Underfloor Heating - Can't understand pipework

championc

Hi all

My folks have problems with their underfloor heating. I think I sort of have it sorted but would be very grateful for some explanations from some knowledgable folk on here.

The bits I'm not too sure of are the functions of the two pipes (5) and (6) since they both seem to bridge the flow (1) + (7) and Return (2) + (8). I know the thermostat is (3) which controls the pump, but is this the pumping of the water around the underfloor loop ? Under the pipe section (5) is a wheel (4) which has a "+" and "-", what is this for ?

The final bit which I did not understand was that the pipe section (9) got hot which seemed to be water coming back up from a T joint just out of shot which is where the general central heating circuit joins this underfloor one to return to the boiler.

heinbloed

Under the pipe section (5) is a wheel (4) which has a "+" and "-", what is this for ?

It seems to be a mixing valve. Hence the by-passes.

An unefficient heating system mixing cold and hot water to get the necessary flow temperature.The pump as far oversized for two loops.

championc

If it's a mixing valve (5), then there is no cold water feed - it's a bridge between the flow and return which somewhat sounds like a safety bypass or something. What I don't understand is the pump (6) linking between the flow and return to the right of this pipe (5) - why would a pump be in here ? Is this maybe as a bypass but should be off most of the time to allow the water to circulate around the underfloor pipes (7) to (8) ? And if so, maybe I need to set this clip on Thermostat (3) to a very LOW temp. ?

heinbloed

If it's a mixing valve (5), then there is no cold water feed - it's a bridge between the flow and return which somewhat sounds like a safety bypass or something.

Yes, a mixing valve it is. It mixes water from the flow and the return,'hot' and 'cold'.

Why there is a pump: well, that's Irish plumbing, 1 zone = 1 pump.
You won't find something like this abroad, central heating means also central distribution there, i.e. 1 modulating pump for the entire distribution. This method of plumbing was featured on TV by certain incompetent 'architects' and other vested parties and had been picked up by the trade.
Expensive to install, expensive to run and difficult to optimise.
And difficult to understand due to it's complexity.

championc

Is there a limit to how hot the water can be going around underfloor ? Seems daft to cool your hot water before sending it underfloor.

And as for the pump, I'm beginning to think it's setup to work when you do not want heat underfloor and is bringing the water from the flow into the return pipes - correct ?

heinbloed

Is there a limit to how hot the water can be going around underfloor ?

Yes, every heating system has it's limits.

And as for the pump, I'm beginning to think it's setup to work when you do not want heat underfloor and is bringing the water from the flow into the return pipes - correct ?

It seems absurd but I wouldn't be surprised.
Get a heating engineer in.

championc

From the picture at the top of this thread, the Flow is the top pipe section, beginning with tap (1) and the return is below it with tap (2). Are the parts used here not fairly standard ones ?

From the picture, you can see that there are only three ways to get from the flow to the return - a) via pipe section (5) with the wheel at the bottom, b) via the pipe section (6) with the Pump or c) via the underfloor circuits.

If a) via section (5), what decides what direction the water is flowing (flow to return or return to flow). I somehow don't see how this could possibly be a mixing valve but I'm all ears

heinbloed

The hot water is sucked (by the pump) via the red valve. We call it the flow pipe.
After passing the pump it is pushed into the distribution , we call it the flow group.
After releasing parts of it's thermal energy it apears again in the return group.
And from there it goes back to the heat source (boiler?).

I think we agree on this basic lay-out.

What is the unknown thing supposed to do?

Immagine the hot water coming from the heat source is 70 degrees Celsius hot.
To hot for the UFH. which would demand (maybe!) a 40 degrees Celsius flow temperature and would then return with (maybe!) 30 degrees Celsius.

What can be done to get the prevalent flow temperature of 70 degrees Celsius down to the demand of the UFH?

Right. We mix it.
For example with the 30 degrees Celsius warm water from the return group of the UFH.
As shown on your picture.

Such a mixing valve can be remote controlled, for example via an outside thermostat. We call this weather compensation.
The colder it gets outside the higher the internal thermal energy demand will be.
So the flow temperature of the UFH can be increased (for example from 40 to 45 degrees Celsius) by mixing less cold water (from the UFH return) to it.
And vice versa, the warmer it gets outside the more water from the return is added to the flow.

