Page 1 of 3 123 LastLast
Results 1 to 10 of 21

Thread: EvoHome - Optimizing

  1. #1
    Automated Home Sr Member
    Join Date
    Dec 2018
    Posts
    52

    Default EvoHome - Optimizing

    This is more of a curiosity than exactly an 'issue'.

    What is the simplest / quickest way for me to start monitoring temperature / boiler usage? I don't particularly want to spend days messing with Python and Rasberry Pi's. I have a constantly on iMac and a MacBook I can hijack. I don't care about output formats and nice graphs or live updates, simple CSV would do the job perfectly well. By profession I'm a software guy, nearly all C++ on PCs, but beyond the stage of wanting to take work home with me. I just want something that works (I could even probably monitor off-site from one of my work PCs ).

    The 'non-issue' I'm having is that it seems no matter what I do to optimise my schedules, I use exactly the same amount of gas every day. The meter goes on 15 units. Has done since I started tinkering at the end of December. In fact, thinking about it, it has done since first using the system at the beginning of December when we didn't even have any HR92s fitted.

    I've reduced some room temperatures / time ranges. I've reduced hot water heating schedules and the tank temperature (no, not unsafely). No matter what I do, when I go to the meter the next day in the hope of having maybe reduced usage... it laughs in my face and says 15 more gone.

    This might be great, as we moved into the house and replaced the heating system from an ancient stinky oil boiler so have no idea what it would have cost to heat with a 'normal' system but I would love to know why, no matter what adjustments I make, the gas requirement is remaining constant.

    The system consists of:
    30kW Worcester CDi Classic
    CenterStore 210l unvented tank
    S-Plan with 2xBDR91 for HW and CH zones
    16 HR92s in 9 zones
    4 hallway radiators with standard TRVs set quite low to keep common areas well balanced
    1 towel radiator, no TRV.

    I don't use the 'optimise' feature of the EvoHome because most of the time I don't care if it gets a room up to temperature precisely at a set time. If I have a set point of 20 degrees at 6pm and the room was already quite warm, it might make it to 20 at 6:10, on another day it may make 20 at 6:20. This doesn't bother me. Is there any other benefit to optimize? Seems a lot of work to re-jig the start times of all my schedules for all my zones for not much difference

    I figured monitoring is probably the only way to understand it but I'm open to any other tips of course!

  2. #2
    Automated Home Legend paulockenden's Avatar
    Join Date
    Apr 2015
    Location
    South Coast
    Posts
    1,601

    Default

    I know it has graphs and stuff which you don't need, but the simplest option is probably to rin Domoticz on one of your always-on Macs.

    http://www.domoticz.com/downloads/

    P.

  3. #3
    Automated Home Sr Member
    Join Date
    Dec 2018
    Posts
    52

    Default

    Brilliant! Thanks!

    I must admit I'd found it's name in a HUGE thread on here spanning a few years talking about monitoring software but after the first xx pages I fell asleep as all I wanted was a nice simple solution. It looked like the EvoHome support wasn't built in to Domoticz at that stage but I see it now is.

    Got it monitoring away and can immediately see my daytime set points are probably too high causing the boiler to fire up. We have an erratic household so it's hard to make a compromise schedule but a few days logging will help me work out what's right.

    (Shhh don't tell anyone, it's running here on my office PC for now .... )

  4. #4
    Automated Home Legend
    Join Date
    Sep 2014
    Location
    Scotland
    Posts
    1,841

    Default

    Quote Originally Posted by Jemster View Post
    What is the simplest / quickest way for me to start monitoring temperature / boiler usage? I don't particularly want to spend days messing with Python and Rasberry Pi's. I have a constantly on iMac and a MacBook I can hijack. I don't care about output formats and nice graphs or live updates, simple CSV would do the job perfectly well. By profession I'm a software guy, nearly all C++ on PCs, but beyond the stage of wanting to take work home with me. I just want something that works (I could even probably monitor off-site from one of my work PCs ).
    Another option if you don't want to be running anything locally is conradconnect.de, which can link to your honeywell account and produce graphs for you.
    The 'non-issue' I'm having is that it seems no matter what I do to optimise my schedules, I use exactly the same amount of gas every day. The meter goes on 15 units. Has done since I started tinkering at the end of December. In fact, thinking about it, it has done since first using the system at the beginning of December when we didn't even have any HR92s fitted.

