Evohome - poor temperature regulation

[JohnnyP]

The 1 minute in 10 minute boiler cycling may be stopped by adding a time delay relay after the Honeywell boiler controller. There's a thread on here from about 2009 dealing with it...

[roydonaldson]

How have you configured your BDR91s for S Plan Plus? What did you do regarding binding the hot water?
Thanks

3 BDR91's connected to 3 zone valves. 1 for upstairs zone, 1 for downstairs zone, 1 for DHW.

Roy.

[The EVOHOME Shop]

3 BDR91's connected to 3 zone valves. 1 for upstairs zone, 1 for downstairs zone, 1 for DHW.

Roy.

I think top brake was asking how these have been configured within the evohome system... Are both heating BDR91's configured as a 'boiler relay', was the hot water system configured as 'two or three port valve' or 'hot water only'?

[roydonaldson]

BDR91's for heating are configured as zone valves, controller by wall sensors in each zone. BDR91 for heating is configured as hot water only. No Boiler Relay setup. (need to get 1 radiator valve changed before I can do this).

Roy.

[erik]

The 1 minute in 10 minute boiler cycling may be stopped by adding a time delay relay after the Honeywell boiler controller. There's a thread on here from about 2009 dealing with it...

Thanks for the hint. I'd prefer it if Honeywell would fix the issue though.

[The EVOHOME Shop]

BDR91's for heating are configured as zone valves, controller by wall sensors in each zone. BDR91 for heating is configured as hot water only. No Boiler Relay setup. (need to get 1 radiator valve changed before I can do this).

Roy.

Hang on, do you have HR92's fitted behind the zone valves? If so what is triggering the zone valves to open when you have HR92 demand and them configured as 'zone valves'?

[roydonaldson]

Heh, thought we'd get there. So, 2 zones, down and up. In hallway (down and up) there are radiators with no TRV's on them and one DT92 in each zone ( up and down). These are set to control the down and up zone valves at the boiler. These effectively control the upstairs and downstairs heating coming on. So, they replicate the CM927's that they replace.

Now, what you were asking. HR92's on radiators in rooms. With no boiler relay, they are effectively just turning their radiators on/off and not controlling boiler demand. But, most of the time, these are either on when the up/down zones are running, or off altogether.

Roy.

[The EVOHOME Shop]

Heh, thought we'd get there. So, 2 zones, down and up. In hallway (down and up) there are radiators with no TRV's on them and one DT92 in each zone ( up and down). These are set to control the down and up zone valves at the boiler. These effectively control the upstairs and downstairs heating coming on. So, they replicate the CM927's that they replace.

Fine...

Now, what you were asking. HR92's on radiators in rooms. With no boiler relay, they are effectively just turning their radiators on/off and not controlling boiler demand. But, most of the time, these are either on when the up/down zones are running, or off altogether.

Roy.

OK, so are you saying these are also bound to evohome as 'zones' behind the zone valve(s)? If so these are gonna have a sod of a job learning and will have little or no chance of being accurate...

[roydonaldson]

Yes, they are set up as zones. It does work surprisingly well. The main advantages for me are that I can have the majority of my house turned off every morning and this works perfectly (ie, the HR92' are told to be closed, so dead easy). It does work pretty well at night as well, as the main hallway is on, so there is demand and the 'remote' zones are effectively controlling how much each radiator consumes of this demand.

The one area where it does not work is if the hallway is satisfied and a remote zone needs to call for more heat. It can't do this at present.

I need to get a plumber in to change the valves on the hallways radiators to have TRV's on them. I've been traveling every week for the last month and a half, so not been there enough to get it done. I also need to find a plumber that will do it without draining the whole house down.

[The EVOHOME Shop]

Yes, they are set up as zones. It does work surprisingly well. The main advantages for me are that I can have the majority of my house turned off every morning and this works perfectly (ie, the HR92' are told to be closed, so dead easy). It does work pretty well at night as well, as the main hallway is on, so there is demand and the 'remote' zones are effectively controlling how much each radiator consumes of this demand.

The one area where it does not work is if the hallway is satisfied and a remote zone needs to call for more heat. It can't do this at present.

I need to get a plumber in to change the valves on the hallways radiators to have TRV's on them. I've been traveling every week for the last month and a half, so not been there enough to get it done. I also need to find a plumber that will do it without draining the whole house down.

I have done a lot of systems and I do not see why you would want to set the evohome system up like this? A configuration like you have completely defeats the point of evohome, as each HR92 controlled zone can only get heat when either of the hallway 'zones' has demand, so in essence is no different to a house with twin zoning and manual TRV's!?

5 minutes of reconfiguration on evohome and I would have your house running 20-30% more efficient than it currently is and a dam sight more comfortable...

[roydonaldson]

I'm open to suggestions.

[The EVOHOME Shop]

Well, I would ignore the two hallways for now and be happy with the fact that these will come on every time you have HR92 or hot water demand. You can even throttle these radiators back using the current manual hand-wheel valve on the radiator so the hallway does not get too warm.

A quick way of doing things temporarily would be to bind in the two BDR91's controlling the heating motorised zone valves together as 'boiler relays'. I appreciate you (at present) will get the radiators come on in the hallways when you have HR92 or hot water demand, hence I suggest throttling them back or just make sure your hot water scheduling is the same as a larger cooler HR92 controlled zone for now.

In a perfect world, I would be getting rid of your two heating motorised zone valves (don't physically have to take them out, just electrically disconnect them and put lever in the manually open position), HR92's on all radiators, 1x BDR91 'boiler relay' (to enable cycle rate and minimum on time feature to fine tune your system) and 'hot water only' for your hot water using 1x BDR91 and CS92. Also ensure you have an Automatic Bypass Valve when you go HR92's all rads.