And of course, yes, we can set the flow temperature of the UFH manually as weel, for example via the wheel.

This seems to be an installation which was added at a later stage, you might still be able to contact the original installer.

What sort of problems do you have with the existing lay-out ?
Except that such a lay-out isn't energy-saving it should technically work.

championc

Thanks Heinbloed

It's starting to make sense now. And to confirm, it's in the floor of a conservatory which, ask you quite rightly deduced, was added later. It's in my parent's house. I have two questions which I hope will get me there 100% to understand how to best manage the heating between winter and spring / autumn.

In my original photo, there are stats above the input (1) and output (2) of the system. So I assume I should see 70 deg C in (1) and 30 C in (2). Assuming I want to have the water in the UFH going in at 45 deg C, what temperature would you suggest I should be getting in the stat at the bottom of section (5) and what temperature should I set the temp on the clamped on Thermostat (3) at the top (since I've messed around with this and don't know where it should be at)

Many thanks, it's greatly appreciated.


C

Plombier

Why there is a pump: well, that's Irish plumbing, 1 zone = 1 pump.
You won't find something like this abroad, central heating means also central distribution there, i.e. 1 modulating pump for the entire distribution.

This wont work where you have two different distribution methods ie. radiators in one section and UF in another, which I assume is the case here.

heinbloed

Championc asks:

In my original photo, there are stats above the input (1) and output (2) of the system. So I assume I should see 70 deg C in (1) and 30 C in (2). Assuming I want to have the water in the UFH going in at 45 deg C, what temperature would you suggest I should be getting in the stat at the bottom of section (5) and what temperature should I set the temp on the clamped on Thermostat (3) at the top (since I've messed around with this and don't know where it should be at)

The setting of the UFH's flow and return temperature depends on the thermal energy demand of the to-be-suported structure and the urge to run the system efficient.

This to figure out is the job of a heating engineer.
Plenty of parameters have to be looked at.

In principal -with every heating system- the chosen flow temperature should be as low as possible.

So it realy depends on what you have and what you want, no one can give an answer to such a question without further research.

A rule of thumb for UFH/surface heating is Delta T of 8 Kelvin. Or in practical numbers: the return temperature of the UFH should be 8 degrees lower than the flow temperature.
And a Delta T of more than 30 Kelvin concerning the temperature fluctuation of a concrete floor/slab (holding the UFH pipes) should be avoided to reduce the risk of uneven and excessive expansion and contraction.
But these are rules of thumb, the most efficient modus of a heating system depends on it's lay-out. Which has to be established by a professional.


Plombier wrote:

This wont work where you have two different distribution methods ie. radiators in one section and UF in another,...

It works in millions of buildings all over the world.

Plombier

With steel panel radiators delta t should be around 10K and a flow temperature of around 75C. the gives a delta t of 50K for selecting the radiator size.
With UFH as you have already said the delta t should be around 8K and a flow temperature of around 40C.
With steel panel radiators if you use a delta t of 8K and a flow temperature of around 40C. gives a delta t of 32K for selecting the radiator size.
the output from the radiatior is not linear, quinn radiators maybe the largest manufacturer in europe use an exponant of 1.3 when applied to these
figures it will reduce the output of the radiator by 80%.
Using a flow temperature of 70C for UFH again wont work.
So you need two mixing circuits and each of these will need a pump.

heinbloed

Plombier wrote:

"So you need two mixing circuits and each of these will need a pump."

This is not true. Contact a plumber on this.

The installation of an efficient heating system is still a mystery to many. But it makes no difference what sort of thermal distribution is choosen, radiator or UFH or a combination.
Each circuit can be run with it's own individual flow and return temperature.
Simply by installing a mixing valve with each circuit. And only 1 central pump.

Hear-say is not a great help, plombier.

Plombier

It seems obvious that you dont understand the difference between variable volume and a variable temperature heating circuits, variable volume will be used for DHW, AHU's etc. but not for UFH or radiator circuits.

