- Joined
- 16 May 2021
- Messages
- 5
- Reaction score
- 1
- Country
I’ve just had a full central heating setup installed into my ‘older’ property. Mostly, everything feels fine (and warm). The one adverse symptom is that one (high capacity) radiator in the living room occasionally but regularly ‘tinkles’. I’ve repeatedly bled the system but this doesn’t alter the noise. There are some aspects that concerned me about the install and the installers and some other things I thought were really good about their work but I have some nagging doubts the install was a bit of an experiment for them.
I have wondered whether the water velocity is too high (which was a design-time concern I had) but some boiler data seems to be inconsistent with that theory.
But first some context:
The property is a two-floor 4 bed thatched house, 162m2 / 1744 sq Feet with a c17 part, a mid c20 extension and a later c20 extension - It’s essentially ‘1 room wide’ so long and thin-ish. Mostly single glazed with perspex secondary.
Tired of finger-in-the-air quotes by various heating engineers who all came up with wildly different quotes, I did my own room-by-room whole house heat loss and I wasn’t surprised the result was about 2.5 times the average house at around 14kW. (The heating industry is screwed in this respect -doing a decent heat loss calc is easily a day’s work for an experienced installer and you can’t do that on every quote - I offered a couple of times to pay for the heat loss first before the quote but twice I was told ‘trust me, I’ve been doing this xx years….’)
Previously heated by electric night storage heaters, this very quickly became non-sustainable in 2023!
The resulting install (Aug 23) is:
A LPG-fuelled Vaillant 835 Ecotec Exclusive Green iQ (VUW 356/5-7) Combi (The Combi was a little controversial since it replaced a 210L direct unvented but the installer convinced me since we only have one bathroom. I am mostly ‘OK’ with the outcome in that respect).
The boiler is installed in a box room on the first floor. The circuit design is 22mm copper a metre up to the loft and then (pre-insulated) Maincor 25 MLCP flow and return running in the loft to drop into a bedroom cupboard at the other end of the landing from the box room and feed a Maincor manifold 13m away. (There is a tee-ed circuit feeding 3 rads that comes off the boiler for two upstairs and one downstairs rad - initially 22mm then 15mm to the second upstairs rad, and 16mm MLCP to the third (which is a high capacity column rad)). Off the manifold, 4 circuits feed two or three radiators each. (Each manifold port has at least one upstairs and one downstairs rad - there is no separate zoning due to the property layout. (There are no temp controls on the manifold).
The downstairs rads are all high water capacity Zehnder Charleston columns. Upstairs are more conventional K-rad K22 or K21s.
I worked with the installer on the specification - any issue with the heat loss assumptions or rad selection is mine - his original spec was much lower than mine but I was keen to see how far I could push the boundary and run as low flow and return temps as possible. So I generally ‘oversized’ rads for even my heat loss calcs and they worked out about double his.
From the manifold, they ran 16mm MLCP. I wasn’t very happy with this when they said this was their plan, given that some of the manifold ports have a couple of downstairs rads each that have a combined output at my desired MWT-AT of over 4kW and water capacity of 60l but he assured me that Maincor had vouched for the heat capacity of the 16mm (Although I’m not sure at what dt) -even though it has a smaller ID than 15mm copper, 16mm MLCP does seem to have a better mass flow rate than 15mm copper (but not 22mm).
Visible pipework to rads is 15mm copper.
The controls on the system are: VR 920 (Sensonet) gateway, VRC 720rf (Sensocomfort), (and the outside weather comp unit that comes with it) VR50 (Ambisense) TRVs on all rads and a VR52 repeater. The choice/level of tech was mine. (Being in IT, I was bracing myself for the Sensoapp being as bad as everybody says it is but it isn’t the worse app I’ve used!)
The Heat curve is currently set at 1.75 - I don’t experience an issue with the time it takes to heat up (Although it hasn’t been bitterly cold just yet). D.000 (Partial load) is set to Auto - so assume should be modulating according to the in-built ‘smarts’.
