Evohome optimisation not working

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts
  • dty
    Automated Home Ninja
    • Aug 2016
    • 489

    Evohome optimisation not working

    Good morning,

    I thought I'd ask here before opening a case with Honeywell just in case I've missed something obvious.

    I have a room which appears to take a long time to heat up. I suspect I have an issue with undersized radiators, very odd thermal characteristics in the room, being at the end of a very long 15mm pipe run, etc. But setting that all aside, the radiators *can* get the room warm if given enough time. However, Evohome doesn't appear to be starting the heat-up cycle early enough and so misses its target (by an hour or more) every single day.

    Here's a graph showing what I mean. Be aware that the times shown are in GMT.

    evohome-optimisation.jpg

    The room is meant to be at 21C at 9am (which is 8am GMT on the graph). You can see that it's taken an hour and a half to raise the temperature by 0.8C (from 20.2C to 21C) and it pretty much hit the target exactly an hour late. I recognise that this is *very* slow, but again, it's a function of a strangely shaped space with odd thermal characteristics, couple with some other bits and bobs, plus the fact that I'm using a wall stat to measure the temperature, not the HR92s, so there's always some lag.

    So my question is, given that I have optimum start set to (up to) 3 hours, why isn't Evohome starting this room a hour earlier so it hits the 9am target?
  • DBMandrake
    Automated Home Legend
    • Sep 2014
    • 2361

    #2
    Are you sure you aren't in "delayed start" mode ? The behaviour you describe is exactly how delayed start mode should work - it will be up to temperature about an hour after the set time. (Personally I think the delayed start mode is a waste of time, but that's a whole different discussion...)

    Another thing to be aware of is that optimal start won't pre-heat a zone if it is already within about 1 degree of the target. It will wait until the set time instead. So in your case it was only 0.8 degrees short, so it wouldn't come on early. So whilst it "missed" the target by an hour, it only missed it by 0.8 degrees...

    If you increase your scheduled set point by a degree or two so that it has over one degree to make up when the schedule switches does it then come on early enough to get up to the target temperature on time ?
    Last edited by DBMandrake; 2 August 2017, 11:17 AM.

    Comment

    • dty
      Automated Home Ninja
      • Aug 2016
      • 489

      #3
      Originally posted by DBMandrake View Post
      Are you sure you aren't in "delayed start" mode ? The behaviour you describe is exactly how delayed start mode should work - it will be up to temperature about an hour after the set time. (Personally I think the delayed start mode is a waste of time, but that's a whole different discussion...)
      I couldn't agree more. Absolutely certain it's in optimum start mode, not delayed start.

      Originally posted by DBMandrake View Post
      Another thing to be aware of is that optimal start won't pre-heat a zone if it is already within about 1 degree of the target. It will wait until the set time instead. So in your case it was only 0.8 degrees short, so it wouldn't come on early. So whilst it "missed" the target by an hour, it only missed it by 0.8 degrees...
      I've never heard of this before. Can you point me at more discussion? This would also not explain why it does start half an hour early. Surely if this were the case it would just come on at 9?

      Originally posted by DBMandrake View Post
      If you increase your scheduled set point by a degree or two so that it has over one degree to make up when the schedule switches does it then come on early enough to get up to the target temperature on time ?
      I'll try it overnight tonight. I'll set the schedule for, say, 23C and see what happens.

      Comment

      • top brake
        Automated Home Legend
        • Feb 2015
        • 837

        #4
        Something amiss with your system setup here. Heating systems are usually sized to raise 3degC per hour.
        Looks like there is a problem with mass flow to me.
        I work for Resideo, posts are personal and my own views.

        Comment

        • dty
          Automated Home Ninja
          • Aug 2016
          • 489

          #5
          Originally posted by top brake View Post
          Something amiss with your system setup here. Heating systems are usually sized to raise 3degC per hour.
          Looks like there is a problem with mass flow to me.
          I recognise that. But the room does heat up, so I want to know why Evohome isn’t starting it early enough to hit its target.

          Comment

          • top brake
            Automated Home Legend
            • Feb 2015
            • 837

            #6
            Originally posted by dty View Post
            I recognise that. But the room does heat up, so I want to know why Evohome isn’t starting it early enough to hit its target.
            what is your preheat time set to?
            I work for Resideo, posts are personal and my own views.

            Comment

            • dty
              Automated Home Ninja
              • Aug 2016
              • 489

              #7
              Originally posted by top brake View Post
              what is your preheat time set to?
              3 hours.

              I have some data from this morning which is interesting. Just heading into a work meeting, so will post about it in a bit.

              Comment

              • dty
                Automated Home Ninja
                • Aug 2016
                • 489

                #8
                OK, so here's what happened when I set it to 23C target temperature instead of 21C.

                evohome-optimisation-2.jpg

                So it definitely started earlier (around 6:30am) but still missed the target by 0.6C at 9am. (Remember the graph times are in GMT.)

                The problem is that in this particular room, that 0.6C feels significantly colder.

                So I'm back to square one... I know optimisation is "working" in the sense that it pre-heats the room, but for some reason the calculation is just not working, so it's missing its targets. Perhaps there's a limit to how slowly it believes a room can possibly warm up, and my room is below that?

                Comment

                • paulockenden
                  Automated Home Legend
                  • Apr 2015
                  • 1719

                  #9
                  There has to be limits on any such algorithm.

                  If a room only rises 2.5 degrees in an hour and a half I'd (respectfully!) suggest that the problem isn't Evohome! ;-)

                  HOWEVER... also bear in mind that it might still be learning. See if that hour and a half gets longer over the following days.

                  P.

                  Comment

                  • dty
                    Automated Home Ninja
                    • Aug 2016
                    • 489

                    #10
                    I acknowledged right back in the first post that the heating system has a problem. But I don't see why Evohome can't just say "Hmm... 2C divided by 0.8C/hr = 2.5 hours". It's clearly trying too seeing as it started much earlier when I scheduled the set point to be higher.

                    As for learning... it's had almost a year!

                    Sounds like I need to open a case with Honeywell.

                    Comment

                    • paulockenden
                      Automated Home Legend
                      • Apr 2015
                      • 1719

                      #11
                      Originally posted by dty View Post
                      As for learning... it's had almost a year!
                      I thought you said you'd only just set the higher temp?

                      P.

                      Comment

                      • dty
                        Automated Home Ninja
                        • Aug 2016
                        • 489

                        #12
                        Originally posted by paulockenden View Post
                        I thought you said you'd only just set the higher temp?
                        Right, yes. But at the lower (normal) temperature it's had a year. In other words, it should know the thermal characteristics of the room by now.

                        The higher temperature was for the purpose of science - to see if optimisation was flat-out not bothering, or whether something else was wrong.

                        Comment

                        Working...
                        X