have a happy and successful New Year!

[RO09]

Hi

It sounds to me like you’re running in single-zone mode, so presumably, on the controller you just have one heating zone that fills most of the screen? I ran Evohome in this way when I first had it back in 2014. My experience was that in single zone mode, the controller applies the proportional band below the set point as usual, but also ABOVE the set point. Effectively I saw a 1.5 - 2 degree overshoot all the time. For example if the set point was 20, it wouldn’t stop cycling the boiler (i.e. it would not skip a 10 min period without firing) until the temperature was around 21.5-22 degrees.

I had the controller replaced by the supplier as it was suspected faulty, but this made no difference. I also tried two slightly different set ups - single zone with no HR92, and single zone with one HR92. Same results. So my conclusion was that this behaviour was by design. I guess it might work quite well in a larger house (..with greater heat loss and distant rooms than my modern two-bed flat I was in at the time).

I suspect that if you acquire another HR92 and then factory reset the controller and set it up with two zones, this behaviour will go away, like it did for me.
I’ve only seen one other person mention this in the past, but I did easily reproduce the issue a couple of years ago, so I’m fairly sure this is what is happening. My controller is the v2 - the first colour one, with a separate remote gateway.

[DBMandrake]

I run a system composed of a Viesmann Vitopend 222 combi-boiler, controlled by an Evohome system (simple setup with 1 BDR91 controlling the boiler and only 1 HR92-as, for the moment, I'm trying to understand how this system works).

What I observed by long monitoring is that the boiler is continuously cycled every 10min (in the proportional band). Except some initial overshoot when the system was forced to keep the boiler off for apr. 1h there is no single observation of any cycle skipping. It regularly fires every 10min. This is because the HR92 will almost never close completely (in fact the minimum valve position I have observed is apr. 20%) so it continuously ask for heat. My first question is if this behavior is normal or there is something wrong with my particular setup.

Are you sure you mean the HR92 valve position (by going to option 10 in the HR92 itself) or do you mean the heat demand for the zone reported in the system devices page in the installer menu on the controller ? The two figures are related but they are not equal.

I ask because any HR92 valve position below about 30% will not generate a heat demand and thus not cause the boiler to come on at all. If you're still getting a heat demand there might be a configuration problem on your system. Check the heat demand for the zones and boiler relay on the system devices page in the controller to get a better idea of what's happening.

On the other hand, at this time of year I would expect there to be a small but continous demand all the time to maintain the room temperature so it's unlikely the boiler would go off completely while the zone is scheduled to be on at a reasonable temperature. As long as there is a minimum of 10% heat demand (based on 6 cycles per hour and 1 minute mininum on time in the configuration) then the boiler will fire for at least one minute in 10 minutes. This is normal.

If this behavior is normal, do you think it can have an impact on the life time of the boiler itself? I don't think it is easy for the boiler being fired so often. On top, in my particular case, there is quite a long overrun time of the circulating pump(~10min) . Overall result is that the pump will run continuously.

Shouldn't have any meaningful impact on the life of the boiler. Keep in mind that (a) most single zone thermostats these days also use TPI control, so Evohome is hardly unique in this regard and (b) even when the boiler relay is on continously if the heat demand from the radiators is below the modulation range of the boiler, the boiler will be forced to cycle the burner on and off anyway - and it is built to be able to withstand this.

I've been using Evohome with TPI on a 25 year old boiler for 4 years which was built long before TPI was a thing and no problems so far related to the on/off switching.

Hi

It sounds to me like you’re running in single-zone mode, so presumably, on the controller you just have one heating zone that fills most of the screen? I ran Evohome in this way when I first had it back in 2014. My experience was that in single zone mode, the controller applies the proportional band below the set point as usual, but also ABOVE the set point. Effectively I saw a 1.5 - 2 degree overshoot all the time. For example if the set point was 20, it wouldn’t stop cycling the boiler (i.e. it would not skip a 10 min period without firing) until the temperature was around 21.5-22 degrees.

I had the controller replaced by the supplier as it was suspected faulty, but this made no difference. I also tried two slightly different set ups - single zone with no HR92, and single zone with one HR92. Same results. So my conclusion was that this behaviour was by design. I guess it might work quite well in a larger house (..with greater heat loss and distant rooms than my modern two-bed flat I was in at the time).

