Optimising Domestic Heating (UFH and Radiator mix)

[AliMcLeod]

For various reasons (I can go into details on a separate thread if anyone is interested) my wife and I had to postpone/put on hold our self build and instead moved house at the end of last year.

 

I’m not convinced the heating setup in this new house is as optimised as it should/can be so am looking for some pointers. I know you guys like details, so here’s what my investigations have found. Apologies if this is too much ?

 

The house was originally build around 15 years ago – it is 1.5 floors with 2 upstairs dormers into the roofspace - and then extended 9 years ago. The extension used the full house height to add a 4 meter high open plan area and new master bedroom suite (again, into the roofspace). So the current downstairs setup is an open plan kitchen-dining-living area, with the double height dining-living area and a double height master bedroom/en-suite. Downstairs also has a separate family room, a small study,  and another downstairs bedroom, also en-suite, but this bathroom is a weird Jack-and-Jill type setup where it also doubles as the downstairs public bathroom.

 

Upstairs has two bedrooms, one en-suite and there is also another public bathroom.

 

The downstairs has UFH throughout and is tiled. Upstairs has 3 radiators with TRVs (one in each bedroom, one in top hall) and there is a towel rail with valve in each of the bathrooms. The stairs and top hall are carpeted. Both bedrooms have laminate flooring. Both upstairs bathrooms have tiled flooring. As well as the 4 showers, we have one bath - a monster in size but only used approximately once a month.

 

I don’t know the makeup of the insulation of the house walls or the UFH. The original house was developer built, one of 3, so I suspect it’ll have passed the reg’s (of the day) but little more. The extension was done by a local builder. We’ve not been in long enough to really understand the heating needs of the house yet - we had a mild winter, then that mad snow in March, so I haven’t yet got a handle on how good or bad it might be. We do have 3 south facing Velux windows (on the sloping part of the roof) and lots of North facing glazing to the open plan living area, another 3 Velux windows  to our master bedroom and 1 to the master en-suite, and we do get quite a bit of solar gain in that space.

 

The boiler is situated in cupboard by the front door and is a mains gas powered Bosch Greenstar 42CDi Classic. The 8 port manifold is around 10 feet away in a cupboard further into the house and was fitted by NuHeat. It serves the 4 downstairs zones (living area, master en-suite, second bedroom en-suite/bathroom/study and family room), each with its own thermostat (Danfoss stats with wireless comms to the two Danfoss receivers fitted above the manifold).

 

As far as I can ascertain, the upstairs radiators/towel rails are fed directly from the boiler but there is no upstairs stat! This means that, unless I am missing something (I can’t follow all the pipework), the upstairs rads can only work when at least one of the UFH zones is calling for heat!

 

There is no Thermal Store (is that @Nickfromwales screams I can hear....) and no easy place to add one. There are no PV panels and no plan to add any.

 

From a usability perspective, just me and my wife (and 2 cats) live here for the majority of the year, so the upstairs is pretty much shut off unless we need to get to storage or our daughter is home from university (not regularly). We do have family visiting 2-3 times a year in which case typically have a full house (we have 5 staying with us at present, 3 adults and 2 teenagers) so all 4 bedrooms are in use.

 

Based on my reading of the threads here’s here, and taking into account the advice posted on ensuring condensing/modulation/reduced short cycling etc, this is what I’ve done with the setup:

 

• I’ve put the boiler DHW setting at the E setting (economy, approx. 50C according to the manual)
• I’ve put the boiler CH setting at around 3.5 (approx. 55C according to the manual)
• Reduced the manifold blending value to 40C (it only goes down as far as 35C and the UFH has not actually been on since I did this so this is a placeholder for future experimentation once the it starts to get colder –typically September in Scotland ?)
• Put the TRVs for the 3 upstairs rads to 15C  and turned off the two towel rails, I’ll change this as we come into winter if we have people to stay or if it starts to get cold in this part of the house – it has been overheating in the summer unless we open upstairs windows.
• I topped up the boiler pressure which had dropped to 0.5bar (the manifold blending valve has a very slight leak)

 

Some questions:

 

• Does the above setup seem a reasonable starting point for me to experiment with further as it starts getting colder?
• Am I missing anything? In particular, could there be a way the upstairs rads call for heat without a stat when all the UFH zones are at their respective setpoints?
• Is there an issue with all my upstairs rads having TRV/values – did I read that there should be somewhere for the heat to flow if these are all off? Could there be something in the pipework that I’m not seeing that is achieving this?
• Any other advice on what other optimisations I could make?

 

To those that made it this far, thanks for reading.

 

(I’m going to try get some manifold pics off my phone and will post them next)

 

 

 

[AliMcLeod]

 

[Ferdinand]

If it is not to passive house levels experience says you need a lot more cats ?.

 

Will try and digest and comment a little later, since the forum helped me readjust my similar setup earlier this year.

Edited August 7, 2018 by Ferdinand

[ProDave]

First thing. there must be a circulating pump somewhere else as well as the one on the manifold. Find this. Then find out if it is running when the central heating programmer is on, but ALL UFH room stats are turned down.  If that is the case then the radiators can run without any UFH zone calling for heat.  If so and there is no room stat upstairs, it will only be the radiator themal valves controlling the temperature in those rooms.

 

[AliMcLeod]

10 minutes ago, Ferdinand said:

If it is not to passive house levels experience says you need a lot more cats ?.

 

Its definitely not to PH standards, but when we had the bad snow in March, the UFH sufficed in keeping us warm. So, I'm not overly concerned we don't have enough to heat the space, just that our fuel bills don't end up silly.

And, i forgot to add - there was a wood burning stove in the double-height living area when we moved in, but it was ugly, verging on unsafe/contravening regs (flue too close to wall) and leaked smoke when we tested it, so we've recently had a gas fire installed in its place (Corner Bell 3 Large). Its as much for room aesthetics as heat, but it will be good to augment a low UFH heating setting for the living area when there's just the four of us (including cats) there.

Edited August 7, 2018 by AliMcLeod

[AliMcLeod]

19 minutes ago, ProDave said:

First thing. there must be a circulating pump somewhere else as well as the one on the manifold. Find this

 

Thanks @ProDave - would the boiler pump not be enough to circulate the water around the rads? 

 

Quote

Then find out if it is running when the central heating programmer is on, but ALL UFH room stats are turned down.  If that is the case then the radiators can run without any UFH zone calling for heat.  If so and there is no room stat upstairs, it will only be the radiator themal valves controlling the temperature in those rooms.

 

I think (but will confirm) that even when the boiler timer has the CH on (I didn't mention that above - i have the boiler stat have the CH on during the day and off overnight but am thinking it should eventually be on 24/7), the burner does not kick in even if the rads are turned up full and I'd be expecting them to call for heat. The burner only kicks in when at least one UFH zones call for heat - but let me confirm. 

 

I can definitely confirm that the upstairs rads heat up when required when the UFH IS calling for heat.

 

Our guests go home tomorrow so I'll have a more detailed look once they're gone - i'm not home till late tonight (Ediburgh Fringe) and don't want to be overheating the house before everyone heads off to bed.

Edited August 7, 2018 by AliMcLeod

[PeterW]

Boiler pump would do the rads but would fight with the UFH manifold potentially. 

 

Any wiring centres anywhere you can take a lid off and have a look at ..?? That way you can see what triggers the boiler switch. 

 

Boiler is a monster combi - shouldn’t have any issues with that ..! 

[AliMcLeod]

Here’s the photos I have. The stat wireless receiver feeding into the (as yet unopened) wiring centre. I’m sure that has the trigger wire to the boiler but will check tonight.

 

 

 

In the second picture, you can see the entirety of the very detailed documentation the previous owners left for us!

[AliMcLeod]

15 hours ago, PeterW said:

Boiler pump would do the rads but would fight with the UFH manifold potentially.

 

Can you elaborate on how that would happen? Do you mean the boiler pump is circulating hot water but it'll get "blocked" at the manifold (blending valve/flow gauge/actuators?) How could I tell?

[Ferdinand]

I cannot find my original thread from the end of last year; it s somewhere ?.

 

My setup is exactly the same Combi as you, ufh downstairs and rads upstairs, controlled via 2 separate programmers. UFH have room room stats, rads have TRVs. There are a couple of small pipe runs that are on whenever the heating is running. UFH was broadly zoned by room. the controllers are overkill imo. I do not have a buffer tank, so there is some cycling.

 

My problem was that the ufh was not getting heat all the way through, which meant that it was unpredictable what was running hot and cold, and I was having to mess about with flow tenperatures too much.

 

The main things that we adjusted under guidance from the gurus,

 

Pump setting, to make sure that the heat was getting all the way round.

Boiler flow temp, to keep the flow/return difference to a minimum.

 

I have ended up with a flow return difference of about 12-20C in winter, which is still quite a bit but far better.

 

Best of luck.

 

F

 

 

 

 

[AliMcLeod]

On 07/08/2018 at 17:08, PeterW said:

Any wiring centres anywhere you can take a lid off and have a look at ..?? That way you can see what triggers the boiler switch. 

 

Here's the wiring. The brown (with grey sheath) wire at the bottom left leads to the boiler. The two black ones nearby are the power from the CU (via an isolator switch) and to the manifold pump. The one weird thing i noticed is that in the connection panel to the actuators, connection 5 leads to two actuators and 6 is empty. Any reason why that might be?

 

 

 

I can confirm that with the UFH off, the upstairs rads do not call for heat.  The only pipework at the manifold is the feed from the boiler and the return to the boiler. The pipework for the upstairs rads is not accessible but I can only assume it is fed directly from the boiler.

 

I also noticed that the manifold pump is on speed 3.

Edited August 9, 2018 by AliMcLeod

[PeterW]

top left of the NuHeat boards is a no volt boiler interlock... he’s not bothered to use it ..!!

 

also is that a single wire (Brown T&E) going to the boiler or is the neutral connected too..??

 

With the zone 5/6 thing you could find that either the relay is knackered on 6 or that 5&6 serve the same area so it’s been zoned so that both circuits come on together.  Bizarrely though it looks like a single stat controls it all so that would be irrelevant anyway. I can’t see properly but are there links between all the “return” wires (white cables are in the way)

 

Has it got the NuHeat controller or stat ..?

[ProDave]

Something missing there.  I would expect the top left "call for heat" contacts to connect to the boiler to turn it on when the UFH wants some heat.  Without that connected, what turns the boiler pump on and off?

 

I suspect at the moment it runs all the time the boiler timer is on, so I would expect the radiators to work.

 

It would be better if this call for heat contact controlled the boiler, and you could put an upstairs room stat in parallel with that, so the boiler would only come on when either the UFH or upstairs actually wants some heat.

[AliMcLeod]

So I did some testing yesterday evening.

The setup

 

- Boiler at 3.5 (55C)

- Manifold pump at speed 3

- Blending valve at 40C

 

I triggered a request for heat via one of the UFH room stats and then the system fairly predictably followed this sequence

 

- The boiler kicks in and the temperate rises on the boiler display

- The actuators (there's two for this stat) open (the top bit kind of rises up showing blue) 

- The manifold flow temperature gauge starts rising

- The boiler temperature rises to around 60C before falling back

- At that point, the boiler stops burning (short cycling?) but I think on on occasion the temperature fell back but the burner was still on (modulating?)

- At the point where the boiler has reached its max, the flow temperature gauge has overshot the blending value setting, maxing out at around 50C, thereafter falling back to 40C - it fairly consistently bounced between the two temperatures

- Throughout this test, the return temperature gauge varied little, between ~18 and ~22

 

Any way to stop the short cycling?

 

[PeterW]

8 minutes ago, ProDave said:

Something missing there.  I would expect the top left "call for heat" contacts to connect to the boiler to turn it on when the UFH wants some heat.  Without that connected, what turns the boiler pump on and off?

 

I suspect at the moment it runs all the time the boiler timer is on, so I would expect the radiators to work.

 

It would be better if this call for heat contact controlled the boiler, and you could put an upstairs room stat in parallel with that, so the boiler would only come on when either the UFH or upstairs actually wants some heat.

 

@ProDave that’s slightly incorrect

 

Looking at how it’s wired the installer has used the manifold pump live to fire the boiler not the no volt contacts. This means the boiler only fires when the manifold pump activates. Not sure why it’s been done like this ....

 

This looks remarkably similar to the wiring that we unpicked for @newhome - maybe it’s a Scottish thing... 

[PeterW]

2 minutes ago, AliMcLeod said:

So I did some testing yesterday evening.

The setup

 

- Boiler at 3.5 (55C)

- Manifold pump at speed 3

- Blending valve at 40C

 

I triggered a request for heat via one of the UFH room stats and then the system fairly predictably followed this sequence

 

- The boiler kicks in and the temperate rises on the boiler display

- The actuators (there's two for this stat) open (the top bit kind of rises up showing blue) 

- The manifold flow temperature gauge starts rising

- The boiler temperature rises to around 60C before falling back

- At that point, the boiler stops burning (short cycling?) but I think on on occasion the temperature fell back but the burner was still on (modulating?)

- At the point where the boiler has reached its max, the flow temperature gauge has overshot the blending value setting, maxing out at around 50C, thereafter falling back to 40C - it fairly consistently bounced between the two temperatures

- Throughout this test, the return temperature gauge varied little, between ~18 and ~22

 

Any way to stop the short cycling?

 

 

Not without a buffer or turning down the boiler temperature but then it won’t condense. 

 

Boiler will hit 60c as the semi-closed actuators will restrict the flow - much prefer the new motorised actuators that also manage the flow and return temps. 

[AliMcLeod]

35 minutes ago, PeterW said:

top left of the NuHeat boards is a no volt boiler interlock... he’s not bothered to use it ..!!

 

also is that a single wire (Brown T&E) going to the boiler or is the neutral connected too..??

 

With the zone 5/6 thing you could find that either the relay is knackered on 6 or that 5&6 serve the same area so it’s been zoned so that both circuits come on together.  Bizarrely though it looks like a single stat controls it all so that would be irrelevant anyway. I can’t see properly but are there links between all the “return” wires (white cables are in the way)

 

Has it got the NuHeat controller or stat ..?

 

No neutral to the boiler, just live (brown).

 

4 stats control the 10 zones (3 with 2 zones and 1 with 4 zones) and i'm pretty sure that's all working fine. By returns, do you mean the "SL Return" block - i do see loops in there. Can post a closer picture if needed? 

 

Here's the inside of the wiring box cover, along with a hand-written piece of paper that was tucked in there when it was changed in 2015. The previous owner bred cats (we're still trying to get rid of the smell from the under-stair cupboard!) hence the "Cats Room"!

 

 

 

No NuHeat controller or stats. 4 Danfoss stats and the 2 Danfoss wireless receivers you can see in the first pic above.

Edited August 9, 2018 by AliMcLeod

[PeterW]

4 stats and 2 receivers ..... 

 

Right - move the white cables out of the way, then take a photo of the switched live connections please. 

 

The single brown feed is the boiler SL - spoke to @Nickfromwales and the WB needs a 230v trigger. Still possible from the boiler  trigger on those boards, just needs a link. 

[AliMcLeod]

Not at home now but here’s what else i took last night.

 

Whilst not necessarily ideal, is there actually anything wrong with triggering the boiler at the same time as the manifold pump is switched on?

 

 

Edited August 9, 2018 by AliMcLeod

[PeterW]

Ok so it looks like they have basically linked 1-4 then 5-10, no issue with that but it could be neater and ideally should all be brown sleeved as it’s all live. 

 

Can’t fathom the two in 5 and no 6, but that’s not an issue. 

 

There is is no bypass valve on the manifold so you really need a method of firing the boiler from the board and a separate upstairs stat as @ProDave said but you need to do some rewiring of that board first as otherwise you will back feed the manifold pump. 

 

Before you do this, have the towel rails for TRVs on or are they just simple valves ..?

 

 

[PeterW]

Oh and what spec are the Danfoss room stats..? Dumb or programmable ..??

[AliMcLeod]

Towel rails have turnable values, but no markings - but I suspect they are TRVs? Room stats are programmable  (TP5000si-RF)

[PeterW]

Ok makes things easy as they are still available. 

 

Worth taking a quick photo of the towel rail valve and posting here if poss as not many TRV options for towel rails and it would give you your bypass rad otherwise you need to consider a bypass valve somewhere. 

[AliMcLeod]

47 minutes ago, PeterW said:

Worth taking a quick photo of the towel rail valve and posting here if poss as not many TRV options for towel rails and it would give you your bypass rad otherwise you need to consider a bypass valve somewhere. 

 

Will take a photo tonight. I did have a brief look at it a month or so back. It turns and the control unscrews to reveal what i'd call a standard radiator valve that a TRV fits to.

 

What issues would i have without a bypass? As i said initially, i was concerned about this as all 3 rads have TRVs and both towel rails have these valves (same towel rail in each upstairs bathroom)

Edited August 9, 2018 by AliMcLeod

[PeterW]

I doubt towel rail has a trv. Without a bypass of some sort you could find the boiler cutting out due to an overheat die to lack of flow