Latest Vaillant config files?

[chrisbev]

Hi newbie hope this is OK to post here.
I'm looking for latest config files that includes SetModeOverride as 2.1.x does not have that. I've searched and found a line to add to the bai.308523.inc file but wondered if someone has this (or more) included or working on this please? Just started, Vaillant 625 ecoTEC plus 001003612 boiler, bit of a learning curve at the moment.

[JonesGW-MGD]

2.1.x are the latest 'official' repository files and as you say, don't include the custom setmodeoverride message definition. To use it therefore, you'll need to gitclone / copy the configuration files to your local instance, where you can edit bai308523 (if that is indeed the boiler file your ebusd instance loads at start) and add the line definition if that makes sense?

[JonesGW-MGD]

I feel your pain!

(Remember, I'm not a heating engineer so take that into account!). I can share some thoughts about what I might do with ebusd in your position and why I think some of the advice you've been given is plain wrong but I'm still not sure what I'd try to fix first - as this veers away from ebusd and more about the behaviour of your specific system, I really recommend you post identical queries about these issues on plumbersforums.net and diynot.com - both have esteemed experts and amateurs alike interested in helping with such issues - and with a lot of Vaillant knowledge too.

First, in terms of using ebusd to implement a control strategy, I would try to mimic PDHW (like 'X Plan' or indeed a Combi). https://heatingacademynorthampton.co.uk/pdhw/ So, basically a call for DHW shuts off Heat. It's highly unlikely you'll notice unless it's -10 outside! That way you'll get the lowest possible return temp, and the highest possible condensing.

(Obviously I don't know your system details, pipe runs, emitter sizing etc. so the flow temp numbers won't necessary be anything more than an example)

For example:

Call for DHW - Set flow temp to 60-65, disable UFH and Rad ZVs
Call for UFH only - Set flow temp to 40-45 (or whatever your UFH heating input should be - no need to waste heat by blending)
Call for Rads only - Set flow temp to 50-55 (assuming you can get away with this - otherwise higher of course)
Call for Rads and UFH - Set flow temp to 60

You could ultimately get smarter - Call for DHW, set flow to 55-60 and then run a anti-legionnaire's cycle at 65 every 24 hours!

(One thing to consider and not ebusd related at all, but if your cylinder doesn't have a restrictor valve (i.e. gate) on the return, might be worth fitting. It's not common practice, but it generally means that most cylinder setups in the UK don't exchange as much heat as they could as the DHW flow goes through the coil too quickly (unless you have a 'high recovery' coil) - consequently it goes to return at too high a temp to be optimal. It is possible that a modulating pump might help with this though - a bit outside my knowledge base)

(To help with the monitoring/tweaking of this what you might find it helpful to do is build a couple of flow/return sensors - sounds like you're capable! I have a couple of sets - ESP32s (running in my case an Arduino sketch) with two DS18b20 probes per device, wifi connected, publishing to HA via MQTT - basically 10 quid each device!). (and yes, I have one of mine on the flow and return near the boiler and reads quite different from D.040 and D.041.)

Now some stuff I don't understand -

You shouldn't need to range rate a system boiler (unless it wasn't sized properly and you really need a 610) with a cylinder (D.000) so no reason why it shouldn't be 'Auto' - range rating is for Combis which may need very high outputs for DHW but less for CH. E.g: My 35KwH combi 'only' needs a max of 16KwH (2.5 times UK average!) to heat the house to design temp (21 deg at -2 outside) - even so, I have mine set to Auto and seems fine.

But even stranger - how can you range rate a 20Kw boiler for 25.5Kw? Have I missed something?!

Even if you range-rate it for 15, that's the max CH load - it's still going to use the same algorithms and pump controls to attempt to optimise modulation, flow temps I think?

Regarding the behaviour you observe re. 'modulation memory' - that's potentially above my technical paygrade - the problem is there are so many setting that inter-relate! Just to check - you have D.017 set to 'flow' I presume? In 'normal operation' if you monitor D.029 (flow rate l/s) - this is what ebusd/HA is great for btw! does it seem 'sensible'? Is D.071 (Max flow rate) and D.128 (min flow rate) still factory defaults?

