Controlling underfloor heating in passive house

[dogman]

Based on the collective opinion of those who have systems what is the best design for controlling underfloor heating in a passive house. 

 

The traditional  current plan is for five zones on the ground floor only.

50lt buffer tank from a modulating air source heat pump zone control and 5 thermostats

 

So do i need five zones 

 

If not how do I control the system. Do i have one zone downstairs, No valves on the manifold. Maybe one control valve to open the flow and one central thermostat and a mixer valve set to supply 23-25 deg floor temp and a circulator running to balance the slab temp

 

Advise and help please

 

 

 

 

 

[Bitpipe]

I have timber UFH on the ground floor only (some electric UFH and towel rads in bathrooms).

 

We have six circuits downstairs - five are in the open plan kitchen / diner / living / study and one is in the hall / stairwell.

 

We control these as two zones separately with wall mounted stats with air sensors and we sometimes run the hall zone a degree warmer than the open plan living area to encourage a little heat upstairs.

 

Seems to work fine - only snafu was not labelling the circuit pairs so a bit of trial and error to figure out which was which - we isolated the hall on second attempt and have never bothered to figure out the rest, much to the sparky's OCD disgust

[Mikey_1980]

On our MBC passive slab We have 11 Zones downstairs all individually controlled via Heatmiser wall mounted thermostats. It works well for us as we can have a 2 degree difference in the various rooms quite easily, ours is supplied by a 70lt buffer on a modulating ASHP.

 

When MBC put the pipes in make sure to label them, we forgot and used a IR camera to detect the circuits when it when live.

[Calvinmiddle]

Seems to be a recurring thing with UFH circuits and MBC labelling them

I'm only on one floor and run the 6 circuits off one thermostat in the open plan living.

 

I have described it here if you haven't seen it.  Seems to work well, even with a mix of tiles and carpet.

 

 

[Jeremy Harris]

I've tied all our UFH zones into a single zone, with a single thermostat.  The heating input is so low that I'm pretty sure I couldn't have got any energy saving by zoning at all.

[Stones]

I have 12 individual loops arranged in 8 zones.  I am currently running this as a single zone off the ASHP master controller / thermostat but did fit individual room stats in the bedrooms so I would have the option of running the bedrooms at a lower temperature than the rest of the house if I wanted.  I can achieve a 2C reduction in room temperature by using those room stats but far prefer running as a single zone.  It's purely a comfort rather than an energy saving device IMHO.

[dogman]

Thanks everyone for the interest and help.

So 

Heat pump has an integral pump that will provide flow to the buffer tank.

The buffer tank will hold a volume of water at a higher temp than the floor.( Is this actually needed if the heat pump modulates?)

Sounds like i then need a pump to feed the blending valve and pump set at the manifold, which in turn circulates the water round the loops.

If I am only running one zone over all the loops then i don't need motorised valves on each circuit, all i need is to lock them open and balance the flow on each circuit.

 

It seems to me that three pumps is not exactly economic

 

If heat pump is set to near the flow temp required by the floor and feeds into the buffer vessel, then what would be the result of doing away with the blender valve set and using a single pump to provide flow through the buffer tank into the floor loops.

 

Control would be a single timed thermostat that would switch the pump on when heat needed in the rooms (somehow). When the temp drops in the buffer tank it will switch on the heat pump and heat the tank. May be able to set some sort of temperature differential to reduce cycling

 

or do away with the buffer tank and the heat pump supplies the blender valve and pump. Will this work?

 

 

 

[Calvinmiddle]

I only had 3 pumps:

boiler to buffer

buffer to manifold

in manifold

due to fact plumber said he wasn't getting "pull" from the buffer by the manifold pump.  I wonder was he not used to sure low temperatures so thought it wasn't working @Nickfromwales was convinced that the buffer to manifold pump wasn't needed, and actually made things more complicated as we had to fit a bypass for when the buffer pump was running but the NC valve on the manifold return hadn't opened yet as this can take several minutes to open once it gets a signal.

 

But Nick is yer man for this one

[PeterW]

The problem with doing away with a buffer - as I am finding - is that the heat pump has to push all of its heat into the water. They tend to have a minimum flow rate on the circulation pump so that dictates your floor flow too - with that you will also have to use the HP to control the water temperature and not directly the floor temperature so any solar gains in one area may skew the temperatures 

 

I plan to use a cheap indirect

copper cylinder as a buffer and heat that to 35c and then let the floor draw off that as and when needed with the HP controlled by a tank stat half way down. This means there is a bulk of water for the HP to work on, and with using W Plan it allows the HP to concentrate on hot water if needed and have the backup of  around 1.6KwH stored in that (if I did the maths correctly..) which should be fine. 

[TerryE]

@dogman, you've got an MBC frame and slab.  You will need to put maybe 1kW of heat on average into the slab in the coldest and darkest months. By all means use 3 or 5 pumps but why? Likewise you can pump heat from zone to zone or you can take the simple path of running all loops in one zone at a comfortable temperature.  Yes, you can make you heating design complex, or you can make it very simple.  Your choice.

 

@PeterW did you follow my Boffin's corner topic?  I can see why a buffer tank will work, but I don't understand why you need one.

[PeterW]

9 hours ago, TerryE said:

@dogman, you've got an MBC frame and slab.  You will need to put maybe 1kW of heat on average into the slab in the coldest and darkest months. By all means use 3 or 5 pumps but why? Likewise you can pump heat from zone to zone or you can take the simple path of running all loops in one zone at a comfortable temperature.  Yes, you can make you heating design complex, or you can make it very simple.  Your choice.

 

@PeterW did you follow my Boffin's corner topic?  I can see why a buffer tank will work, but I don't understand why you need one.

 

@TerryE I have but there are a couple of differences :

- we are not passive slab 

- heat losses are slightly higher due to conversion / extension rather than air tight new build. 

 

The ASHP needs a minimum flow so whilst I could just plumb direct into the UFH it would stop every time the DHW called for heat. We also have most of the solar gain in a morning (in fact we gained 0.8 bar in water pressure on Monday just from the sun before we cast the slab) so there are benefits from running the pumps independently. 

 

Final benefit for me is running everything on the heat pump side at around 1 bar (as per @JSHarris design ) then I only have limited need for antifreeze and inhibitor as it's the ASHP circuit only rather than the whole system. 

[dogman]

As @TerryE stated it can be complex or simple.

 

Our current cottage system has the heat pump feeding the buffer vessel and the underfloor heating pumps draw from the buffer vessel. Its controlled by simple electronic timers and two poly pipe 4 zone controllers,that cost next to nothing on ebay.

It does have zone control but as others have found you tend to set every zone downstairs to the same temp as the MVHR tends to even out the temperature.

 

It looks like the best option is Simple  so do away with the pump set and blender valve. 

Use the heat pump to supply the buffer vessel at low temperature.

Use a pump to supply the manifold 

don't use zone control but install the wiring should it be needed in future

Install one master time and temp controller and use the heatpump controller to manage the system on and off times.

 

Question about the buffer tank is as stated by @PeterW should the buffer tank be a heat exchange to reduce inhibitor costs or just a vessel to hold a volume of water.

If so is it best to get a dedicated heat exchanger or use a tank with a coil

 

 

 

[Calvinmiddle]

54 minutes ago, dogman said:

 

.....as others have found you tend to set every zone downstairs to the same temp as the MVHR tends to even out the temperature.

 

 

 

Not trying to be pedentic here but its not the MVHR evening out the temperature, airflow rate too low to do this, its the fact there is a well insulated passive slab.  there is 30mm of EPS in the passive slab so the heat can't go down, it can go up out of the slab, but air is a good insulator, so it's going to flow through the slab from hot areas to cool areas as concrete is a better conductor (worse insulator), over a while this will mean the slab is all the same temperature.  That is what evens the temperature out.

 

And that should say 300mm of EPS

Edited March 15, 2017 by Calvinmiddle

[dogman]

Don't worry, its why i ask questions and eventually learn

 

[Nickfromwales]

Buffer tank with a coil is best imo, but @JSHarris has made clever use of convection to preheat his PHE so there is little / no delay in producing cold mains uplift to his SAPV's which is exactly what I would recommend ( if the DHW manifold / distribution point is any real distance from the uplift buffer. 

Remotely mounting the PHE closer to the manifold location could also be negated by having a buffer in the airing cupboard and also locating the DHW manifolds there. Latent heat loss would be minimal, but useful too then. 

I cannot see the point / need for a pump to circulate beteeen the buffer and the Ufh manifold. The manifold pump will suck that through with ease unless it's a massively long and complex pipe run. The biggest problems I've had is stopping unwanted ( summertime ) convection circulation from the ( TS ) buffer to the Ufh manifold ! The bloody things pull through without even turning the manifold pump on . That I dealt with by means of a 2-port zone valve, also a requisite if you intend to circulate the Ufh water for solar gain absorption and redistribution throughout the slab, e.g. flow but no heat input. 

Whovever said these haven't pulled through on their own, I seriously doubt. 

[Nickfromwales]

4 minutes ago, Calvinmiddle said:

Not trying to be pedentic here but its not the MVHR evening out the temperature, airflow rate too low to do this, its the fact there is a well insulated passive slab.  there is 30mm of EPS in the passive slab so the heat can't go down, it can go up out of the slab, but air is a good insulator, so it's going to flow through the slab from hot areas to cool areas as concrete is a better conductor (worse insulator), over a while this will mean the slab is all the same temperature.  That is what evens the temperature out.

Only 30mm 

[dogman]

Just checking the acronyms and abbreviations back in a second

 

[dogman]

Ok air source Heatpump is for heating only and not  DHW

 

There is a 6m run between Heatpump and manifold 

 

 

[Calvinmiddle]

27 minutes ago, Nickfromwales said:

Only 30mm 

 

Didn't want to go over board

[Calvinmiddle]

4 hours ago, Nickfromwales said:

 That I dealt with by means of a 2-port zone valve, also a requisite if you intend to circulate the Ufh water for solar gain absorption and redistribution throughout the slab, e.g. flow but no heat input. 

 

Nick surely another way is to have a normally closed ( NC ) valve on the return of the manifold.

 

I can have my manifold pump on all the time, this circulates the water in the loops when the NC valve is closed. NC valve is connected to the thermostat so when it calls for heat, the NC valve opens and thus completing the circuit to the buffer allowing the manifold pump to pull water from it (allegedly). 

[jack]

1 hour ago, Calvinmiddle said:

Not trying to be pedentic here but its not the MVHR evening out the temperature, airflow rate too low to do this, its the fact there is a well insulated passive slab.  there is 30mm of EPS in the passive slab so the heat can't go down, it can go up out of the slab, but air is a good insulator, so it's going to flow through the slab from hot areas to cool areas as concrete is a better conductor (worse insulator), over a while this will mean the slab is all the same temperature.  That is what evens the temperature out.

 

Feedback from an unintentional experiment:

 

We had a leak in the UFH in one room (downstairs study).  The rest of the slab is treated as a single circuit, heated by an ASHP.

 

I haven't managed to find time to repair the leak, so we've had an entire winter with the study effectively unheated.  While the study floor was noticeably colder than the rest of the house, subjectively the air temperature hasn't been all that much different.  I think part of this is the fact that the house temperature isn't actually all that high - we once reached an accidental peak of around 24.5C when experimenting with heating, but more generally we spent winter drifting between 20-22C.    

[dogman]

trying to get my head around this

 

found this on line

 

 

Taking into account that this is probably two floors so you can discount  zones 3 and 4 from what i am trying to establish

 

This shows a heat pump pumping into a TS or BT. A pump with bypass valve 

it then goes to the zones where there is a motorised valve a blending valve then the manifold pump

 

so the system has three pumps 

 

 

This is what i understand to be a blending valve set

 

Will one of these pull water from the buffer tank 

 

if they do why have a pump at the buffer tank?

[TerryE]

@dogman, I used a Wunda pumpset simply because it was the cheapest way to buy the pump and connectors.  If you look at my setup then you will see that I have a TMV blender but since my circulation is always cooler than its minimum setting, it will always be open.  I have a PHE for a different reason, but this could just a easily be used to link in an ASHP.  Also read my blog post on heating the slab and the linked Boffin's corner article.  I currently use an immersion heater in a Willis jacket for heating my slab.  IMO, if you aren't using the buffer to preheat the DHW, then you don't need one unless you have a gas boiler.  You can modulate the output temp of the ASHP down to room temp + 7°C say and at this temperature you can pump the headt directly into the slab and use the slab as the buffer.

[Jeremy Harris]

1 hour ago, TerryE said:

 You can modulate the output temp of the ASHP down to room temp + 7°C say and at this temperature you can pump the headt directly into the slab and use the slab as the buffer.

 

Not that easy, I'm afraid.  Anything lower than about 30 deg C causes the heat pump to cycle too much and it will probably go into anti-short cycle mode, staying off for 15 to 20 mins even when there's a call for heat.  Heat pump life is closely related to the number of start cycles, rather than the run time, too, so excessive cycling is quite likely to cause early failure.

 

I tried to run ours into the UFH only and ran in to this problem, which is why the logic in our control system now automatically turns the valve to the buffer tank on whenever there is a call for heat.  I think it's probably good practice to treat heat pumps like boilers, and try to arrange to run them for as long a period of time as possible when they turn on, and then leave them off as long as possible.  Our 6 kW to 7 kW ASHP will modulate down to about 1.5 kW output, minimum, and that is usually far more than the UFH needs.

[Nickfromwales]

I'd 100% be using the ashp for DHW uplift. Why have such a good resource and not use its full ( good CoP ) potential?

A buffer with DHW uplift coil ( so actually a TS ) is a no-brainer afaic, unless you have a good PV array where the argument becomes less one sided. 

With a typical target 'set' buffer / TS temp of 40oC, and showering requiring 38-40oC max, most of your DHW production can then be from low grade / high CoP heat energy, rather than having the ASHP running in DHW mode, at 55oC, with zero or negative CoP. 

Edited April 18, 2018 by Nickfromwales
Good CoP instead of low CoP.....D'oh!