Manifold puzzle clarity

[Sumo]

Hello, I am fairly convinced I have a manifold flow direction issue. I thank JohnMo and ProDave for there contributions in my introduce yourself post.

 

I have a fifteen year old ashp-UFH, manifold system that I can not believe has been plumbed in REVERSE and thus has been costing my old parents a fortune all this time. Also confused how a Mitsubishi accredited engineer could comission it as fit, but that is not the issue.

 

I know a manifold can be plumbed into the left side for flow and return whilst having a circulating/mixer pump on the right side, that would not be an issue. It is the fact that the FLOW is entering at the bottom and the RETURN is comming from the top, see below for the numbers, the flow should be through the top.

When I disable the mixer pump I get zero flow/min through the meters(are they one-way flow meters? thus hitting the flow meters from the wrong side) dispite having two additional circulation pumps and the ashp pump working. With the mixer pump I get 1.5 to 3l/min depending on pump speed, though as you can see from the numbers below , it is not going in the right direction.

It is like it is working against itself.

I removed the actuators and opened up the temperature /mixer valve plus opened the manual ball valve whilst trying to fathom this, the mixing pump runs on wall stat demand? The infra red heat gun figures are the key.

Any clarity to say yes you are right, swap the flow and return around , would be helpfull.

 

 

[ProDave]

I repeat the diagram from the other post.  Everything is working fine.  Boiler flow 48c into bottom left boiler return 36c out from top left.

 

UFH loop flows out from the floe meters on the top manifold 36c  UFH loop returns back into the bottom manifold 26c

 

 

What is the actual fault you are trying to correct?

[Sumo]

Sorry, I still do not understand. If my flow temp in is 48 but my flow through the loop is 36, where has the 12 degrees c gone? And if my return flow into the manifold is 26, how does the return flow out of the manifold reach 36 suddenly?

Also, not wanting to be difficult but how can we trust your picture when it is clearly labled in a contradictory manor.

If you look at your picture, on the right hand side, it says "manifold(flow) and on the left it says "flow to underfloor loops" pointing to the top section. With "manifold(return)" on the right and "flow from underfloor loops" written underneath, pointing at the bottom section of the manifold. 

Then, if you look at the detail you are relying on, which is written to the left, it states "boiler return" from the top section (already twice discribed as the flow) and for the bottom section it is "boiler flow"(already twice discribed as the return)

So the picture makes no sense. Sorry John, I have two heating engineers look over the system and neither has picked up on this as a fault, so I have to look closley for the solution.

[ProDave]

You don't need to understand the inner workings in detail.

 

UFH manifolds work by taking the incoming water from the boiler (or other heat source) and blending it with colder return water to give a controlled flow temperature into the UFH heating loops.  If you just wanted to pipe raw hot flow water from the boiler into the heating pipes you would not want the blending system incorporated in this manifold.

 

Most of them have the boiler flow and return on the pump / blending valve end so intuitively make more sense.  I had not see your type before to I searched and found that diagram that seems to indicate it is connected correctly.  Whether you or I understand what is going on inside does not matter.

 

Yes the annotation on that drawing is poor, but it seems to be doing what it is supposed to.

[JohnMo]

Certainly looking at the annotated photo, you have hot water going through the flow meters and return water at the bottom. How the mixing occurs doesn't matter as @ProDave says. It looks to be doing what it should be. With all UFH mixers there will be a higher temp going into the mixer than out to the UFH loops, 10 to 15 degs is pretty normal and that's what you have. The mixer takes some hot water from the boiler, adds some return water. So if you have 10l/min around the loops, you may take 6 to 8l/min from the boiler supply, the rest is reused return water.

 

If you want to remove the pump and mixer and direct couple, you would put the flow in at the top and flow out at the bottom.

 

Switching the pump of will give you zero flow, as the pump pulls the water through the mixer, the mixer acts as hydraulic sepereration.

[Temp]

20 minutes ago, Sumo said:

Sorry, I still do not understand. If my flow temp in is 48 but my flow through the loop is 36, where has the 12 degrees c gone? And if my return flow into the manifold is 26, how does the return flow out of the manifold reach 36 suddenly?

 

The boiler flow (48) is mixed with the UFH return (26) to produce the UFH flow (36). We can even work out what fraction of each the mixing is using to make the UFH flow..

 

48X + 26Y = 36(X + Y)

X + Y = 100%

 

Solving these simultaneous equations gives..

X = 45.5%

Y = 54.5%

 

So the 36C UFH flow is made up of 45.5% Boiler flow at 48C  and 54.5% UFH return at 26C.

 

[Sumo]

I think I know how and why mixers work, they seperate and mix flow and return dependant on target to protect the floor when you are running a system at a higher target temp to feed the rads than the ufh needs.

My Thermostatic valve is set wide open above 60 degrees, so with a flow in of 48 and return of 36, no mixing is needed or should be happening?

I checked the TV is opening and closing by listening to the flow changes/feeling resistance and also the pump tone changes.

I plan on maybe removing the tv blender and just having the pump or a new grunfos model3, on pressure setting, at the flow entry. I just want to block heat my slab on economy 7, with no restrictions.

 

How can you say it is working efficentley when actual flow temperature through the underfloor is a max of 36 with a delta 10 for a 26 return, yet the ashp is heating at a target of 48 and is prone to cycling off every seven minutes when it overshoots its target. If I set the ashp to a target of 37, I get 25 in to the loops, so it is not as though the blender is specificaly targeting the 36 I have now.

 

I think I should be getting say 46.5 in through the loop with the ashp at 48, with a delta between 5 and 7 across the floor, so returning to the ashp with a delta 10 max.

 

 

 

[JohnMo]

7 minutes ago, Sumo said:

so with a flow in of 48 and return of 36, no mixing is needed or should be happening

The way they are made there is alway flow mixing you cannot stop it. I tried two different makes before I realised. I have no mixer and no pump, just use circulation via the ASHP circulation pump.

[Sumo]

Hello Temp, thanks for joining in.

Yes, I had spotted the equation with in the numbers and I can see the theory. However I cant understand why any blending is needed. It is a bit like a false positive signal.

Surely with the pump to the right , it is pulling water from the manifold in flow point, and the assumed lower return plastic pipes, this is then sent up towards the return flow and the assumed in flow floor loop. My question is, would all the water not take the path of least resistance, through the big copper return pipe, that has two additional pumps pulling on it, rather than through the loops. It is as though the water through the loop is only there as it is been dragged away from the return by the mixing pump, as opposed to being sent and circulated through the system.

[John Carroll]

What way is the arrow on the manifold pump pointing, is it up or down?

[Sumo]

Hello John, the pump arrow is pointing Upwards.

So, the flow is entering via the lower "return manifold" where it is being "mixed" with the return's from the ufh loops. Then up through the pump, over the flow valves and out through the upper "flow manifold".

Complete nightmare.

Also the TMV is not a direct mixer valve, it is a single bored piece, with no direct mixing facility. The mixing being permenant, via the suction on the ufh return loops as mentioned, mixing the 48 celsius input with the 26 degree uf loop return's.

[Sumo]

On 26/09/2023 at 16:42, Temp said:

 

The boiler flow (48) is mixed with the UFH return (26) to produce the UFH flow (36). We can even work out what fraction of each the mixing is using to make the UFH flow..

 

48X + 26Y = 36(X + Y)

X + Y = 100%

 

Solving these simultaneous equations gives..

X = 45.5%

Y = 54.5%

 

So the 36C UFH flow is made up of 45.5% Boiler flow at 48C  and 54.5% UFH return at 26C.

 

Thank you Temp, you have summed it up well. My ASHP is only having 45.5% of 9l/min of water used.

So, with a theoretical flow of 15l/min for the whole system flow rate, in theory we are wasting 2/5th's plus 55.5% of 3/5th's. A very inefficent setup when you want to block heat your slab only on economy 7.

[Sumo]

On 29/09/2023 at 20:26, John Carroll said:

What way is the arrow on the manifold pump pointing, is it up or down?

Up.

[JohnMo]

On 26/09/2023 at 14:16, Sumo said:

I am fairly convinced I have a manifold flow direction issue

So how it works

 

Water comes in to the system as noted in @ProDave drawing, via a temperature regulating valve. It comes in via a regulating valve which is sensing the water temperature of the upper manifold - the hot water (from boiler is mixed the return water in the manifold. The pump pushes water from the lower to upper manifold, around the loops and back to the boiler. On the upper manifold water is restricted from going back to boiler by the lockshield valve. The lockshield valve balances what goes back to the boiler and what gets pushed through the UFH loops.

 

Your manifold is installed as it should be.

[John Carroll]

14 hours ago, Sumo said:

Hello John, the pump arrow is pointing Upwards.

So, the flow is entering via the lower "return manifold" where it is being "mixed" with the return's from the ufh loops. Then up through the pump, over the flow valves and out through the upper "flow manifold".

Complete nightmare.

Also the TMV is not a direct mixer valve, it is a single bored piece, with no direct mixing facility. The mixing being permenant, via the suction on the ufh return loops as mentioned, mixing the 48 celsius input with the 26 degree uf loop return's.

Yes, but achieves exactly the same objective, to give you the required hot manifold  UFH temperate, I know these controllers as tapstats. 

There is no loss of heat, you have a lower flowrate with a higher dT on the heat pump side and vice versa on the UFH side, both balance.

The HP will run with reduced efficiency if its required temp of 48C is required elsewhere like for HW production but your floor covering dictates a max water temp of 36C.

[Temp]

On 02/10/2023 at 18:42, Sumo said:

Thank you Temp, you have summed it up well. My ASHP is only having 45.5% of 9l/min of water used.

So, with a theoretical flow of 15l/min for the whole system flow rate, in theory we are wasting 2/5th's plus 55.5% of 3/5th's. A very inefficent setup when you want to block heat your slab only on economy 7.

 

I'm not quite sure how ASHP handles things but it's not normally wasted..

 

With a gas boiler the unused hot flow returns to the boiler raising the return temperature. The boiler modulates down to keep the flow temperature constant. So the gas used depends on the mixing ratio at the UFH manifold. 

 

On an oil boiler (these can't modulate the burner) the flow temperature rises and the boiler cycles the burner on and off. That can reduce efficiency but not much unless its short cycling a lot. We have a thermal store that feeds our UFH so the oil boiler runs flat out until the store is back upto temperature and then shuts off. This reduces short cycling but wastes some through heat loss to the room.

 

 

 

[Sumo]

On 02/10/2023 at 19:29, JohnMo said:

So how it works

 

Water comes in to the system as noted in @ProDave drawing, via a temperature regulating valve. It comes in via a regulating valve which is sensing the water temperature of the upper manifold - the hot water (from boiler is mixed the return water in the manifold. The pump pushes water from the lower to upper manifold, around the loops and back to the boiler. On the upper manifold water is restricted from going back to boiler by the lockshield valve. The lockshield valve balances what goes back to the boiler and what gets pushed through the UFH loops.

 

Your manifold is installed as it should be.

Thank you for your patience JohnMo. I get it now. Once I got my head around the fact that it was actually meant to be this way.  Uponor have recently been selling them, see below.

[Sumo]

On 03/10/2023 at 09:27, John Carroll said:

Yes, but achieves exactly the same objective, to give you the required hot manifold  UFH temperate, I know these controllers as tapstats. 

There is no loss of heat, you have a lower flowrate with a higher dT on the heat pump side and vice versa on the UFH side, both balance.

The HP will run with reduced efficiency if its required temp of 48C is required elsewhere like for HW production but your floor covering dictates a max water temp of 36C.

Thank you John Carroll.  I now understand a lot more and it is good to put a name to the tapstat.

So my primary-ashp circuit is flowing across the manifold with a delta T of a) at x) flow rate but my secondary ufh circuit has a smaller delta T of b) and faster flow y) through the manifold.

It is like the heat is being taken from one circuit to the other, as opposed to being given to the circuit. When pople have mentioned a figure of eight I could not conceptualize the flow's but if I see it as two slightly overlaping circles, with one trying to feed from the other it makes sense.

On the floor covering/flow temp debate, we have stone tiles and carpet on concrete, would I be right in thinking they could take a 48 degree flow if I could achieve it? If so, I am wondering if a different manifold setup with more control on the minimisation of the mixing could be more efficient. I am wanting to avoid short cycling, which we often have, as when combined with the defrost cycles it means we are not using much of the economy 7 power for heating. It is more cost effective to run with a higher flow temp for a short period than a lower rate for longer when the electric is 3 times cheaper.

 

[JohnMo]

3 minutes ago, Sumo said:

different manifold setup with more control on the minimisation of the mixing could be more efficient

Simple answer is they all mix flow and I tried a few before I dumped them all and now don't have a mixer. Flow out of ASHP equals flow to and through manifold.

 

 

[Sumo]

On 04/10/2023 at 12:34, Temp said:

 

I'm not quite sure how ASHP handles things but it's not normally wasted..

 

With a gas boiler the unused hot flow returns to the boiler raising the return temperature. The boiler modulates down to keep the flow temperature constant. So the gas used depends on the mixing ratio at the UFH manifold. 

 

On an oil boiler (these can't modulate the burner) the flow temperature rises and the boiler cycles the burner on and off. That can reduce efficiency but not much unless its short cycling a lot. We have a thermal store that feeds our UFH so the oil boiler runs flat out until the store is back upto temperature and then shuts off. This reduces short cycling but wastes some through heat loss to the room.

 

 

 

Hello Temp. Maybe wasted was the wrong word for me to use. I can now understand the heat transfer between the circuits, so there is no loss it is transfered heat, it just seems very inefficient to have an ASHP targetting 48 c when the secondary achieves 36-41 at best. I think it maybe best to forget/remove the mixing section and swap the flow and return primarys around and then put the pump on the primary flow?

 

I know there are saftey concerns when removing the mixer, from a double fault occuring,ie, the ashp goes awol and heats at 60 and not 48, and then the room stats also fail, so allowing constant heating, also that room stat failure would overide the 5 degree setback which limits the heating for 17 hours, and then an eldery relative falls unattend onto the floor and gets burned. It is possible but I could not guess the probability of those three events happening at once.

[JohnMo]

14 minutes ago, Sumo said:

saftey concerns

Easy way around that (if you ae concerned) is to have a pipe stat, that breaks the pump supply (normally closed) and if the temp goes to say 50 the stat contacts break to trip the pump

[Sumo]

1 hour ago, JohnMo said:

Easy way around that (if you ae concerned) is to have a pipe stat, that breaks the pump supply (normally closed) and if the temp goes to say 50 the stat contacts break to trip the pump

Thank you JohnMo, that sounds like an easy solution. I may have to treat the uf system to a new grundfos upsm3 along with the pipe stat.