The system has been in a couple of months and used in earnest this last few weeks. Aside from the tinkle noise, there are a few things that I think are ‘odd’ that I wanted some advice on:
From the diagnostic settings on the boiler:
The flow and return delta is really low every time I look (D.040 and D.041) - like 5 degrees - while I’m happy in general with what I hope has ended up being a low return temp and dt solution - it seems suspicious - might this indicate a too high velocity with the resultant heat not getting a chance to dissipate in the emitter network?
But D.029 (Water circulation volume) always seem to be under 1000l/h - which by my maths works out at 0.3m/s. And the pump is set to ‘auto’ (D.014) and I haven’t seen this above 15% (D.015 - Actual)
So to recap my queries:
What generally is the explanation for a ‘tinkle’ sound (!)?
Do my stats look odd in anyway?
Any other pointers?
(The couple of things that troubled me about the installers were - he didn't install an aav at the high point of the system (the loft) - he said they go wrong and that makes them a problem especially in the loft (I would have put a lever valve to isolate it if I had done it). He also said he didn’t believe in balancing radiators - ‘if you have to, you had designed the system wrong’. I think this is a fairly significant misunderstanding of what balancing is and does personally, although I do think that having smart TRVs on each rad and having the rads in ‘star configuration’ hanging off a manifold may mitigate things - but only lockshields restrict flow to a radiator surely? To be fair - I do not experience an issue with some rads heating up well before the others but currently all the lockshields are ‘wide open’. The final thing may just be a personal thing - he used some nice press fittings to the boiler and the odd half decent solder joint (limited because of thatch and his insurance) but under the floor and where not generally visible it's bucketloads of push fit - and IMHO he used way too many elbows, when a little thought with the MLCP or even swept copper would have looked neater, and reduced frictional resistance. It seemed like he didn’t know you could get MLCP press-to-Copper press fittings - I would certainly have paid any uplift but actually the Maincor press to press are cheaper than speed fit! I would say without exaggeration he used about 35-45 elbows in a property with already longish pipe runs. He also failed to insulate the copper flow and return that joined the MLCP in the loft and also the cold feed (that was previously insulated to the unvented). Maybe being picky but it made me maybe unfairly question the outcome)
I have wondered whether the water velocity is too high (which was a design-time concern I had) but some boiler data seems to be inconsistent with that theory.
But first some context:
The property is a two-floor 4 bed thatched house, 162m2 / 1744 sq Feet with a c17 part, a mid c20 extension and a later c20 extension - It’s essentially ‘1 room wide’ so long and thin-ish. Mostly single glazed with perspex secondary.
Tired of finger-in-the-air quotes by various heating engineers who all came up with wildly different quotes, I did my own room-by-room whole house heat loss and I wasn’t surprised the result was about 2.5 times the average house at around 14kW. (The heating industry is screwed in this respect -doing a decent heat loss calc is easily a day’s work for an experienced installer and you can’t do that on every quote - I offered a couple of times to pay for the heat loss first before the quote but twice I was told ‘trust me, I’ve been doing this xx years….’)
Previously heated by electric night storage heaters, this very quickly became non-sustainable in 2023!
The resulting install (Aug 23) is:
A LPG-fuelled Vaillant 835 Ecotec Exclusive Green iQ (VUW 356/5-7) Combi (The Combi was a little controversial since it replaced a 210L direct unvented but the installer convinced me since we only have one bathroom. I am mostly ‘OK’ with the outcome in that respect).
The boiler is installed in a box room on the first floor. The circuit design is 22mm copper a metre up to the loft and then (pre-insulated) Maincor 25 MLCP flow and return running in the loft to drop into a bedroom cupboard at the other end of the landing from the box room and feed a Maincor manifold 13m away. (There is a tee-ed circuit feeding 3 rads that comes off the boiler for two upstairs and one downstairs rad - initially 22mm then 15mm to the second upstairs rad, and 16mm MLCP to the third (which is a high capacity column rad)). Off the manifold, 4 circuits feed two or three radiators each. (Each manifold port has at least one upstairs and one downstairs rad - there is no separate zoning due to the property layout. (There are no temp controls on the manifold).