    I've reduced some room temperatures / time ranges. I've reduced hot water heating schedules and the tank temperature (no, not unsafely). No matter what I do, when I go to the meter the next day in the hope of having maybe reduced usage... it laughs in my face and says 15 more gone.

    This might be great, as we moved into the house and replaced the heating system from an ancient stinky oil boiler so have no idea what it would have cost to heat with a 'normal' system but I would love to know why, no matter what adjustments I make, the gas requirement is remaining constant.
    "15 units" ? Is your meter imperial or metric ?

    If it's imperial that would be 474kWh (ouch!!) or if its a newer metric meter that would be 167kWh. For what it's worth, our worst day in the last couple of weeks is around 150kWh, but averaging quite a bit lower than that. Old draughty not so well insulated house with antiquated boiler!
    I don't use the 'optimise' feature of the EvoHome because most of the time I don't care if it gets a room up to temperature precisely at a set time. If I have a set point of 20 degrees at 6pm and the room was already quite warm, it might make it to 20 at 6:10, on another day it may make 20 at 6:20. This doesn't bother me. Is there any other benefit to optimize? Seems a lot of work to re-jig the start times of all my schedules for all my zones for not much difference
    I think you misunderstand the benefit of Optimal start. The point of it is that you state to the system (by setting your schedules) when you want to inhabit the rooms at a comfortable temperature and let it figure out on a day by day basis how early zones need to come on to meet that, rather than having to second guess how long your rooms take to warm up on any given day.

    So say you get up at 6am in the morning, you would have 6am as your set point in your schedule. While you may not care if it meets that target exactly each day (it usually doesn't anyway, especially in changeable weather) and think it's a hassle to now reprogram your existing schedules that were not set with optimal start in mind, in the long term it's a convenience because you just have to say when you want the rooms to be comfy and let the controller adjust the start point as the weather changes.

    It uses both the temperature difference between current room temperature and next set point, and also the previously learnt warm up rate of the room to estimate how early a radiator has to come on to hit the target. This lead time is calculated separately for each zone on the fly.

    The system will adapt immediately to day to day differences in starting temperature. For example If a room is to reach 20 degrees and started at 14 degrees yesterday and needed a 2 hour warm up, but today started at 17 degrees it will only come on an hour earlier. (Roughly 3 degrees per hour in this example) This difference in starting temperature may be a result of the room being heated until later in the night than the previous day for example.

    The system will also adapt to different warm up rates (different flow temperatures, different outside conditions) as well but that takes longer - it can take several days to a week or so to adapt to a sudden cold snap, or a change such as a radiator being replaced with a different one that results in a faster or slower warm up rate.

    It can be remarkably accurate with some rooms, less so with other rooms. (It can be a long way off for rooms that struggle to reach their set points)

    The point is that if you set your start times manually based on how long you think it will take to heat up, any time the weather gets a bit warmer it's running for longer than it needs to be, and you need to reprogram your schedules as the seasons change as well. Some of our rooms take up to 3 hours to get up to temperature in the middle of winter (solid brick internal walls so loads of thermal mass to heat up, and some slightly undersized radiators) but maybe only an hour in cool autumn/spring weather.

    Think of optimal start not as turning zones on "early" but as an "opportunistic delay" system. What I mean is that if you want the house at 20C at 6pm for getting home, with a manual system you'd have to guess that it should maybe come on at 4pm in cold weather. But then in warmer weather it still comes on at 4pm even though it doesn't need to and you may not even be aware of that if you are not at home and not graphing your temperatures.

    With optimal start you are saying "20C at 6pm please, go to it". If it thinks it needs to come on at 4pm to manage that in a particular zone it will do that. When the weather warms up it will automatically, progressively delay the start times towards the set point schedule so it might come on as little as 30 minutes before. Nobody is manually going to adjust their schedules with this level of granularity on a daily basis.

    It's also good in situations where different rooms heat at different rates but you might want them all "ready" at the same time, say 6pm in the evening. If you are working this out manually are you really going to say "Ok, living room at 4pm, kitchen at 4:30pm, bathroom at 5pm, that should get them all warmed up for 6pm". You're more likely to schedule all these rooms with different warmup rates to come on at the same time say 4pm, which means some rooms will be up to temperature far too early and running longer than they need to. (Does the quickly heating bathroom really need to be up to temperature an hour early just because the living room takes longer ?) This applies to traditional single timer for the whole house systems as well of course where inevitably some rooms come on long before they need to.

    Optimal start automatically accounts for the differences between rooms such that even if the schedule set points for the zones are all the same, they all come on staggered at different times as needed based on their individual temperature differences they need to make up and their individual warm up rates. This surely must save some gas compared to the quickest room coming on at the same time as the slowest room...?

    So I think there is a pretty good case for Optimal start. It generally works well but under certain circumstances it can have slightly pathological behaviour, so it's useful to keep an eye on it to see that it is working sensibly. In particular it seems it will only adapt so far in relation to room warm up rate.

    It appears to use a warmup rate of 3C per hour as it's initial estimate in an unfamiliar zone, so a room that warms up slower will initially fail to meet its target in time and a room that warms up faster than this will initially reach it's target too soon, but it will adapt over a few days by starting later or earlier by about 15 minutes per day until it hit's its target. As conditions change it will keep making these small daily adjustments to try to hit the target.

    However it doesn't have infinite adaptability. In my observation any room that heats faster than 6C per hour will always be ready too soon, and any room that heats slower than about 1C per hour will always fail to reach the target in time - these seem to be its limits of adaption which it won't go beyond.

    Optimal stop is a completely different thing and without going into detail in this post, I haven't found it particularly useful so I leave that off.
    Last edited by DBMandrake; 16th January 2019 at 04:31 PM.

  5. #5
    Automated Home Ninja
    Join Date
    Mar 2017
    Posts
    298

    Default

    Have to say that usage seems rather high, if DBMandrake is right in his calculations that's roughly 180 a month in Gas. Probably unlikely, but have you considered the fact you might have a gas leak?

    Also you've not really commented how big the house is, or how well insulated it is.

    We have a well insulated 4 bed detached house (Built 2000), 170L unvented cylinder, and the original 18kw boiler (Non condensing). Our average usage per day for two people last January was 81kwh per day. I'm currently trying an experiment with leaving the heating on all day rather than our usual Morning/Evening times and we're currently averaging out at 79.5kwh per day. However the issue with this is the electricity usage will be increased due to the 5 minute pump overrun every-time there is a demand from any zone during the day (Which you wouldn't usually have as the system would be set back).

    I'd say our usage is pretty high for two people as we like it warm, our showers consume a lot of hot water (11.5l Min hot) and we also use gas for cooking too. Temps we usually set at around 21oc (Couple of zones boosted to 22.5 in the evening) with a set back to 16oc at night and daytimes.

    Unlike DBMandrake I do use Optimum Stop along with Optimum Start (Set to Max 1 hour) which I find both work very well.

    Is it possible you've got a HR92 that's in a cold spot or draft that keeps calling for heat when not really needed?

    You've also got to consider that your Worcester Bosch CDI classic only has a 1:4 modulation range, so it'll only modulate down to 7.7kw. The issue with this being is you're only getting one zone calling for heat, i'll have to output a minimum of 7.7kw into the system every-time there's a demand. This can be an issue if your boiler has been oversized for the property or situation, especially with Evohome when it's possible you'll only have a couple of zones calling at once when they have reached desired temperatures.
    Last edited by mtmcgavock; 16th January 2019 at 09:33 PM.

  6. #6
    Automated Home Guru
    Join Date
    Feb 2016
    Posts
    229

    Default

    I must admit I'd found it's name in a HUGE thread on here spanning a few years talking about monitoring software but after the first xx pages I fell asleep as all I wanted was a nice simple solution. It looked like the EvoHome support wasn't built in to Domoticz at that stage but I see it now is.