As you say Hear-say is not a great help.

galwaytt

heinbloed wrote: »

Each circuit can be run with it's own individual flow and return temperature.
Simply by installing a mixing valve with each circuit. And only 1 central pump.

Hear-say is not a great help, plombier.

Well mine's not hearsay: I've had UFH since the early '90's, and a single central pump is neither sufficient, nor efficient. It leaves you exposed to poor fluid dynamics most commonly exposed to performance dicatated by poor installation or restrictive layout on site. You need to take control of each circuit, and control it, and not rely on incidental convection to move it around. Whether that means adding 1, 2, or 22 pumps, then so be it.

Even in the last week alone, when the UFH pump in my basement went down, manually opening the mixing valve, and relying in the central pump on the buffer was pointless. A quick manual re-wire of the pump got the circuit pump going again, and heat restored inside............15 minutes ?? YMMV and all that,.........

galwaytt

OP - just a quick (related) question: is that flow/return pipework to the manifold's, un-insulated ? And, if I see it correctly, everything is just screwed to a masonry (aka, cold), wall ?? Is that an external or internal wall, btw ?? You don't want to be trying to heat the wall as well !!

heinbloed

OP - just a quick (related) question: is that flow/return pipework to the manifold's, un-insulated ? And, if I see it correctly, everything is just screwed to a masonry (aka, cold), wall ?? Is that an external or internal wall, btw ?? You don't want to be trying to heat the wall as well !!

Well spotted, galwaytt.

Insulation wouldn't be expensive. The hot flow pipe coming from the central heating should be insulated, at least separated by distance from the rest. Any thermostatic valve or thermometer installed near to an uninsulated heat source will give wrong readings and won't deliver to what is was designed for.


About zoned heating and pumps:

I quote Plombier first:

It seems obvious that you dont understand the difference between variable volume and a variable temperature heating circuits, variable volume will be used for DHW, AHU's etc. but not for UFH or radiator circuits.

It seems to me that the idea/concept of hydraulic adjustment (restricting the flow volume with valves to optimise the system ) isn't understood at all.
Check the picture posted by the OP again. Hydraulic adjustment (reducing the flow volume of individual cicuits individually) is planned for, is ready to be used, is payed for. But not understood,well, not by the OP or you.
The OP asks friendly and will get a friendly answer to this as well.

It does not need a pump for ever differing flow rate, plombier.
How do you do the hydraulic adjustment in your installation, Plombier? With an individual pump for each circuit ? Absurd. But technically no problem...


About the pump being to weak to run both, the UFH and the radiators together as galwaytt is experiencing:

Well mine's not hearsay: I've had UFH since the early '90's, and a single central pump is neither sufficient, nor efficient. It leaves you exposed to poor fluid dynamics most commonly exposed to performance dicatated by poor installation or restrictive layout on site. You need to take control of each circuit, and control it, and not rely on incidental convection to move it around. Whether that means adding 1, 2, or 22 pumps, then so be it.

Even in the last week alone, when the UFH pump in my basement went down, manually opening the mixing valve, and relying in the central pump on the buffer was pointless. A quick manual re-wire of the pump got the circuit pump going again, and heat restored inside............15 minutes ?? YMMV and all that,.........

Poor flow dynamics are the system designer's fault, not the fault of the pumps(smiley). But to the point of curing the problem:
Calculate for the pump the hardest job first and the easiest job afterwards. Install the modern, modulating pump (those which change automatically the flowrate/pressure according to demand). And then the radiators.

Both distribution systems can be hydraulically adjusted in case the hydraulics don't fit - which is usually the case. This also an EN standard which MUST be adhered to, not doing so will cause the informed consumer to call you back. He/she has the right get the entire system optimised, i.e. hydraulically adjusted. This is part of any central heating installation job.

This hydraulic adjustement must be done on each circuit but can be done seperatly for each group (radiator group, UFH group) as well.
If supporting the entire distribution system (a combination of UFH with radiators) with 1 pump the hydraulic adjustment for the radiator group is strongly recommended to avoid flow noises in the piping and at the radiaotor valves.
Similar to adjusting the flow TEMPERATURE of each group the flow VOLUME can and MUST be adjusted as well. And such a set-up realy needs only 1 pump.