The downstairs rads are all high water capacity Zehnder Charleston columns. Upstairs are more conventional K-rad K22 or K21s.
I worked with the installer on the specification - any issue with the heat loss assumptions or rad selection is mine - his original spec was much lower than mine but I was keen to see how far I could push the boundary and run as low flow and return temps as possible. So I generally ‘oversized’ rads for even my heat loss calcs and they worked out about double his.
From the manifold, they ran 16mm MLCP. I wasn’t very happy with this when they said this was their plan, given that some of the manifold ports have a couple of downstairs rads each that have a combined output at my desired MWT-AT of over 4kW and water capacity of 60l but he assured me that Maincor had vouched for the heat capacity of the 16mm (Although I’m not sure at what dt) -even though it has a smaller ID than 15mm copper, 16mm MLCP does seem to have a better mass flow rate than 15mm copper (but not 22mm).
Visible pipework to rads is 15mm copper.
The controls on the system are: VR 920 (Sensonet) gateway, VRC 720rf (Sensocomfort), (and the outside weather comp unit that comes with it) VR50 (Ambisense) TRVs on all rads and a VR52 repeater. The choice/level of tech was mine. (Being in IT, I was bracing myself for the Sensoapp being as bad as everybody says it is but it isn’t the worse app I’ve used!)
The Heat curve is currently set at 1.75 - I don’t experience an issue with the time it takes to heat up (Although it hasn’t been bitterly cold just yet). D.000 (Partial load) is set to Auto - so assume should be modulating according to the in-built ‘smarts’.
The system has been in a couple of months and used in earnest this last few weeks. Aside from the tinkle noise, there are a few things that I think are ‘odd’ that I wanted some advice on:
From the diagnostic settings on the boiler:
The flow and return delta is really low every time I look (D.040 and D.041) - like 5 degrees - while I’m happy in general with what I hope has ended up being a low return temp and dt solution - it seems suspicious - might this indicate a too high velocity with the resultant heat not getting a chance to dissipate in the emitter network?
But D.029 (Water circulation volume) always seem to be under 1000l/h - which by my maths works out at 0.3m/s. And the pump is set to ‘auto’ (D.014) and I haven’t seen this above 15% (D.015 - Actual)
So to recap my queries:
What generally is the explanation for a ‘tinkle’ sound (!)?
Do my stats look odd in anyway?
Any other pointers?
(The couple of things that troubled me about the installers were - he didn't install an aav at the high point of the system (the loft) - he said they go wrong and that makes them a problem especially in the loft (I would have put a lever valve to isolate it if I had done it). He also said he didn’t believe in balancing radiators - ‘if you have to, you had designed the system wrong’. I think this is a fairly significant misunderstanding of what balancing is and does personally, although I do think that having smart TRVs on each rad and having the rads in ‘star configuration’ hanging off a manifold may mitigate things - but only lockshields restrict flow to a radiator surely? To be fair - I do not experience an issue with some rads heating up well before the others but currently all the lockshields are ‘wide open’. The final thing may just be a personal thing - he used some nice press fittings to the boiler and the odd half decent solder joint (limited because of thatch and his insurance) but under the floor and where not generally visible it's bucketloads of push fit - and IMHO he used way too many elbows, when a little thought with the MLCP or even swept copper would have looked neater, and reduced frictional resistance. It seemed like he didn’t know you could get MLCP press-to-Copper press fittings - I would certainly have paid any uplift but actually the Maincor press to press are cheaper than speed fit! I would say without exaggeration he used about 35-45 elbows in a property with already longish pipe runs. He also failed to insulate the copper flow and return that joined the MLCP in the loft and also the cold feed (that was previously insulated to the unvented). Maybe being picky but it made me maybe unfairly question the outcome)
Last edited:
