The Efficient Home Forum

Hi all,

I have recently upgraded my Evohome from the BDR91 wireless relay box to the R8810 OpenTherm module.
After doing this, I've noticed an interesting/strange behavior of the boiler pump.

My setup consist of:
A Buderus Logamax Plus GB012-25K V2 combi boiler (On/Off and OpenTherm compatible)
5 zones ( four with HR91 TRVs and one with HR92 TRV)
The CH flow temperature is set to 60 degrees from the boiler's front panel.
The boiler pump is running on its modulating speed setting.

My experiment data:
I've set all the zones to 17 degrees. At this moment, there was no heat demand displayed on Evohome's System Summary screen and the boiler was off.
After a while, a zone (just one zone) displayed a 2% heat demand. Luckily, it was the zone having the HR92. I had a look at the valve position. It was 34.
The 2% head demand started the boiler. However, the flame indicator was off. Only the pump was running. The flow temperature, as displayed on the boiler, was 35 degrees.
I've monitored the pump, and it ran for 30 minutes. After that, I had to leave home. When leaving, it was still on.
During all this time, there was no other zone calling for heat.

Is this behavior normal ?
What is the effect on the pump ? ( I assume that it's affecting its lifetime)
Is this harmful ? I've switched back to BDR91 until I can (hopefully) clarify this.

Looking forward for your opinion about this!
Thank you.

Yes this is normal behaviour for load compensation systems such as OpenTherm.

Will it do any damage to the pump? No, pumps are used to pumping and normally its the stop/starting under TPI (relay) control is more likely to cause them to fail.

OpenTherm control is a much smoother control type than TPI control and with many benefits such as extended appliance life, more efficient boiler operation, less burner hours and increased comfort especially at zone level.

This is an annoying trait. Honeywell's new firmware allows you to monitor each zone. My Intergas HRE 24 OV will switch on the CH pump, fire up, burn some gas, switch off (code 1). The pump will run for anything up to an hour with zero call for heat from any zone. Periodically it will switch off and the boiler will show code 0.

Within a few minutes however it will repeat this over and over. It's almost as if OT is trying to maintain a minimum water temp. All it's doing however is wasting gas keeping the bypass circuit hot as no water at any point goes to a radiator.

I should note that this ONLY happens when a zone is close to its set point. Why run the pump and periodically fire the boiler when all zones are showing 0%?

Over the course of a day this consumes about the same amount of gas as it takes to heat my hot water cylinder daily...

DAVS wrote:This is an annoying trait. Honeywell's new firmware allows you to monitor each zone. My Intergas HRE 24 OV will switch on the CH pump, fire up, burn some gas, switch off (code 1). The pump will run for anything up to an hour with zero call for heat from any zone. Periodically it will switch off and the boiler will show code 0.

Well code 1 is burner shutdown, which means the boiler has reached the 'Control Setpoint' the R8810 OpenTherm Bridge requested, so there must be some sort of demand from the evohome Controller?

DAVS wrote:Within a few minutes however it will repeat this over and over. It's almost as if OT is trying to maintain a minimum water temp. All it's doing however is wasting gas keeping the bypass circuit hot as no water at any point goes to a radiator.

Honeywell uses a system called 'Low Load Control' which identifies the boiler is short cycling and then implements a demand similar to what you are seeing, however there must be something calling for this demand? If there is no demand from the R8810 OpenTherm Bridge, the boiler will not run.

DAVS wrote:I should note that this ONLY happens when a zone is close to its set point. Why run the pump and periodically fire the boiler when all zones are showing 0%?

Right, so when you say 'close to set point' how close? Within 1 degree either side of set point, evohome will see this as a potential 'demand' depending if the temperature is falling or raising. At the end of the day evohome is attempting to maintain the room temperature you have set. If the room is losing energy, it needs to be replaced at the same rate it is lost otherwise the room temperature will fall.

I would suggest potentially your HR92's are not calibrated properly and I would be calibrating them as an initial measure. The communications between the evohome Controller and the OpenTherm Bridge may also be causing the issue, so check the locations in accordance to Page 4 and Page 48 of the evohome Installation guide.

I hope that helps.

Apologies, that was a bit of a bad explanation. It happens when the room temperature and set point are equal. I.e. Room temp shows 19deg with a set point of 19deg. If it's 0.5deg or 1deg lower the HR92 will open and demand shows in the evohome with heat going to where it needs to be. Similarly, if room temperature is above set point all is well and the boiler is not fired, so ok.

The system has recently been upgraded from an old TPI Gloworm boiler to the OT bridge and Intergas HRE and it did strike me that there may be some teething problems initially. It may be a red herring but the zones that this is occurring in have multiple HR92s set as multi-zone and not single-zone. So I can try reverting back there.

I'll check out the HR92s in terms of calibration and see if that solves it. I just thought it was odd behaviour for the OT bridge to fire the boiler when no demand is show on the controller.

On the flip side, when there is demand the system works beautifully.

The plot thickens...

I recalibrated all HR92s which made no difference.

