Flow gauges for central heating pipe work

[John.g]

"If there are two flow gauges, one each side of the pump, they must both read the same; but what it reads will depend on which zones are open. You can't have 10 litres leaving the pump but only 8 litres entering. Where have the two litres gone to, only to magically reappear at the pump? "

But 10 litres is 10 l per min, so the water hasn't "disappeared" its just moving slower due to the restriction in the system?

For example some people get 20l per min from their kitchen tap, I only get 12l min due to my water main performance being worse than theirs.

So if there was a gauge at the pumping station it would read 20l min but my tap 10 miles away only 12l min?

NO, the gauge on the pumping station would read exactly 32 LPM. If you fitted flow meters at each end of your branched supply then both would read exactly 12 LPM and ditto on the 20 LPM branch. If you require 20 LPM without any piping alterations and if your water supplier was willing then because flow is proportional to the square of pressure, the pressure would have to be increased by a factor of ((20/12)^2)), 2.78 but then the 20 LPM line would flow ((20*sq.rt.2.78), 33.3 LPM and the pumping station flow meter will read 53.3 LPM.

 

[doitmyself]

But 10 litres is 10 l per min, so the water hasn't "disappeared" its just moving slower due to the restriction in the system?

It has if 10 lpm is leaving the pump but only 8 is returning. Where have the two litres gone and where have they come from? Does the pump create water?

For example some people get 20l per min from their kitchen tap, I only get 12l min due to my water main performance being worse than theirs.

So if there was a gauge at the pumping station it would read 20l min but my tap 10 miles away only 12l min?

But that's an open system; water is leaving your tap. Assuming you were on an exclusive pipe from the pumping station, if 20 litres is leaving the PS then 20 litres must be leaving the tap.

"Heating engineers do not understand how systems work" is an inflammatory statement, I'd question your motives in making such a statement.

I should have said: "Too many heating engineers do not appear to understand etc." Why do I need to have a motive?
You only have to read what is written by some "heating engineers" to come to the same conclusion.

I know how to balance a system, I was taught to obtain an 11 degree difference, now 20 degrees incidentally.

I'll have to take your word for that. But what do you do if you can't obtain 11C/20C? Do you know why you may not be able to obtain it?

But again the reality of domestic heating systems is they are often undersized, they are often tampered with and they are often poorly maintained.

What evidence do you have that they are undersized? There seems to be more cases of oversized systems on this and other heating forums. As for tampering and maintenance, I can't comment.

Better information about the system would lead to better solutions and greater understanding, which is what this post is about.

I tend to agree. A "heating engineer" should be more than a tradesman, which implies skilled manual labour. They need to be properly trained and qualified, by which I mean more than the current ACS and Gas Safe qualification.

 

[Jones82]

My experience of under sizing is usually the original system being modified, by adding an extension for example, but the heating system is just connected onto, anywhere they can, which results in half the house on a 15mm circuit, resulting in poor performing or even cold radiators.

If I couldn't get the temp diff then I'd use the lock shield valves, rads closest the pump lock shield almost closed, rads furthest away fully open. I'd have temp gauges on the pipes if I was being really keen but I'll be honest and say I don't need these most of the time.

This would change the frictional resistance of the system... which would affect the pressure / flow of the system?

How much would you pay per hour for one of these "qualified " engineers and which quals would you request?

I'm a qualified plumbing and heating engineer with nearly 20 years experience but I am by no means a physicist.... always willing to learn mind

Also my college days are a long time ago now... and I'm not even sure we discussed pumps and frictional resistance, to this degree, even then!

 

[John.g]

From your experiences do you find it more difficult to balance a oil fired system vs gas fired? as you can get a very steady/stable flow temperature from a gas fired boiler due to modulation but the oil fired flow temperature is always changing by the boiler stat hysteresis of 10/12C at a rate dependent on the boiler output power vs the required power.

 

[Rob Foster]

Granted, but I don't think any installer would mind having this sort of info available.

that kind of info is gold dust to some of the organisations we deal with...as I stated above mini bms are on their way
centralheatking

 

[doitmyself]

If you require 20 LPM without any piping alterations and if your water supplier was willing then because flow is proportional to the square of pressure,

NO. Flow rate is approximately proportional to the square root of the pressure difference between the two ends of the pipe. So, assuming the open end is at atmospheric pressure and the supply end is at 1 bar above atmospheric, increasing the supply to 2 bar above would only increase the flow rate by √2, i.e. 1.414 times, not by 4 times which you are suggesting.

 

[John.g]

NO. Flow rate is approximately proportional to the square root of the pressure difference between the two ends of the pipe. So, assuming the open end is at atmospheric pressure and the supply end is at 1 bar above atmospheric, increasing the supply to 2 bar above would only increase the flow rate by √2, i.e. 1.414 times, not by 4 times which you are suggesting.