    Got it monitoring away and can immediately see my daytime set points are probably too high causing the boiler to fire up. We have an erratic household so it's hard to make a compromise schedule but a few days logging will help me work out what's right.
    Yes, that Domoticz thread definitely goes into far more detail than most users need and Gordon's recent update to the Domoticz Evohome Web API implementation makes set-up really easy - just install Domoticz, add your Honeywell account details to the Evohome Web API hardware device and then all your Honeywell zones should appear automatically. I've updated the Evohome Domoticz wiki to put this at the top as I would recommend this set-up as the starting point for all new users.

    https://www.domoticz.com/wiki/Evohome

    I recently helped a friend get Domoticz set-up following their new Evohome install and apart from an issue with their Honeywell account, it was pretty straightforward.

    Dan

  7. #7
    Automated Home Sr Member
    Join Date
    Mar 2017
    Posts
    85

    Default

    see if you can use your phone to take some hourly or something photos of your meter.. that way you get some timestamped readings. make sure that between at least 2 of your readings you have your heating off, then stick it all in a spreadsheet, make a few graphs (easier to visualise things) and try and match usage to heating...

  8. #8
    Automated Home Sr Member
    Join Date
    Dec 2018
    Posts
    52

    Default

    Darn it, notifications for this thread dried up, I see there's a few posts, not really sure where to start so in answer to a few people...

    Quote Originally Posted by mtmcgavock View Post
    Have to say that usage seems rather high, if DBMandrake is right in his calculations that's roughly 180 a month in Gas. Probably unlikely, but have you considered the fact you might have a gas leak?

    Also you've not really commented how big the house is, or how well insulated it is.
    Ok, so to start with the house. It's about 100 years old, 4 bed, 3+ reception, 2 bathroom + downstairs wc. Detached, badly insulated, solid brick walls for the most part and a nice draughty basement. We knew it was never going to be the easiest to heat when we moved in last September. The installed heating system was probably Edwardian too! Haha! It was an old 19kW oil burner probably from the 60s-70s. It would take 2 hours to heat up the radiators so it had to go. My gas supplier said they could estimate me a monthly bill based on number of radiators (20) and number of people in the house (4) and said 195 I said no thanks.

    Quote Originally Posted by mtmcgavock View Post
    You've also got to consider that your Worcester Bosch CDI classic only has a 1:4 modulation range, so it'll only modulate down to 7.7kw. The issue with this being is you're only getting one zone calling for heat, i'll have to output a minimum of 7.7kw into the system every-time there's a demand. This can be an issue if your boiler has been oversized for the property or situation, especially with Evohome when it's possible you'll only have a couple of zones calling at once when they have reached desired temperatures.
    In our last house (4 bed semi, similar age but cavity walls) we had upgraded to a Worcester 24i combi boiler and were happy with it so decided to go for another Worcester combi ... that plan fell through when our plumber got cold feet about the pressure of a combi on the older pipework so we went for a standard Worcester 30 and an unvented tank to get the higher pressure for showers / sinks while keeping the heating as a standard vented system (20 radiators). More expense but less risk as we're trying not to disturb too much of the original fabric of the house, and ultimately a better system than a combi in a house of this size.

    We initially weren't thinking of adding any form of smart controller but then figured if we fitted an EvoHome controller it would work as a normal thermostat for the time being for only a small additional outlay. At a later date we could add HR92s if we felt like it. The more I read, the more I thought it was a great idea for a house this size so I ended up buying 16 HR92s. Our fitter said he was Honeywell approved and had an EvoHome system himself. He has a good reputation as a plumber but as the job went on it became apparent his own EvoHome install was just a controller and he hadn't fitted others, so I guess that's why he spec'd a boiler that maybe wasn't optimal (along with a tank that didn't have a pocket for the temp. sensor, another thread, but that's solved now).

    We're a family of 4-5 depending on time of year and probably have between us an average of 3 showers a day. People are also not out all of the day so I can't just shut it off and warm it up for 6pm unfortunately.