You can use your DTS92's as the sensors in any of your rooms that are HR92 controlled, should you feel they would benefit over being controlled locally at the HR92.

At least with the above configuration the evohome system will genuinely only heat the room(s) you are in (with exceptions of the hallway for now) to the correct set point at the correct timings. Hot water doesn't take long to heat, so don't worry too much about this heating the hallways for now as the energy will not be wasted.

Forgot to say, make sure you delete the two hallway zones.

[G4RHL]

G4RHL, sadly that doesn't work in my case. I've been testing while being out of the house whole day, perfect conditions, and saw very much sub-perfect results.

But with the old wired Honeywell Round I'm seeing very good results. So I know that Honeywell CAN work very well in my home. I just wish it had zoning

Erik, you will like this! Sitting in my front lounge this morning I could hear the boiler going on and off, sometimes on for less than a minute but the gaps between varied from a few minutes to 10 or more. Reason? Perhaps because it is a warm day here at present, downstairs (4 zones) is set to 18 and all were being kept to that or close to it - 17.5 to 18.5 although the radiator next to me has just gone up to 19. On the other hand is not this what heating is meant to do - maintain a temperature close to that set?

As I typed the last sentence I heard the boiler come on for a short time, then go off, then back on again. In each case for less than a minute but no doubt the reason for that is 4 zones wanting 18 and all increasing or decreasing in temperature at different rates and therefore one is calling for heat when another does not need it. Again that seems normal to me.

It makes me wonder whether I should combine 3 of the zones into one. I kept them separate initially to get used to it all but now three months in more often than not three of the zones are at similar temperatures most of the time and doors between them tend to be left open. Am learning all the time!!

[erik]

I'm almost always testing with just 1 zone at a time enabled to make sure the different zones aren't affecting each other's heat demand etc. If there's heat demand, I know exactly wich zone is responsible for it.

[greyhound1234]

Observation/test result: yesterday, I reset/rebinded everything and put my living room to zone 2 instead of zone 1 and the result had not improved. Actually, I noticed the HR92's were being closed while still asking for heat. I changed the HR92's to full stroke and tested again today. The change resolved the HR92's being closed while asking for heat. However, I'm now seeing the same behavior again: putting in heat every 10 minutes for an hour or two. Creating a build up to at least 0.5 above setpoint (Evohome shows 0.5 above setpoint. My own thermometer actually goes up a whole degree). Afterwards, it cools down for about 2 hours, right untill setpoint is reached. Then it starts heating again, building up the 0.5-1.0 overshoot again. So the conclusion is: using zone 2 instead of zone 1 does not improve my setup.

If anyone has other suggestions of things I might try, I'd like to hear it.

I will jump back in on my own post now. Wife has basically got fed up of me constantly watching the Evohome controller like a hawk, and told me to give it a rest for a bit...
Some other interesting discussions going on, please continue with them.

So I rebound in a different order, and like Erik found, it makes no difference to the bathroom overshoot. There goes that theory.

So since I did my full reset, I still have not been able to reproduce the 'repeated driven overshoot' that I used to get. I am even wondering if I only saw that with thermostat mode, and that the rest of the time I have been seeing effects of TPI.

In the case of the bathroom, this is more like a passive overshoot...it is the smallest room so I would expect that room to be hardest to control (plus I don't really mind it drying the towels, and I could always use the lockshield valve if need be). Bedrooms and lounge are well controlled.

Certainly, some of the behaviours I have observed are explained by an over-zealous proportional-integral controller (i.e. the integral component causes overshoot). One case that really threw me at first, having come from a proportional only thermostat, was the following. Set point and actual = 17. Then override set point to something obscene like 23. After an hour, when the actual temperature gets to 19, then set it back to 17 again. You will find that not only will the temperature keep increasing due to the heat in the system, the large integral-error term that has build up will mean for the next 30 mins or so, it will actually keep calling for heat.
This is obviously an unrealistic case but demonstrates what can happen if you give it a big disturbance.

Large disturbances to the system can really make it unstable, so I have now changed my program to avoid any steps in temperature of more than a few degrees. It is much better.

When settled - and that means after 2 hours of each zone being at the set point - the system can maintain the desired zone temperatures perfectly (within 0.5), even when zones are at temperatures of 3 degrees different from each other. It works very well if you hold temperatures constant in all zones, and do not have too many switch points.

This accumulated integral error effect (which having googled it, is called 'integral wind up') does not however explain the 1.5 degree driven bouncing that I definitely see in thermostat mode (which I don't use anyway). I have not been able to reproduce this effect since my reset (which is both good and bad).

So the warning is: this is a complicated, intertwined system driven by a P-I algorithm. In my house, it takes 2-3 hours to settle to an equilibrium so that each zone is at the temperature required, and receiving the heat required to maintain that temperature for the loop temperature. if you 'fiddle' with the system to see if it can hold a temperature or see how the other zones behave, it will disturb everything, and will take a few hours to find that new equilibrium (in mild weather...only 30-60 minutes in cold weather so you have a larger sink for heat).
Of course the controller is also trying to take that demand, and put it into 'lumps' for the boiler. From what I have seen, it is also trying to decide when the temperature error is small and not to trigger the boiler and wear it out. For example, if all zones are at set point except one zone which is 0.5 lower than set, it won't do anything for 30 mins or until another zone also needs heat before firing. This is called a 'deadband' according to my googling :-)

See, if someone had put that in the manual, that that we have is a series of closely coupled non-linear systems driven by a P-I algorithm, I would have treated it much more gently! I guess people don't write user manuals for physicists.