As long as this single pump can deliver the maximum demanded pressure it will do the rest as well.

A modulating pump - correctly sized and installed - will deliver as much as necessary, be the demand low or large.

With flow restricting valves EVERY pump can be used to deliver various amounts of liters per minute at every group of the system.
We call this hydraulic adjustment.


About over-running a dead pump with a live one:

Qote:

" ....when the UFH pump in my basement went down, manually opening the mixing valve, and relying in the central pump on the buffer was pointless. "

I believe you every word, it is logic: the dead pump works as a break, restricting the flow.

heinbloed

This is a hydraulic scheme showing the set-up of central heating system with only 1 central pump, supporting hot heating water at differnt flow rates and temperatures to various parts of a building. Basic plumbing since wet central heating systems are used:
http://www.bfw-gohl.de/images/Schema mit hydraulischem Abgleich klein.jpg[/IMG]






Only 1 pump.
[/LEFT]

heinbloed

Plombier

Your image shows one unmixed heating circuit so its not relevant.
Follow the link below and you might understand how its done.


http://www.buderus.co.uk/index.php?fuseaction=schematic.detail&content_id=1446

championc

galwaytt wrote: »

OP - just a quick (related) question: is that flow/return pipework to the manifold's, un-insulated ? And, if I see it correctly, everything is just screwed to a masonry (aka, cold), wall ?? Is that an external or internal wall, btw ?? You don't want to be trying to heat the wall as well !!

Hi Galwaytt,

This was originally an outside shed and the wall you see is the back wall of the house. This shed is now indoors and now a small room off the conservatory. The pipework leading to the pipes on the left is insulated to just below the photo. However, the orange and white underfloor heating pipes are un-insulated from the manifold down to where they disappear through the floor.


C

heinbloed

@ Plombier:

You shouldn't mix-up cascade systems with single boiler systems, this is a different story.

The link provided by me does show a multy storey appartment block after a correct hydraulic adjustment.
It needs only 1 central pump in the basement to deliver hot heating water at various pressures (i.e. flow rates). Different pressures at every storey because of the height difference.
The scheme's details, where the hydraulic adjustment valves are placed, the thermostats, the group for each apartment and so on are NOT shown.

But this issue is getting-off topic, if you wish to then open a new thread concerning optimisation of central heating systems.

Plombier

The installation of an efficient heating system is still a mystery to many. But it makes no difference what sort of thermal distribution is choosen, radiator or UFH or a combination.
Each circuit can be run with it's own individual flow and return temperature.

What you are describing in this statement and in the schematic is a means to balance the system and deliver the correct amount of energy to each part but it will not control temperature being delivered.

Simply by installing a mixing valve with each circuit. And only 1 central pump.

Using a mixing valve without a pump on the load side will just make the valve operate as a diverting valve.

This is where you have gone wrong you can not have individual circuits operating at different temperatures to do this you need a pump for each circuit.

heinbloed

Plombier wrote:

Using a mixing valve without a pump on the load side will just make the valve operate as a diverting valve.

This is where you have gone wrong you can not have individual circuits operating at different temperatures to do this you need a pump for each circuit.

No, wrong.

But you better try it out youself, I'm getting tired with this.....

Check your local plumbing supply shop for the various valves they have, thermostatic mixing valves are ready available. Maybe someone there can explain to you how the different types work. A plumbing course might be helpfull as well.

Open a new thread, this might atract some plumbers/heating engineers to advise you on the issue.

Plombier

Look carefully at these examples.

With the pumps as shown the same volume of fluid flows through the secondary circuit at all times and varying amouts are injected from the primary circuit in a mixing fashion thereby reducing the temperature to the secondary circuit as required.

If we put the pump before the valve the opposite will be true.

The same volume of fluid flows through the primary circuit the valve will now work as a diverting valve and no mixing will take place but instead it will vary the volume through the secondary circuit.

To achieve different temperatures to different circuits the example in the picture is how it is done.

If you can show me any example of how it can be done differently then please do.

But all you have done so far is said that it can be done but without explaination.