HR92 Observations on temperature and control processing

[g6ejd]

1. Now I have the ability to do an IR Image assessment (thanks PaulO for the FLIR link) I have discovered in some instances of my 14 HR92's that the ambient temperature around them does not correspond with the set temperature or displayed temperature and typically some (but not all) of the HR92's read about 1°C higher than actual. So have have used the calibration facility to remove 1°C (setting = -1) where the HR92 is over reading.

2. What's been happening is there are a few locations where as I call it, a micro-climate exists around the HR92, for example one is in the corner of a room with little or no air movement around it and I can now see this creates a hotter spot and the offset is now corrected for. So you may wish to consider the same scenarios in your installations. Of course you can compensate by increasing the zone temperature to above that expected, perhaps 23°C for an achieved 21°C ambient.

3. Recently I set only one zone to come on early; our bedroom and heard the valve open at 04:00 (actually 04:02) as programmed, the heating began as I lay watching the room temperature projected on the ceiling () but to my surprised I heard the valve close a little after about 10-mins, the room turn-around time is about 20-mins from cold, then it turned (I presume) the heat off even more as it approached temperature. I found this interesting because it was the only zone on at that time, so why constrict water flow to achieve heat control, why not use the TPI to turn off demand earlier and leave the valve open? Interesting control logic I think.

[Rameses]

3) to keep heat there and not lose the stored heat in the rads. If another zone had kicked in it would have flushed your lovely warm rad which was still perfectly delivering heat to the room. The system was utilising the stored heat and then calculating the demise curve, depending how long you asked the setpoint for it would have kicked in again. Btw Not sure what size room you or what temp you were going from and too but 20mins is not a long time.

[Rameses]

2) good observation and yes. This is why many customers also opt for dts91 to get a sensing point in the room nearer their comfort area. Or just set it to what you feel comfortable anyway. Measurement and what you feel are key factors. The system provides an idex of measurement but what you feel comfortable at is ultimately your choice and what we think is most important

[g6ejd]

3) to keep heat there and not lose the stored heat in the rads. If another zone had kicked in it would have flushed your lovely warm rad which was still perfectly delivering heat to the room. The system was utilising the stored heat and then calculating the demise curve, depending how long you asked the setpoint for it would have kicked in again. Btw Not sure what size room you or what temp you were going from and too but 20mins is not a long time.

Thanks for the insight, but if I I understand you correctly my system does not have a boiler relay and so can't utilise stored heat, if the pump is on, the boiler is on. I might get an additional BDR91 to utilise that capability in this Y-plan system. The house is fairly new and so needs little heat to get it to temperature, the bedroom is 6Mx3.5M and has two radiators one at either end.

[Rameses]

Thanks for the insight, but if I I understand you correctly my system does not have a boiler relay and so can't utilise stored heat, if the pump is on, the boiler is on. I might get an additional BDR91 to utilise that capability in this Y-plan system. The house is fairly new and so needs little heat to get it to temperature, the bedroom is 6Mx3.5M and has two radiators one at either end.

Yes but if your pump /boiler turns of its not as if your radiators are suddenly ambient room temp are they? Two large bodies of water and metal filled with 60+ degrees water will continue to drive heat for some time.

[rcopus]

3) to keep heat there and not lose the stored heat in the rads. If another zone had kicked in it would have flushed your lovely warm rad which was still perfectly delivering heat to the room.

This doesn't make sense on a modern heating system, every radiator in my house has its own flow and return feed so there's no way to flush it with cold water by another HR92 onpening.

I factory reset my Evohome and stated from scratch yesterday because the temperatures were either below set-point of shot way above them, hopefully it will 'learn' something better after the reset.

Part of the problem (in my setup) I see with how the valves work is that they don't open anywhere near as much as they could in order to heat a room. It seems Evohome would happily run the boiler for longer/more often than just open the valve more. My house is brand new and the radiators all reach a temperature too hot to touch within minutes.
Then there's the opposite problem with valves not being closed enough when another zone demands heat causing the zone that was at temperature to massively overshoot.
I've also noticed that when a room needs to heat up from cold (say 18 to 20.5) it runs the boiler for immense amount of time to the point where 3 towel radiators can't dissipate the heat and the boiler starts cycling. Then I end up with an overshoot because the boiler could of heated up the radiator in that zone for 5 mins and then waiter another 5 mins before adding more heat etc.