Also my understanding re modulation differs from your comment: ... the flow temp (i.e. driven by low modulation) stayed low. I thought the boiler would fire high, meet set point and then modulate down to maintain set point, not the other way around?

(This is your installation manual in case you don't have it: https://professional.vaillant.co.uk/downloads/aproducts/new-ecotec-plus/firstspirit-16874195715150020308118-05-2681222.pdf - codes at the end)

On pumps -

In general my basic understanding is you want pumps to work as little and as low as possible while maintaining your design criteria - i.e. keep you warm!

So - on your TMV/Pump set - my u/s is that the pump is 'sucking' the water around the UFH loops as the boiler pump would be unable to 'push' it around. (Page 52 of the manual does suggest the boiler has some capability to satisfy UFH without a secondary pump but might be too limited - no idea how much pipe/area you are covering). Consequently it's after the TMV/blender - I'd still set this as low as possible so that you get the heat transference and a low return temp but I wouldn't have thought it would significantly interfere with the boiler pump - is the manifold very close to the boiler?

Setting D.170 to 'fixed' makes absolutely no sense to me. You've got a sophisticated boiler with a modern pump - and basically the advice has been to cripple it like a pump from the 1970s? The only scenario where I could see setting the pump to fixed is if you have hydraulic separation - either a buffer tank or low loss header for example. In this scenario you're simply interested in the Boiler>LLH>Boiler loop being constant and the off board pump(s) then handle the heating circuits which are hydraulically separated. (and your manual more or less confirms that - page 50)

And 80% seems high - my pump is on auto (I know my set up is different) and this is the last 7 days pump % and modulation:

I have a 'lot of water' - 7 out of 13 are high capacity radiators (over-sized Zehdner Charlestons), and miles of MLCP and copper - probably 250-270 litres at least.

If you monitor d.015 is it actually operating at 80% when it's on?

I'd be inclined to check the manual and try setting the pump to default - do you know what it was: 1?
I'd also be inclined to set burner anti-cycle and pump overrun back to defaults too (D.002 & D.001)

And then maybe make small adjustments after having ebusd/HA monitor some of this stuff for a while?

(just to answer your weather comp query - the boiler uses a heat curve to take into account outside temp with desired (room) temps - essentially it determines the DesiredFlowTemp - which is why with weather comp you don't want to be manually changing this). The practical implication of this is the flow temp is lower when the outside temp is higher and therefore the resulting return temp and condensing benefits are better. A lower heatcurve is going to mean a lower flow temp and the heating taking longer to reach desired room temp of course. If I had two or more heating zones, each could have their own heat curves - again I suspect you'd really need to use Vaillant controls to make this work, but as you say you can certainly emulate weather comp by implementing a simplified curve using ebusd - the typical Vaillant curves are on page 52 of your installation guide. I run my heatcurve at about 1.6 (and tweak it if it is going to be arctic to 2) - the Vaillant TechSupport told my installer to set this to between 2.5 and 3 as my old house couldn't cope with it being lower. I hate to say it but I think the standard of heating and hydraulics in the heating industry in the UK is worse than poor - we've had decades of installers smashing in massively over-sized boilers because 'they work' - The heat loss in the 'average' house in the UK is between 6-7Kw - when have you ever seen a boiler that small? Modulation of course helps make it even easier for installers without understanding it!).

Hope some of this helps!

[JonesGW-MGD]

While I'm thinking.... Your internal expansion vessel is 'only' 10l I think - rule of thumb is 10% so might be worth checking your total capacity esp if you have extensive UFH. It's probably not causing any of the issues you're experiencing but since you're going to the trouble of optimising your heating system, no harm in checking? (You can work on 16x2mm UFH being 0.12l / m, 15mm copper 0.14 l/m and 22mm 0.31 l/m, plus what's in the boiler, and the rads)

I also suspect your Hives may not helping. From what I recall, the Hive devices have little or no hysteresis curve - so you set the temp for 20 and boiler calls for heat, the room reaches 20 (or likely overshoots a bit with a fast running pump) but when it falls to 19.8 or 19.5 (I've seen various reports of 0.2-0.5, it kicks the boiler back in. I think this is less of a problem with a room stat controlled environment but where you have Hive TRVs on Hive's Heating on Demand setup which allows the TRVs to essentially call for heat, then it might be making it more sensitive to cycling - with the temp from the TRV near the metal mass cooling down quicker than the room air near a roomstat and more devices that can effectively call for heat . If individual TRVs make a call for heat often, and for short periods, the boiler may not get a chance to modulate before it satisfies the desired temp anyway but this is much more conjecture than the earlier response!

[chrisbev]

Good thought about expansion vessel will do some sums.
Nice bit on your system thanks some questions to follow and answers to yours, in the mean time.....
Yes the hive trvs and HoD! I've plotted out the room temperatures in ha as the hive app only does the thermostat. My first mod cardboard/foil between the trv and rad helped but a mixture of overshoots or not as the trvs learnt after a few days? Some good moved to sensitive positions others in don't care position! Web says the hive trvs have 3 temp sensors, I presume from that one water and 2 air to predict the tempersture gradient and get to in room temperature, a tricky job for the algorithm? This is what I have so live with it but thinking of adding a room sensor into ha and ramp the trv set point to satisfy the room temperature required? Yes to me hysteresis very small but annoyingly the HoD works 0.5C below the set point (hive say to stop excessive cycling but to me just moves it to a different point?), imagine evening watching TV all good room temperature set to be comfortable. Schedule change to get bedroom ready for bed time now bedroom HoD calls for boiler, the lounge trvs now get hot water and control to the set point which is 0.5C above what it was! To me I can notice that and not satisfactory. What I did was at bedroom schedule time lowered the lounge set points by 0.5C, not 100% if another room calls early but works most of the time! I have ha automation instead of hive HoD so can modify the point at which the boiler is called etc. work in progress but delayed by new boiler. This is another reason to get temperature compensation working so trvs have less to do.

[chrisbev]

From earlier. I have flow and return pipe temperatures into a raspberry pico so have plots and then wifi to ha, isn't ha good for yhat? I had wondered about the dhw return temperature being quite high and just thought an old cylinder with scale? Fitted a water softener as very hard water here and no noticeably difference so your restricter could be worth a consideration on dhw, the boiler is quite low modulation so I thought higher flow the better but that presumably is false thinking?
It is a 625 (not a 620) and the setting says 25.5 the reason for this was vaillant's way of overcoming some shortfall in the boiler algorithm (I believe it is a new algorithm for this updated model but same name ecoTEC Plus?).
For problem 1 the booler modulation was ~20% with flow temp 40C and didn’t drive to set point setting D.000 to 25.5 rather than auto fixed that but it still modulates to get up to/down to set point just doesn’t stick down there (I assumed remembering the last load conditions) maybe after hours/days it might 'recover'? I think this is where my 625 differs from previous 428 that had the reverse problem it fired full blast and with low load ran into high temp trip before it modulated low. Setting D.000 lower had no effect I was told it had to fire high for a certain time to ensure good ignition, the new one has IoniDetect for mixture control for different modulations so can keep good combustion to a lower level to get the 10:1 range and presumably start low and then increase (or supposed to). All be it they haven’t got it quite right for my system hence the fudge?
D.001/002 I have not changed and at defaults 5 & 20.
D.071 for me is max target flow temperature not flow rate and at defaults 75C.
D.128 min flow temp not rate and at 30C (default 40C but not changed by me).
With this ufh blender the hot port is to boiler so if satisfied (never quite gets there?) would shut to the boiler so has to have a separate pump? Common port goes off to the ufh pump and floor coils. Cold port tee'd into coils back to boiler pipe. To use the boiler pump I expect boiler temp is controlled and no blender?
Problem 2, fixing the boiler pump stage to me is also a hack to get around the boiler algorithm not being suitable for my sysem or could say it the other way round but I did ask the installer (and later vaillant) if the boiler was suitable for the zones dhw, rads and ufh?
With the 2 hacks now seems to start and get going on the first try. Am I a happy Customer? Yes and No I suppose?