I suspect that if you acquire another HR92 and then factory reset the controller and set it up with two zones, this behaviour will go away, like it did for me.
I’ve only seen one other person mention this in the past, but I did easily reproduce the issue a couple of years ago, so I’m fairly sure this is what is happening. My controller is the v2 - the first colour one, with a separate remote gateway.

This scenario has been discussed very recently:

https://www.automatedhome.co.uk/vbul...ing-other-room!!

In short, it is normal for the boiler to continue firing intermitently slightly above the set point, within the +/- 1.5 degree proportional band, and this is the case regardless of whether you have one zone or many. If it just stopped firing completely as soon as you went above the set point the temperature would then plummet well below the set point before it recovered. This is not how proportional control works.

And no, the proportional band is not responsible for temperature overshoots, this is a common misunderstanding. Have a look in that thread for some ideas on how to deal with overshoots.

The common reasons are (a) oversized radiators (b) flow temperature too high for the conditions (radiator size, heat loss of the house, outside temperature etc) (c) valve bodies that are not calibrating correctly to open at 30% valve position.

Keep in mind that if there is a large change in the operating conditions, such as a sudden shift in outside temperature from one day to another, (which we have been getting here lately) it will take a few hours of runtime near the set point for the HR92 to gradually adapt to this. (This is integration occuring in the PID controller, which takes time)

In the meantime you might see a persistent overshoot or undershoot of approx 0.5C depending on which direction the conditions changed, but it will adapt if you give it time.

I can assure you that I do NOT see 1.5-2 degree overshoots in my system, apart from a couple of pathalogical zones where it is fully expected to happen due to where the temperature is being sensed.

Typically I never see any more than 0.5C overshoot in all my zones with remote temperature sensors where true room temperature is being sensed. Overshoots measured by the HR92 as a sensor are not always "real" due to the proximity to the heat of the radiator, so you are seeing a localised swing in temperature close to the radiator as the radiator panel temperature changes which is greatly exagerated compared to the real room temperature changes at a distance.

[asimion]

Are you sure you mean the HR92 valve position (by going to option 10 in the HR92 itself) or do you mean the heat demand for the zone reported in the system devices page in the installer menu on the controller ? The two figures are related but they are not equal.

Yes, I'm reading the value shown on the valve (under pos. 10 of the menu). Btw, what is the meaning of the heat demand for the zone reported in the system devices page in the installer menu on the controller?

I ask because any HR92 valve position below about 30% will not generate a heat demand and thus not cause the boiler to come on at all. If you're still getting a heat demand there might be a configuration problem on your system. Check the heat demand for the zones and boiler relay on the system devices page in the controller to get a better idea of what's happening.

This is new to me, would be the way I would have implemented the system if I were the designer, but....I have to ask if you are sure on the value of 30%. What I observed is that the system only disables cycling when the HR92 reports 0 as valve position. As mentioned, cycle skipping happened also in my case at ramp-up (as there was a quite strong overshoot at the beginning--apr.2C) but it never happens in the proportional band.

This apparently is confirmed by Honeywell: https://ensupport.getconnected.honey...language=en_US.

Regarding the wrong configuration, I don't say this is impossible but I would think it is unlikely. It's not arrogance but, because the system in my case is extremely easy, I do not see what could I do wrong (wiring is trivial, setting up also, the distances is below 10m with direct visibility between elements). Any ideea of what could have went wrong with the installation would be highly appreciated. I will check the heat demand as you mentioned and post it back here later when I'm at home.

I'm quite familiar with the PID algorithms of control (and TPI is an implementation of PID). I know if is the optimum way to control but, the case of Evohome is a bit more special than a simple TPI thermostat. This is because there are in fact two regulating loops that attempt to control the same parameter (T). One loop is done through the duty-cycle of the boiler and the other is done internally by the HR92s. It is a matter of how the priorization between these loops is done. In the case of Honewell, at least in my case, it appears to settle somewhere at 20% valve pos and 40% duty cycle of the boiler( you could as well get the same T by setting a higher vave pos and lower DC of boiler). I miss a functionality where you can choose between optimizing for comfort and optimizing for lower cycles/hour.

Anyhow, I'm already more relaxed now if you say cycling every 10min is not a problem for the boiler system ( I was most afraid about the circulating pump being switched so often).

This scenario has been discussed very recently:

https://www.automatedhome.co.uk/vbul...ing-other-room!!

Btw, what is the "the integrated balancing insert"?.

@RO09 : I will try mounting all the valve controllers and see. Thnks for the tip.