I thought the logic on Evohome was more intelligent than just having a constant demand at the boiler until the zone is nearer the set point. In my setup this just results in 20+ minutes of the boiler trying to heat water that's already returning very hot to where the boiler can't modulate any lower and starts cycling. Unfortunately due to the fact the boiler put in by the house builder is a BAXI it further compounds the problem by automatically relighting three minutes after turning the burner off regardless of what the flow return temp is, which of course is still too high causing the burner to cut out again seconds after relighting.

[Rameses]

This doesn't make sense on a modern heating system, every radiator in my house has its own flow and return feed so there's no way to flush it with cold water by another HR92 onpening.

I factory reset my Evohome and stated from scratch yesterday because the temperatures were either below set-point of shot way above them, hopefully it will 'learn' something better after the reset.

Part of the problem (in my setup) I see with how the valves work is that they don't open anywhere near as much as they could in order to heat a room. It seems Evohome would happily run the boiler for longer/more often than just open the valve more. My house is brand new and the radiators all reach a temperature too hot to touch within minutes.
Then there's the opposite problem with valves not being closed enough when another zone demands heat causing the zone that was at temperature to massively overshoot.
I've also noticed that when a room needs to heat up from cold (say 18 to 20.5) it runs the boiler for immense amount of time to the point where 3 towel radiators can't dissipate the heat and the boiler starts cycling. Then I end up with an overshoot because the boiler could of heated up the radiator in that zone for 5 mins and then waiter another 5 mins before adding more heat etc.

I thought the logic on Evohome was more intelligent than just having a constant demand at the boiler until the zone is nearer the set point. In my setup this just results in 20+ minutes of the boiler trying to heat water that's already returning very hot to where the boiler can't modulate any lower and starts cycling. Unfortunately due to the fact the boiler put in by the house builder is a BAXI it further compounds the problem by automatically relighting three minutes after turning the burner off regardless of what the flow return temp is, which of course is still too high causing the burner to cut out again seconds after relighting.

Give it 7-10 days then DM me your Honeywell login and we will look at how your system is doing.

[DBMandrake]

1. Now I have the ability to do an IR Image assessment (thanks PaulO for the FLIR link) I have discovered in some instances of my 14 HR92's that the ambient temperature around them does not correspond with the set temperature or displayed temperature and typically some (but not all) of the HR92's read about 1°C higher than actual. So have have used the calibration facility to remove 1°C (setting = -1) where the HR92 is over reading.

It's also been my experience that HR92's over read by about 1 degree when the radiator is hot due to direct heat pickup so most of mine (not all) are set to -1 calibrate as well. Of course this over-reading only applies when the radiator is hot so when the radiator is cold the indicated temperature with calibrate -1 is now 1 degree below actual and this can cause the radiator to come back on again when the room is not in fact too cold. Also when there is no convection in the room due to the radiator being cold you get temperature stratification between the floor and ceiling - if I check with an IR gun there is more than 2 degrees difference between the floor and ceiling for example. Because the HR92 is close to the floor I find it typically reads an additional degree lower than a wall mounted DTS92 due to the area near the floor being genuinely colder. (So together with -1 calibrate the reading is typically 2 degrees lower than the "real" reading when the radiator is cold)

A remote sensor like a DTS92 solves both these issues as it is both away from the radiator and up away from the floor. (Recommended height, 1.2 metres) No calibration offset is required and the reading is always correct regardless of whether the radiator is hot or not. (As long as the DTS92 is well positioned and at least 1.5 metres from the radiator)

2. What's been happening is there are a few locations where as I call it, a micro-climate exists around the HR92, for example one is in the corner of a room with little or no air movement around it and I can now see this creates a hotter spot and the offset is now corrected for. So you may wish to consider the same scenarios in your installations. Of course you can compensate by increasing the zone temperature to above that expected, perhaps 23°C for an achieved 21°C ambient.

