Low temperature thermostatic mixer

[MarkB]

I'm looking for a thermostatic mixer valve, of the type used for UFH systems that can operate over a range of 20C to 25C.

Normally they operate at a minimum output of 35C, which is too hot for my requirement.

Any ideas?

[dhaslam]

There seems to be  some available  but expensive.  These  Watts ones seem to  need a high  minimum  temperature   on the hot inlet that might not suit.

http://media.wattswater.com/1915905.pdf

[ecogeorge]

Are you sure you want as low as 20-25c ? Will be very slow to heat , especially from cold.
Are you running from GSHP ? and 24/7 ?
rgds George.

[pb]

I'm looking for a thermostatic mixer valve, of the type used for UFH systems that can operate over a range of 20C to 25C.

Normally they operate at a minimum output of 35C, which is too hot for my requirement.

Antares have one that is specified for 20-65C, their article E.149.

http://www.antaresint.com/PUBLIC/php/first.php?middle=scheda_prodotto.php&COD_CAT=E.149&messaggio=Thermostatic%20mixing%20valve%20for%20underfloor%20heating%20systems.&right=0.php

20C does seem quite low though.  What are you planning to do with it?

[MarkB]

Thanks for the answers and suggestions.

The Watts Water website seems to be down, so I'll try again later.

The Antares looks like it will do the job, although it is a very large adjustment range (45C in one rotation of the control) which leads me to wonder how accurate the mechanism is. Unfortunately there doesn't seem to be any datasheets without beign registered. I'll try some googling and see what I get.

The interest is probably more of a thought experiment that anything else. There is a dilemma for the best means of providing comfortable heat to a very low energy house, of the standard of a PassivHaus or better. Contrary to popular opinion these can need some heat in the coldest of weathers.

Warm air systems are complicated because they need high temperature water (60C) to heat the supply air from the MVHR and very careful design to make sure the heat gets to where you need it. This makes the use of renewable energy or heat pumps impractical.

Radiator systems have advantages in being quick to respond, well known technology, relatively cheap and can work with reasonably low water temperatures (assuming well oversized) and can be matched to heat pumps. You still need 40C plus water, even with very large oversizing. However, they take up wall space and lead to air currents in the room with the warmer air being more likely to be near the ceiling.

Fan heaters are a neat possibility, with low capital outlay. They're generally ugly (arguably Dyson excepted), but can be stored away except when not needed. Unfortunately they use grid electricity unless there is enough solar for PV. I'm also not sure how well the heat is distributed in a large room.

UFH is a problem, because it typically runs at 35C and once you have heated a slab to 35C it will be pumping heat into the room long after any thermostat has turned it off. This risk serious overheating, opening windows and therefore innefficiencies. It's also very high in capital cost but it leads to very comfortable heat and is well matched to heat pumps.

However, very low temperature UFH is an interesting concept. If you can run the slab at low temperatures, say 21C, then there is zero risk of overheating - once the room reaches 21C then there is no more heat transfer. A typical UFH density leads to between 8 and 11W/m2 per degree C above ambient, which is just about perfect for a PassivHaus. Any heat is well distributed, and there is no space lost for radiators. Being such low temperature, there is a chance that it could be used with low-grade heat, possibly from evacuated tubes, on a bright winter's day, or at a high COP from a heat pump. It would be slow to respond, but in a high thermal mass PassivHaus this would rarely be a problem - it would take days to cool down even in the coldest of weather. Of course, it's a very expensive system for such a small amount of heat required, hence it being a thought experiment more than anything else.

If anyone has any other suggestions on low temperature mixers then I'd be interested to hear them.

[KLD]

Mark

Are motorised mixing valves out of the question? You could then adjust the controls to your heart's content, have sensors where you need them, etc.

Just a thought
Klaus

[MarkB]

Are motorised mixing valves out of the question? You could then adjust the controls to your heart's content, have sensors where you need them, etc.

A motorised valve would be a definite possibility. A controller could theoretically monitor the supply and return temperatures on a manifold and adjust the valve to keep the slab at 21C. The loop time would be so slow compared to the temperature sampling rate that it probably wouldn't need any clever control system, just something basic to cope with initial cold scenario (in this case it could ramp the temperature up to normal UFH for a short period to get the heat up rapidly).

I also suspect it would be within my capabilities to knock something up, maybe Arduine/Nanode based, but I'd prefer to find a foolproof, off the shelf solution which would be in the realm of non-engineers. I'm always paranoid about doing anything custom - it might work great, but it would be a nightmare to maintain, especially for any future owners if we ever came to sell the house.