    Quote Originally Posted by DBMandrake View Post
    Another option if you don't want to be running anything locally is conradconnect.de, which can link to your honeywell account and produce graphs for you.

    "15 units" ? Is your meter imperial or metric ?

    If it's imperial that would be 474kWh (ouch!!) or if its a newer metric meter that would be 167kWh. For what it's worth, our worst day in the last couple of weeks is around 150kWh, but averaging quite a bit lower than that. Old draughty not so well insulated house with antiquated boiler!

    I think you misunderstand the benefit of Optimal start. The point of it is that you state to the system (by setting your schedules) when you want to inhabit the rooms at a comfortable temperature and let it figure out on a day by day basis how early zones need to come on to meet that, rather than having to second guess how long your rooms take to warm up on any given day.
    Yes the usage is high. Too high (it is metric so I'm glad to say it's your lower figure per day at the moment), but on the other hand, when I looked back at my bills for last year, the old house was using about 100kwh per day average in the winter quarter so half-again isn't ridiculous.

    I figured that if a room gets up to temperature 30 minutes early, surely the HR92s sense this and close off the radiator to just a trickle, so it's not like the heating is going to be full-on for that half an hour extra. I didn't want to introduce the extra variability of the optimal start until I got the usage reduced and the system set right (takes a while to tweak at 9 zones to a point where we're all happy).

    Got to say I also felt a little bit out-of-control until I got Domoticz up and running a couple of nights ago. Now I can see the temperature history, even if not the actual boiler demand. I turned on optimal start last night. It will be interesting to see how much it alters the usage (haha I'm guessing not at all, still be 15 units ) but EvoHome turning on the bedroom heating at 4:45 for a 6:30 wake-up this morning was a little excessive. A bit of adjustment is definitely required as that's 1hr 15m more than I'd previously been using to get a little heat to get up to.

    Of course, in my constant battle, the weather keeps getting colder. Makes it hard to tell if I'm making a difference by tweaking or not. So far, from Domoticz, I've identified the overnight 15 degrees wasn't low enough to stop the heating kicking in at 3am for some rooms. I've corrected that one. Similarly in the late mornings it was also running pretty much pointlessly. I can also see one bedroom that's unable to reach its set-point, new radiator required and some insulation I think. Hopefully in the next week I'll get the usage down.

    It's amazing such data isn't readily available direct from the controller. I don't know how anybody could possibly work these things out in any other way on anything other than the simplest installation. Many thanks to all who've helped me out here!

  9. #9
    Automated Home Sr Member
    Join Date
    Dec 2018
    Posts
    52

    Default

    Quote Originally Posted by Stevedh View Post
    see if you can use your phone to take some hourly or something photos of your meter.. that way you get some timestamped readings. make sure that between at least 2 of your readings you have your heating off, then stick it all in a spreadsheet, make a few graphs (easier to visualise things) and try and match usage to heating...
    Or something like this clipped to the meter would be handy.

    https://www.loopenergysaver.com/

    Wonder if that interfaces to Domoticz.

    Darn it. To get the gas monitor I need to buy the electric monitor, but I don't want to monitor electric as (a) I already have an Owl Energy Monitor and (b) so far my electric usage is pretty much exactly what I budgeted for.

    Oh well. Maybe there's other similar devices...
    Last edited by Jemster; 18th January 2019 at 10:15 AM. Reason: Found the Gas Only version! :) Oh no I didn't.

  10. #10
    Automated Home Legend
    Join Date
    May 2014
    Posts
    988

    Default

    In my experience, it is worth getting a simple gas usage monitor. I managed to get a NorthQ gas starter kit off EBay last year for 25. In the days when I had Evohome it showed that there was merit in setting Evohome to run 16/8 (or 24 hours with lower set temps overnight). As I have posted before, the issue for me was Evohome's fuzzy logic which demands a high boiler flow temperature when one or more zones are outwith their set temperature range. My 15 minute usage graphing gave me the confidence/knowledge just to set and forget. That said, I appreciate that no two homes are the same in terms of usage and lifestyle.
    Last edited by HenGus; 18th January 2019 at 10:27 AM.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •