Tracking down random boiler demand with Evohome

[jonstatt]

Plus the weather is kind of irrelevant -

The weather is relevant in the sense that if it's very cold outside, then the chances are overnight the rooms are colder to start with in the morning. As I mentioned earlier, the fundamental dynamics consist of the delta between start temperature and target temperature. If the start temperature is only 1 degree below target, then of course the heating can start later. If it is a room with no windows, it will still have lost heat overnight, and the chances are the colder the weather, the more it lost through the walls of the house. However it will warm up quicker in that room, than one with windows. This is where the learning bit comes in. How long does it take the kitchen to get from 15 degrees to 20 degrees? How long does the bathroom with no windows take to get from 18 degrees to 20 degrees? I would hope the system can cope with all common room variants, but like any learning algorithm, it will have parameter limits. Just to add that it is possible to enable/disable optimisation on a room by room basis if one particular room isn't playing well with it.

[top brake]

The weather is relevant in the sense that if it's very cold outside, then the chances are overnight the rooms are colder to start with in the morning. As I mentioned earlier, the fundamental dynamics consist of the delta between start temperature and target temperature. If the start temperature is only 1 degree below target, then of course the heating can start later. If it is a room with no windows, it will still have lost heat overnight, and the chances are the colder the weather, the more it lost through the walls of the house. However it will warm up quicker in that room, than one with windows. This is where the learning bit comes in. How long does it take the kitchen to get from 15 degrees to 20 degrees? How long does the bathroom with no windows take to get from 18 degrees to 20 degrees? I would hope the system can cope with all common room variants, but like any learning algorithm, it will have parameter limits. Just to add that it is possible to enable/disable optimisation on a room by room basis if one particular room isn't playing well with it.

Well put

Consider this:

I have overnight setback to 15 degrees and morning setpoints at 20 degrees

A radiator heating system is sized to raise 3 degrees per hour so on a normal day it will take anything between 1 and 2 hours to raise temperatures to comfort level

If it is a colder day with more heat loss it may take a little longer, conversely in the spring it may take a shorter time. Hence enabling optimum start allows the evo to hit setpoint regardless of the weather.

[orange]

How does an evohome system with no gateway know the outside temperature ?

Does optimisation not work on an 'unconnected' comtroller

What I'm saying is that 'current' outside temperature has no direct effect on the optimisation function. Obviously it would effect the starting temp and curve but this changes slowly over time as part of the learning process. Check the posts on the subject.

That assumes the function works - And there is some doubt about that

[orange]

http://www.wordpress-1219309-4387497.cloudwaysapps.com/vbulletin/showthread.php?4033-evohome-amp-outside-temp-amp-window-sensors&p=19750#post19750

And

http://www.wordpress-1219309-4387497.cloudwaysapps.com/vbulletin/showthread.php?4048-Evohome-outside-temp-Part-2&p=19866#post19866

--------

But what seems to be happening is that optimisation doesn't work for some people - the outside temperature is a RED HERRING.

In these situations 'optimisation' is a total waste of energy. We just don't know what situations these are and if Honeywell will fix it ( I know the answer to the 2nd question)

[orange]

A radiator heating system is sized to raise 3 degrees per hour so on a normal day it will take anything between 1 and 2 hours to raise temperatures to comfort level

If it is a colder day with more heat loss it may take a little longer, conversely in the spring it may take a shorter time. Hence enabling optimum start allows the evo to hit setpoint regardless of the weather.

Are you saying 'officially' that that is how evohome works ? That it 'knows' the outside temperature ? That it allows 1 hour for every 3 degrees. That, even if it only needs to up the temperature 1 degree, it will switch on 1 hour early to ensure it hits the switch point ?

Please read the document you posted earlier. It LEARNS the room curve. Johnstatt's post is correct IMO - my gripe is that these are not odd rooms with weird curves - and it shouldn't default to just wasting gas.

Top brake - you've already said that optimised start works properly for you - which is great. But are you not interested on why it doesn't work properly for other people? Are you interested in why we are getting these inconsistent boiler startups? Are you not interested in getting answers to the questions raised in this thread ?

[top brake]

Are you saying 'officially' that that is how evohome works ? That it 'knows' the outside temperature ? That it allows 1 hour for every 3 degrees. That, even if it only needs to up the temperature 1 degree, it will switch on 1 hour early to ensure it hits the switch point ?

Please read the document you posted earlier. It LEARNS the room curve. Johnstatt's post is correct IMO - my gripe is that these are not odd rooms with weird curves - and it shouldn't default to just wasting gas.

Top brake - you've already said that optimised start works properly for you - which is great. But are you not interested on why it doesn't work properly for other people? Are you interested in why we are getting these inconsistent boiler startups? Are you not interested in getting answers to the questions raised in this thread ?

Recommend that you log a call with consumer support, thanks

[erik]

Wow, lots of posts here.

And the answer to harder questions (like why doesn't the optimizing work?) seems to be:

-it works for me, if it doesn't work for you, just turn it off.

or

-call support (who will probably say you need to go to your 'installer').

Ouch. It would be really really nice to have some more in-depth insight in the reason why the system makes the decisions that it does. That would really help clearing cases like this up. Now it's just wild guessing.

[jonstatt]

How does an evohome system with no gateway know the outside temperature ?

Does optimisation not work on an 'unconnected' comtroller

What I'm saying is that 'current' outside temperature has no direct effect on the optimisation function. Obviously it would effect the starting temp and curve but this changes slowly over time as part of the learning process. Check the posts on the subject.