(Still primarily a thought experiment.)

[billt]

You might find that a standard TMV will work at lower temperatures than specified. I'm experimenting with low temperatures as my screed dries and I've found that this TMV http://www.underfloorheating1.co.uk/22mm-blending-valve.html seems to operate satisfactorily at 22C, although it's only supposed to go down to 30C.

I'll probably end up running the TMV at about 35C and using the manifold valves to control the floor temperature.

[Iain]

Hi
My underfloor heating has water straight off the boiler ,at the moment about 45/50 deg c, via a motorised valve. My remote stat has its sensor in the floor screed. The temp I set, 18/20 deg c is the temp the floor stays at. The motorised valve switches off/on as required to keep it stable. The thermal mass of the floor gives hysteresis so the valve doesn't just keep switching on and off all the time. Seems to have worked quite well for about 5 years. If you used the standard 35 deg c TMV the valve should cycle even slower.
Iain

[MarkB]

You might find that a standard TMV will work at lower temperatures than specified. I'm experimenting with low temperatures as my screed dries and I've found that this TMV http://www.underfloorheating1.co.uk/22mm-blending-valve.html seems to operate satisfactorily at 22C, although it's only supposed to go down to 30C.

I'll probably end up running the TMV at about 35C and using the manifold valves to control the floor temperature.

That is interesting. Is 22C as low as it goes, i.e. the valve turned all the way down?

I didn't realise that the manifold valves had any sort of temperature control - I thought they were for initial balancing of flow rates in different zones and soem with actuators to turn the flow on and off. (I guess that the temperature can be controlled by slowly pulse-width modulating just using on and off, but you'd still need some sort of controller - what are you planning to use?)

[MarkB]

Hi
My underfloor heating has water straight off the boiler ,at the moment about 45/50 deg c, via a motorised valve. My remote stat has its sensor in the floor screed. The temp I set, 18/20 deg c is the temp the floor stays at. The motorised valve switches off/on as required to keep it stable. The thermal mass of the floor gives hysteresis so the valve doesn't just keep switching on and off all the time. Seems to have worked quite well for about 5 years. If you used the standard 35 deg c TMV the valve should cycle even slower.
Iain

Ah, so just by burying the thermostat in the screed for each "zone" it does exactly what I want. Are these motorised valves just the ones that fit into the manifold to turn flow on and off?

Are you also running this in a low energy house to get away with slab temperatures that low?

I would be interested in more details about your system if available - any diagram or system reference would help confirm my understanding.

[billt]

That is interesting. Is 22C as low as it goes, i.e. the valve turned all the way down?

I didn't realise that the manifold valves had any sort of temperature control - I thought they were for initial balancing of flow rates in different zones and soem with actuators to turn the flow on and off. (I guess that the temperature can be controlled by slowly pulse-width modulating just using on and off, but you'd still need some sort of controller - what are you planning to use?)

There is a bit more valve movement available, but the temperature is so close to ambient it's a bit difficult to set.

The manifold has balancing valves on the flow side and powered wax on-off valves on the return. Although the valves take nearly 2 minutes to open and close, the response of the floor is even slower so it should be possible to maintain an even floor temperature with them. I'm using a windows program called Homeseer to monitor and control things so I'll write a script for that.

[Iain]

Hi Mark
We just use it to keep the chill off the room so keep it low. I just use a standard 22mm SUNVIC MoMo zone valve, it feeds one of those 22mm to 4 x 10mm adaptors. I have 4 (10mm barrier pipe) loops under the Sitting/Dining room.
It is an older house, solid brick!! I use a Danfoss room stat with an external sensor. My system is on a suspended wooden floor, I have 50mm kingspan between the joists with a 25mm air gap between the Kingspan and the flooring sheets. The 4 wet loops are in this 25mm air space along with the temperature sensor for the thermostat.
So with a solid slab floor the Hysteresis would be a lot greater and the temp would be quite stable I would imagine. I use a MoMo valve so that it is only using power when the valve is actually moving and not when it is fully open or shut.
Iain

[MarkB]

Thanks for the info Ian. I'll think some more.

[Iain]

Hi Mark
If you do go down that route it is worth embedding a 15mm copper tube in the screed for the temp sensor. If the sensor ever fails you can just pull it out of the tube and slot in a new one. I actually used a piece of HEP but had a piece of copper on the end for the sensor to be inside.
Iain