Measuring pressure vs flow to balance manifolds

[Aunt Maud]

I would like to balance a couple of manifolds for a manifold fed radiator system.

It goes like this:

From the boiler the flow goes to manifold 1 which splits it in two and each goes to its own zone manifold 2 & 3 which distributes hot water for the radiators. The return is a mirror image of the flow and a pump on the return sends it all off back to the boiler.

As you can imagine I'm having fun balancing this thing, as there are no flow meters or gauges on any of the manifolds, so I would like to add two gauges to help me balance the flow.

I was thinking of adding pressure dial gauges to the zone return or supply manifolds 2 & 3 and a valve on each to balance them out. Dial gauges would be my first choice, as they don't interrupt the flow of water in the manifold and are easy to fit.

Would measuring pressure at the manifold instead of flow work ?

[PaulF]

Based on my (dimly remembered) Chemical Engineering Fluid Mechanics course I'd say only if both legs were identical: same diameter, length and geometry. For example a narrower pipe will have a higher pressure for the same flow rate. So it might give you an approximation but I suppose it depends on the dgree of precision that you require.

At least I think that's right...

[Aunt Maud]

The manifolds on zone 2&3 are similar in configuration, but 3 is a lot further from the pump than 2.

I was thinking that if I placed the gauge on each return manifold, as the temperature was low enough for the dial gauge, I could get a rough estimate to the state of the flow for comparison.

[rafletcher]

3 will have a greater resistance than 2, so will need a higher delivery pressure to have the same flow rate. Ergo the inlet to 2 needs to be restricted to send flow to 3. I’d probably fit a needle valve in the delivery line to 2 and do it empirically - feel when first radiator gets hot - meaning higher flow to 3 if simultaneous - then back off the needle valve a bit.

[matthew]

If you balance the pressures at the two manifolds then that is all you have done the flow will now be balanced according to the frictional losses through the pipework, most of the flow will go through the larger, shorter pipework as the overall pressure drop has to be the same at each end of the supply and return to the common pipework at the pump.

Effectively you are trying to use the valves to equalise the pressure drop through each route so that the flows are equalised. However because the flows recombine the system will naturally equalise this by varying the flow. The only way to test this will be to isolate manifold 2 and measure the pressure difference between the distribution and return manifolds through Manifold 3. The isolate manifold 3 and measure the losses through manifold 2 (at the same flow), use any valves to adjust the pressure drop to match that of manifold 3 and the flow split between 2 and 3 should balance. If you want to do the same within manifold 3 and manifold 2 then you will have to individually check the routes within each manifold

Code: [Select]

     
             v1                 
           ----------------------------
        p3|   v2                       |
  p1      |----------------------------|p4
    ------|                            |---------
   |  v4  |____________________________|         |
   |              v3                             |
---|                                             |--------
   |           v6                                |
   |       ----------------------------          |p2
   |      |       v7                   |         |
   | v5   |----------------------------|         |
    ------|                            |---------
       p5 |____________________________| p6       
                    v8


So either you test between p1 and p2 to confirm the two zones are balanced using valves v4 &v5 or you can test between p3 and p4 then p5 and p6 to confirm each manifold is balanced using valves 1-3 & 6-8   before then conforming that 1 and 2 balances the the two manifolds

[matthew]

Because the test will have to be full pump flow through one manifold and then through the other when the flow is split it won't be exactly equal because of the different effects of pipe length and fittings but it will be fairly close.

Should there then be thermostats in each leg on each manifold then these will further unbalance the flows but that should also be fairly minor.

What sort of valves do you plan to use for the flow balancing as gate and butterfly valves are not as good as globe valves.

[Aunt Maud]

I knew it was worth asking here.

I'm not sure what kind of valve as yet, but I have looked at 3/4" BSP needle valves and they are expensive, the manifolds are 3/4" brass. I have full flow ball valves on each manifold, which I use to isolate them when I add a radiator, but they get turned on and off at regular intervals, so I was hoping to fit a balancing valve or two so I didn't have to re-balance each zone when I added a new radiator.

Reading what you have all posted, it looks like I'll have to balance it every time I add a new radiator to either zone anyway, so I thought a pressure gauge on each manifold might help speed things up, as I end up chasing my tail a bit every time I do it.

I like dial gauges too, so I'd like to pander to my analogue fetish for weirdo kicks.

[matthew]

Needle or diaphragm valves would be the normal flow balancing valves provided you can get a diaphragm that accommodates the temperature. As you say they are pricey as only the most technical or sensitive systems would go to this level of detail.

Yes if you add more radiators to one of the manifolds you have just added another flow path so the system would need to be rebalanced, though not within the other manifold just the modified manifold and the two manifolds. Personally I would just try to ensure that the flow to each of your zonal manifolds is even as any thermostatic valves within the manifolds will juggle the flow within the manifolds most of the time anyway.

[rafletcher]

What sort of valves do you plan to use for the flow balancing as gate and butterfly valves are not as good as globe valves.

Globe valves really are expensive. As you say, needle, or diaphragm will be ok.  AM - the reason you would use either of that type is they are much more proportional. Gate or butterfly (cheap) won’t modulate well, 10% open and you’ll have 90% flow.

[Aunt Maud]

Do you think I'd get away with one Needle valve on the supply of manifold 2, as there's one on Fleabay ?

https://www.ebay.co.uk/itm/Swagelok-SS-4GUF8-G-3660psi-252Bar-3-4-Stainless-Steel-Needle-Valve/112617930999?epid=2075317554&hash=item1a388d40f7:g:8bAAAOSwPC5ZySoY

[rafletcher]

I think so, as per my first post. That valve is certainly good enough.

[matthew]

Provided you are confident that there is less flow to manifold 3 than manifold 2 then installing a valve on the return side of manifold 2 before it recombines with the return from manifold 3 will be sufficient to add the pressure drop to the manifold 2 route and even the flow between the two zones.

[Aunt Maud]

Hmmm, well, that's the thing.

I intend to add at least another four radiators to each manifold, so each manifold supplies 7 radiators. At the moment three radiators on each manifold are balancing well and providing sufficient heat, with a guess it/empirical type balancing procedure.

I need to dance about for a few hours to get it to work though and I'm getting fed up with going up and down the ladder (no stairs) and wearing a groove in the floor.

He wants £60 posted for the valve, BTW.

[Aunt Maud]

Back to the manifold fun.

I have the low energy secondary pump on the return to the boiler, just after manifold 2.

If I want to throttle the flow at manifold 2, to increase it to manifold 3, would it be better to have the pump on the flow side, or doesn't it matter ?

Also, in a manifold does the last port in line have a better flow than the first, or are they all the same ?

[matthew]

If the manifold is set up as:

Code: [Select]

__>_____________________________________________________
        |       |       |       |       |       |       |



Then the last off take is likely to see marginally more flow than the others. This is due to the end effect and the propensity for water flow straight rather than turn down the branch till it reaches the corner on the end. It will be worst if the end connection is straight on through a reducer from 25mm to 15mm.


If the manifold is like this:

 

Code: [Select]


                           |
                           \/
        ________________________________________________
        |       |       |       |       |       |       |



Then the center ones may see more flow if they are in close proximity to the supply and again the ends will see more flow.

Each time you add an extra radiator to a manifold you will increase the available flow paths and it will upset the balance between the two manifolds. However it does seem that you are achieving sufficient balance within the manifolds through your current method.


Regarding the pump, if you put it on the flow side rather than the return the only difference is the pressure in the radiators. The valve can be positioned anywhere in the supply to, or return from, manifold 2 wherever is most accessible.

[The French Tandem]

Would measuring pressure at the manifold instead of flow work ?

Balancing pressure or flow in a central heating system is totally irrelevant. What you really want is to balance the temperature in every room of your house. What I do is :
1)  I set all the valves half open, 2) I run the boiler for a couple of hours, and 3) If a room is too cold, I open up the valve slightly, or close the valve if it's too warm. What else could someone wants?

A

[Diver300]

Would measuring pressure at the manifold instead of flow work ?

Balancing pressure or flow in a central heating system is totally irrelevant. What you really want is to balance the temperature in every room of your house. What I do is :
1)  I set all the valves half open, 2) I run the boiler for a couple of hours, and 3) If a room is too cold, I open up the valve slightly, or close the valve if it's too warm. What else could someone wants?

A

I agree that pressure and flow balancing is irrelevant. However I would say it's optimistic to assume that a single setting for each valve could keep the temperatures balanced over a range of conditions.

My suggestion would be the Honeywell Evohome system.

[Aunt Maud]

I agree that pressure and flow balancing is irrelevant. However I would say it's optimistic to assume that a single setting for each valve could keep the temperatures balanced over a range of conditions.

My suggestion would be the Honeywell Evohome system.

Very nice, but 14 radiators it would nearly double the cost of the whole heating system. Unfortunately I don't have that money to spend on it, as the budget for the whole house renovation is very small.

I'll try fettling the valves first, as manifold 2 is hogging the flow.

[Brucey]

this may be b-obvious but if you have TRVs then any attempt to balance the flow between manifolds is pretty futile (or at least will only apply when the system is cold); as soon as the TRVs start to work the pressure drop in each manifold will start to vary and so will the flows.

If you fit a balancing valve then (providing it does not reduce the flow to a very low level)  it will arguably just alter the time it takes in each branch before the various rooms reach the desired temperature. If the flows are always very uneven from cold then it may make sense to have separate zoning valves and to heat each zone separately (at different times separated by about 15 minutes) from cold. Once the zones are hot the TRVs kick in and the flow imbalance will (usually) be reduced.

BTW you can have (almost) fully independent control of radiators with TRVs using a cheap mains timer and a few other bits and pieces. The method is to use the timer to control a small transformer that drives a low power resistor that is set to heat the TRV capsule directly.  Thus if you want a specific radiator to be 'blocked off' at times when the rest of that zone is 'hot' all you need to do is to set the timer so that the TRV is heated electrically at this time  instead. This typically consumes less than 5W of electricity for as long as the chosen radiator is 'blocked'.

cheers

[Aunt Maud]

That's interesting Brucey, I was thinking that I might get a result using the lock shield valves on the radiators.

Anyway, I'm quite pleased with my little system. Mrs AM was able to bleed the air out of the whole thing in two minutes, and she doesn't do plumbing.

[the snail]

I was thinking that I might get a result using the lock shield valves on the radiators.

That's how it's normally done (if at all) - take the heads off the trvs, so that the valves are open, use thermometers to measure the temperature drop across each rad, use the lockshield valves to equalise temperature drops.