That assumes the function works - And there is some doubt about that

That's not what anyone is saying. Evohome doesn't know the outside temperature and it doesn't need to. The FAQ referred to "weather". It doesn't really know the weather, but the weather will affect how the temperatures drops in every room over night so there is still a direct relationship between "the weather" and how EVOHOME reacts. All EVOHOME knows is the temperature in any room at any given time, and learns how long it takes to get from that temperature to the target one. That's it. All we are saying is that the weather affects what temperature deltas EVOHOME will have to deal with throughout the course of the year. It's a causal relationship.

[HenGus]

That's not what anyone is saying. Evohome doesn't know the outside temperature and it doesn't need to. The FAQ referred to "weather". It doesn't really know the weather, but the weather will affect how the temperatures drops in every room over night so there is still a direct relationship between "the weather" and how EVOHOME reacts. All EVOHOME knows is the temperature in any room at any given time, and learns how long it takes to get from that temperature to the target one. That's it. All we are saying is that the weather affects what temperature deltas EVOHOME will have to deal with throughout the course of the year. It's a causal relationship.

An interesting thread. One of the warmer nights last night. First radiator 'whirred' 5 minutes before its programmed start up time. On checking, when I got up, all the zones that were on were 2 to 3C degrees above the target temperature according to the controller. I assume this is because the radiators get hot, stay hot for longer and room temperature is slow to fall away. Two hours on, all the zones are within 1C of their target temperatures. I conclude that the system is telling me that the boiler temperature is probably set too high for the current ambient conditions. Colder weather is expected over the weekend so I will leave things as they are for the moment.

[Mavis]

Right, last night I went to bed with the controller and kept watch.

The general lowest temp during the night around the whole house (apart from our bedroom) was 16c. All radiators apart from our bedroom have optimisation activated.

Hall - Set point 19c at 6am. Rad came on at 5.10 with an actual temp of 17c. Temp at 6am - 19.5c. Radiator felt cold at 6am (ie it had been off a while)
Bathroom - - Set point - 19c at 5.50. Rad came on at around 5.20 with an actual temp of 16.5c.
Sitting Room - Set point - 18c at 6am. Rad came on at 6am with an actual temp of 17c
Bedroom 2 - Set point - 18c at 6am. Rad came on at 6am with actual temp of 17c
Kitchen - Set point - 19c at 6.20am. Rad came on at about 6.05am with actual temp of 17c
Bedroom 1 - Set point 17c at 6.40am Rad came on at 6.40am with actual temp of 12c (no optimisation set) (Window was open during the night but think there is an issue with this rad setup as I took the DTS92 out of this room. Room temp at 17c is red hot so will maybe delete and rebind)


So my very basic conclusion is that optimisation only kicks in if the temp is >1c from the setpoint. But still not sure why it kicks in so early.

[orange]

That's not what anyone is saying. Evohome doesn't know the outside temperature and it doesn't need to. The FAQ referred to "weather". It doesn't really know the weather, but the weather will affect how the temperatures drops in every room over night so there is still a direct relationship between "the weather" and how EVOHOME reacts. All EVOHOME knows is the temperature in any room at any given time, and learns how long it takes to get from that temperature to the target one. That's it. All we are saying is that the weather affects what temperature deltas EVOHOME will have to deal with throughout the course of the year. It's a causal relationship.

agreed - (I think we're saying the same thing in different ways) - but it's (IMO) it's essentially irrelevant to the way that optimisation works

[orange]

Right, last night I went to bed with the controller and kept watch.

The general lowest temp during the night around the whole house (apart from our bedroom) was 16c. All radiators apart from our bedroom have optimisation activated.

Hall - Set point 19c at 6am. Rad came on at 5.10 with an actual temp of 17c. Temp at 6am - 19.5c. Radiator felt cold at 6am (ie it had been off a while)
Bathroom - - Set point - 19c at 5.50. Rad came on at around 5.20 with an actual temp of 16.5c.
Sitting Room - Set point - 18c at 6am. Rad came on at 6am with an actual temp of 17c
Bedroom 2 - Set point - 18c at 6am. Rad came on at 6am with actual temp of 17c
Kitchen - Set point - 19c at 6.20am. Rad came on at about 6.05am with actual temp of 17c
Bedroom 1 - Set point 17c at 6.40am Rad came on at 6.40am with actual temp of 12c (no optimisation set) (Window was open during the night but think there is an issue with this rad setup as I took the DTS92 out of this room. Room temp at 17c is red hot so will maybe delete and rebind)


So my very basic conclusion is that optimisation only kicks in if the temp is >1c from the setpoint. But still not sure why it kicks in so early.

in my view they all look about right - ignoring bedroom 1 the only one with a potential issue is the Hall that came on 50 minutes early to heat up by 2 degrees which it did quickly (as radiator had cooled by 6am)

[G4RHL]

Well put


A radiator heating system is sized to raise 3 degrees per hour so on a normal day it will take anything between 1 and 2 hours to raise temperatures to comfort level

If my heating took that long to warm I would be complaining.

[G4RHL]

[QUOTE=Mavis;21300]Right, last night I went to bed with the controller and kept watch.

Thats not a nice name to give your husband!

[jonstatt]

If my heating took that long to warm I would be complaining.

That's easily how slow my kitchen is to warm up and its a relatively modern house (built 2006) with polystyrene in the cavity of the walls. The trouble is that the windows have to be wooden framed due to being a protected area. And I don't like to keep the door closed on the kitchen so some of that heat escapes into the hallway