Yes this can certainly be an issue, especially if the HR92 is right in a corner, boxed in or otherwise obscured - if you don't get a good strong convection current around the room the HR92 will read too high, possibly several degrees, leading to a room that is colder than claimed. This can also be a problem with old fashioned non-convector panels as they (not surprisingly) don't convect very well and rely more on direct IR radiation. If you're seeing more than one degree of offset you're probably better off with a remote wall mounted sensor near the occupants, as applying a large correction like 2 or 3 degrees will only be right under specific circumstances and only make the problem of the radiator coming on when it shouldn't (as described above) even worse. The radiator will tend to cycle on and off a lot if the HR92 is suffering from localised heating and poor convection instead of maintaining a steady temperature.

Our kitchen has this problem as the radiator is boxed in by two kitchen cabinets and a worktop, the HR92 is literally about 2 inches from the side of the cabinet. It doesn't get a good reading of the room temperature at all and tends to cycle a lot and shows a large amount of overshoot on my graphs, although the overshoot is really only the temperature near the HR92 overshooting not the room as a whole. Again a remote DTS92 would solve this completely but I'm not bothered enough about the exact temperature of the kitchen to spend the money on that - as long as its warm when I'm up for breakfast I'm happy.

3. Recently I set only one zone to come on early; our bedroom and heard the valve open at 04:00 (actually 04:02) as programmed, the heating began as I lay watching the room temperature projected on the ceiling () but to my surprised I heard the valve close a little after about 10-mins, the room turn-around time is about 20-mins from cold, then it turned (I presume) the heat off even more as it approached temperature. I found this interesting because it was the only zone on at that time, so why constrict water flow to achieve heat control, why not use the TPI to turn off demand earlier and leave the valve open? Interesting control logic I think.

It's normal for it to start closing the valve progressively well in advance of reaching the target - its a PID controller so has a differential component to the control - if the temperature in the room rises quickly (as it does in our bedroom too, because there is a fairly large type 22 double panel convector radiator) then it will learn this and learn to prevent overshoot by progressively closing the valve as the target is reached, ideally stopping on a dime at the correct temperature. Radiators have a lot of thermal inertial due to the heat storage of the steel and water which would cause an overshoot of 2-3 degrees or more if it waited until the target was reached before turning off the radiator.

Keep in mind that just because you can hear the motor turning to close a bit does not mean the flow is always actually reduced - there is quite a bit of "dead band" on either side of the active part of the valve operation. For example if the black wheel can turn 10 turns between fully closed and fully open, you'll probably find that from the point where the valve starts flowing until it is fully flowing is only about 2-3 turns. (Typically 40-80% indicated valve opening in the menu is the full useful range of the TRV valve body)

So when a large temperature jump is scheduled it will fully open the valve, as the target approaches it will keep making small adjustments to progressively close it but initially these adjustments won't have any real effect on the flow of water until you approach the action point of the valve.

Regarding TPI - it actually does both valve control and TPI (via sending heat demand messages) in parallel so the final result is a combination of the two. The way it works is quite clever.

The heat demand sent from a single HR92 is quite "low" in proportion to how far its valve is open. What this means is that when only one radiator is calling for heat (for example the bedroom at night) and only a small amount of heat is required for example to maintain 16 degrees, it reaches a balance where the valve is largely open but the TPI duty cycle is low. I have seen 80% indicated valve opening and an average flow temperature of 50 degrees for example. However as soon as other zones start calling for heat the total call for heat goes up, TPI increases the duty cycle so the average flow temperature goes up, now the HR92 will detect that the room is starting to get too hot so it will close the valve a lot more. Now you might have 4 zones running, the original bedroom zone now has the valve open 40% but with a 70 degree flow temperature to maintain the same average amount of heat going into the room.