[JonesGW-MGD]

sorry my bad - not sure why I read 620 instead of 625 and yes d.071 and d.128 - I meant temps not rate my bad again.

If setting d.000 works for you - then hey - stick with it! These devices are 'black boxes' and we'll never know what they really do so it's very much trial and error - my setup is working well but one of the reasons I ventured into ebusd/HA is I think (ironically since I wsa determined to have a really low flow, return and dt!) my delta T is too low (might benefit from an external bypass?) and my D.029 flow rate (confusingly displayed on screen in litres and different metric in HA) seems too high so always learning!

I think your (Vaillant's!) pump hack is more suspect - of course running the pump high and fast will work - like they did decades ago - just not sure it's the best your setup deserves! You could start experimenting here though with data feedback from HA.

I think your setup is obviously capable - just not sure it (/Vaillant) 'plays well with others'. (- annoyingly they don't comply with OpenTherm - except in the NL where they legally have to - grrrr!! Try buying a Vaillant VR33 OpenTherm module in the UK....) - but to fix that you'd need to replace your controls with a VR71 ebus wiring centre, a VRC720f or similar controller and if you wanted see the results remotely - a VR920/1 internet gateway- basically another 700 quid plus install! Oh, and another 100 quid per ambiSENSE TRV!....You seem to be fairly close to where you need to be - at least with HA and ebusd you can collect a sh*t-tonne of data which is what I like!

[chrisbev]

Thanks again just a quick question on your graph if I may then just as I have the ha data of my boiler in ha all-be-it some entries repeated a few times and some entries with the same name have different values to sort and about to try receiving and sending single lines ready to add to automations I have to take a break from ha for a week or so but considering the green flow, yellow return and blue required/set point lines. I see green and yellow close together for a lot of time so assume boiler is off or not heating, fair enough but why is the blue line above and not seem to tie up for most of the time? Also there is one point 04/10 01:00 ish where return is above flow and the required dips? Middle of night? There is at 5:00 ehat looks like it could be the required rising and flow climbing but why so slow for flow to get to required temperature at 8:00? Just asking to try and understand the logic not a criticism.

[JonesGW-MGD]

To. answer your question, I'm not fully sure of the behaviour of my system! From an 'input' perspective, each TRV has it's own programme but broadly - heating is off overnight (23:00-07:00 - setback to 15 or so), downstairs (and some upstairs) is warmed 08:00-10:00 (20-21) and then sets back to 15-16 or so and then back on to 20-ish for the evening.

This might be a better graph - longer period, colder period (end of Feb, early March)

I don't understand why the 'DesiredTenp' 'oscillates' at night (could be the frequency of plot is off-putting) - obviously as the outside temp falls, the target flow is going to shift anticipating satisfying setback temp?.
I'm not too surprised the flow lags the desired - I read that as a function of the low-ish heat curve but sometimes it leads according the chart?
I've always been concerned the flow and return DT is really low. Low is generally good particularly since my objective is to run a low temperature heating system but the usual recommendation is DT = 0.3 x Flow, it implies the heat isn't given a chance to 'emit' but the rads definitely warm the rooms.

I have zigbee temp gauges in most rooms but haven't yet overlaid that data.

The Vaillant setup claims to be 'adaptive' and the TRVs are certainly very 'chatty' with the controller, making audible regular adjustments - but it's hard to work out what is 'learning' algorithms and what is unwanted behaviour

TBH generally happy with the system - was such a 'low bar' coming from electric. Have also had to resolve some issues left by installer - like air noise since they didn't 'believe in airvents'... Just need to find time to leverage the ebusd data to optimise further!