It appears this phenomenon only occurs in the 2 zones that have 2 HR92s in them (as there are 2 radiators in these zones). If the actual temperature of the zone is equal to target temperature the boiler fires and then reverts back to code 1 shortly after. If I monitor the HR92 valve positions one HR92 goes to P14 and the other stays on P00. At these levels the valve isn't open enough for any water to flow so no heat is entering the zone. At no point does any load show on the system summary on the evohome controller.

It's as if the evohome is attempting to add heat to the zone to maintain target temperature but is unable to due to there being no actual radiator flow. I have even tried setting the HR92s to full stroke but this doesn't change the outcome.

If I simulate the same scenario in any other zones that have one HR92 this does not happen and the boiler remains off.

This is suggestive that there's an issue with zones that have multiple HR92s in them or my valves could be the problem. Valves seem unlikely though as they were all brand new <2 years ago and are supported Danfoss RAS C2. All other zones function as expected that have single HR92s. This is quite interesting I must say, if rather frustrating.

The next thing I could try is to bind my evohome controller to one of these zones with multiple HR92s and see if that still results in the same outcome. I do feel like I'm now grasping at straws though...

Well, another day's testing reveals more on Opentherm behaviour.

I have run more tests on both zones with single HR92s and those with 2 HR92s.

I have essentially reset the entire system and re-bound everything, including valve calibrations. None of that has made an ounce of difference. I also carried out a full RF comms check with no issues there.

For zones with a single HR92, the boiler will actuate when actual temperature is 0.5deg below set point. The HR92 will open and the boiler will fire up. Valve position will be something like P64 which is enough for heat to get into the zone, this equates to something like 26% demand. Once the target temperature is reached the HR92 will start to throttle back and demand will gradually go from say 26%, to 22%, 18% and finally 13%. This is roughly equal to a valve position of P22 I think it was. Here's where the system starts to come undone. At that valve opening there is no flow to the radiator but still a call for heat so the boiler does it's usual thing of firing up for a few seconds, going to code 1 and the pump runs, which repeats endlessly. The rad valve is never moved to P00 and the boiler shut off, it will just stay there indefinitely heating up the main circuit even though target temp appears to have been satisfied. Why is there demand if target temperature has been met?

In zones which have 2 HR92s this seems to be exacerbated as the demand appears to be shared between both HR92s, so each HR92 will open to say P24 and P22 which is not sufficient to allow any heat to reach the zone from the start. So the system does not function unless the delta-T is great enough for it to sense a load that is enough for the HR92s to open enough to get heat into the respective zone. This is proven by setting the target temperature to 1deg above actual. However, as with my testing on the single zone, as the zone approaches and reaches target temperature the HR92s are throttled down to the point that flow is reduced to zero but the HR92s still remain open at say P12 and P10 and the boiler goes off on it's endless firing ritual heating nothing but the main loop. Again, if target temp is reached, why doesn't this call for the HR92 to close to P00?

Does anyone know how Evohome is set up? Is the HR92's valve position a crude representation of load or is this based on delta-T? Something is definitely goosed here but I don't seem to be able to figure it out.

Unless my installer has any bright ideas, or we can find some sort of fault somewhere (perhaps a stray wire etc), I think the R8810 will have to go and we'll have to revert back to our original BDR91 as, so far, this is sadly performing worse than my previous 30 year old Glow Worm boiler.

I'll admit this seems like a bit of a ramble but I thought it would be good to document this encase anyone else may come across similar teething troubles, especially if we manage to solve this

I wish I had time to go through this with you but unfortunately my own customers tech support must take priority.

There are just too many variables to consider and evohome is much simpler than the depths you are going into to find a solution... The fact is we would be here hours going over it with you in text form.

Where there are issues like this I would normally FaceTime, WhatsApp or Skype my customers to get a feel for what the install looks like and what things are occurring. You say you have re-paired the entire system, but I would just be starting with one zone only bound to the evohome Controller (the so called trouble zone) and get this working 100% first, before introducing any further variables or zones. You also state that RF signal is ok, but how have you proven this especially with the R8810A OpenTherm Bridge? It sounds like you are creating demands, switching them off, creating demands and switching them off and evohome will not function correctly like this.

The R8810 OpenTherm Bridge receives exactly the same demands the BDR91 Wireless Relay Box does, so swapping the two out will not give you a solution, as the evohome Controller is creating the demand not the R8810. Have you tried cold resetting the OpenThem Bridge (turn boiler off for at least 2-3 minutes and back on)? I would be interested to see its position and how its been wired (picture please).

I suggest giving Honeywell a call on 0300 130 1299 and see if they can remote diagnose any issues first or take my advise by starting with your trouble zone and then moving forwards one zone at a time.

What valves do you use?
I've had a similar problem with soma Italian valves that did not open as quickly as Evohome expects. So the boiler runs but no flow through the radiator. My solution was to set the HR92 to full stroke.

Bart wrote:What valves do you use?
I've had a similar problem with soma Italian valves that did not open as quickly as Evohome expects. So the boiler runs but no flow through the radiator. My solution was to set the HR92 to full stroke.

They're all HR92s. Travel setting doesn't change anything and wouldn't anyway based on the issue at hand.