It automatically finds the optimum equilibrium that satisfies the individual rooms with the lowest TPI duty cycle and flow temperature possible. The only drawback is there can be noticeable overshoots in one zone like the bedroom when other zones switch on as the first zone was relying on the low TPI duty cycle to maintain the right temperature so it has to detect an overshoot and close the valve reactively - it has no knowledge that other zones have just caused the TPI duty cycle to increase dramatically. All it knows is that its own room is suddenly getting warmer and that it needs to close the valve down and signal less heat demand.

[DBMandrake]

This doesn't make sense on a modern heating system, every radiator in my house has its own flow and return feed so there's no way to flush it with cold water by another HR92 onpening.

On the contrary, it can and does happen. If you have one radiator flowing and all the rest are off your boiler will be modulating its burner (or cycling on and off) to try to maintain your flow temperature based on the warm water coming back from that one radiator.

Now what happens if 5 cold radiators open simultaneously ? All that cold water will come rushing back to the return pipe of the boiler and cool the heat exchanger right down - even if the boiler is burning at maximum burn the output flow temperature can't help but drop dramatically, maybe 20 degrees initially, and it could take a good 5-10 minutes for the flow temperature to recover as the 5 additional radiators warm up. In the meantime if the first radiator was still flowing it would be cooled down by this drop in flow temperature.

I factory reset my Evohome and stated from scratch yesterday because the temperatures were either below set-point of shot way above them, hopefully it will 'learn' something better after the reset.

Part of the problem (in my setup) I see with how the valves work is that they don't open anywhere near as much as they could in order to heat a room. It seems Evohome would happily run the boiler for longer/more often than just open the valve more. My house is brand new and the radiators all reach a temperature too hot to touch within minutes.

Not sure what you mean there - how are you determining how far open the valves are ? And at the time you check what is the set point and what is the measured room temperature ? The HR92's will always open 100% if the set point is at least 1.5 degrees above the current measured temperature. If your radiators are not flowing properly under these conditions your valves might not be fully compatible with the HR92. (Pin dimensions/travel etc) You could try enabling full stroke mode if this is the case.

Then there's the opposite problem with valves not being closed enough when another zone demands heat causing the zone that was at temperature to massively overshoot.

What do you call massive ? It is true if you had one radiator running at a low temperature then another zone is scheduled to come on you might see an overshoot of up to about 1 degree initially.

I've also noticed that when a room needs to heat up from cold (say 18 to 20.5) it runs the boiler for immense amount of time to the point where 3 towel radiators can't dissipate the heat and the boiler starts cycling. Then I end up with an overshoot because the boiler could of heated up the radiator in that zone for 5 mins and then waiter another 5 mins before adding more heat etc.

I can't actually follow what you're trying to say there...if a room is below temperature it needs to call for heat from the boiler does it not ? Is it a modulating boiler or just an on/off thermostat type. (like mine)

I thought the logic on Evohome was more intelligent than just having a constant demand at the boiler until the zone is nearer the set point. In my setup this just results in 20+ minutes of the boiler trying to heat water that's already returning very hot to where the boiler can't modulate any lower and starts cycling.

Again, I don't really understand what your perceived issue is - I think you're misunderstanding how things are supposed to work.

If zones are 1.5 degrees or more below their set points the system will call for the boiler to be on 100% - why wouldn't it ? The rooms need to heat up...

If the return flow is too hot that suggests your pump speed may be set too high. If the boiler can't modulate low enough that suggests it might be oversized for the house... but there is nothing wrong with the boiler cycling of and on when the demand is low, and if its oversized for the house and doesn't have a sufficient modulation range then that's what it will have to do.

None of this is really anything to do with the Evohome.

Unfortunately due to the fact the boiler put in by the house builder is a BAXI it further compounds the problem by automatically relighting three minutes after turning the burner off regardless of what the flow return temp is, which of course is still too high causing the burner to cut out again seconds after relighting.

Sorry but that sounds like a design flaw or maladjustment of the boiler - and completely out of scope of the Evohome system.

Edit: one other thing occurs to me, do you have HR92's on all radiators, and if so do you have an automatic bypass valve in the system ?

[rcopus]

Give it 7-10 days then DM me your Honeywell login and we will look at how your system is doing.

Thanks Rameses thats kind of you.