Quite correct, I should have stated that pressure is proportional to the square of the flow, so as my figures suggested, if the supply end was 1 bar then to increase the flow from 12 LPM to 20 LPM would require a supply end pressure of 2.78, ((1*(20/12)^2.)

 

[chris watkins]

NO. Flow rate is approximately proportional to the square root of the pressure difference between the two ends of the pipe. So, assuming the open end is at atmospheric pressure and the supply end is at 1 bar above atmospheric, increasing the supply to 2 bar above would only increase the flow rate by √2, i.e. 1.414 times, not by 4 times which you are suggesting.

Quite correct, I should have stated that pressure is proportional to the square of the flow, so as my figures suggested, if the supply end was 1 bar then to increase the flow from 12 LPM to 20 LPM would require a supply end pressure of 2.78, ((1*(20/12)^2.)

And the practical application of all this higher level math is what may I ask ?
I love to learn but have to understand what it can do for me have you any examples for a lonely plumber or heating engineer, please?

 

[Jones82]

Just reading some interesting articles on pumps and understanding them further. It seems quite common, on an industrial scale, as a previous poster mentioned, to position gauges "on the suction" side of the pump and "on the discharge" side of the pump. I might buy a couple of low ranging gauges and experiment with them on my own system.

It does seem that a pressure gauge is the way to go.

Its odd that boilers measure flow rate, as we've mentioned, using electrical sensors, they then use this information to modulate accordingly. However an electronic flow sensor hooked up to a digital display doesn't seem to exist?

I had an ideal engineer out to a boiler recently. He hooked up his laptop and the flow sensor reading was displaying on his software.... So the technology exists but perhaps just isnt on the market?

 

[John.g]

I would venture that the ideal engineer was picking up that info on a App on his laptop linked to the circulating pump.
Grundfos have a Alpha 3 Model B 15-50/60 (6M pump) which you can link to your laptop/smart phone with the grundfos go App and from what I can see its a pandora's box of info/settings.
I don't know what boiler type/pump location you have but if its a externally mounted pump then you might consider installing one of these pumps or something much more down market (and cheaper) like my pump which is a Wilo Yonos Pico 1-6 (6M pump) which has plenty of settings but only then displays the power in watts while running but you can then derive the flow rate and the pump head from its pump curves quite easily.
I think you would learn far more from something like this rather than just installing two pressure gauges.
Incidentally, what make/model pump have you installed, it may already be a "smart" A rated pump which may already display the power.

Edit: Some more info here which may be of interest:
How to choose the correct speed control for heating systems - Lowara - International

 

[Jones82]

It was a grundfos 25-80, beast of a pump, think it was 20 litres a min on his laptop but the pump just had the usual 3 settings, it didn’t display watts.

Also I suppose it’s not really the performance of the pump I’m interested in. I’m interested in the performance of the system.

So how much force is the pump generating and how much is lost going round the various pipe runs.

I realise we could work this out on paper if we knew every bend and pipe run but in reality it’s either impossible or just too much grief!

So back to the original premise. Is there a way of measuring the frictional resistance of a heating system using gauges / flow meters / electronic sensors?

 

[doitmyself]

Some more info here which may be of interest:
How to choose the correct speed control for heating systems - Lowara - International

Very informative. The bit about constant pressure and underfloor heating confirms what I had read in some Grundfos literature - not that I can find it now. I guess that would also apply to microbore systems where the rads are fed, individually, from a manifold.

The other bit which struck me was the statement that current high efficiency pumps "are seriously lacking in sensory input. They can detect the system resistance and, if they have a built-in temperature sensor, they can detect system temperature." This means that all the "data" which is presented by the Grundfos Alpha 3 App is derived (from frequency, current etc?), not measured directly. I have read criticisms of the Alpha 3 that the data is not always accurate.

 

[John.g]

Yes, all these pumps derive the head and flow from the power (easily measured), the frequency (again easily measured) BUT they must also know the pump&motor efficiency at every point on every curve setting, which is literally hundreds of "numbers" as a lot of these pump's PP and CP settings can be increased from say 0.5M to 6M in 0.1M increments, I would love to know how this is done (efficiency calculations). I used my (oil) boiler rated output and the measured deltaT to calculate the flow rate and it compared within 10% of the derived flow from the Wilo power so I suppose fair enough for most purposes.

 

[John.g]

It was a grundfos 25-80, beast of a pump, think it was 20 litres a min on his laptop but the pump just had the usual 3 settings, it didn’t display watts.

Also I suppose it’s not really the performance of the pump I’m interested in. I’m interested in the performance of the system.

So how much force is the pump generating and how much is lost going round the various pipe runs.

I realise we could work this out on paper if we knew every bend and pipe run but in reality it’s either impossible or just too much grief!

So back to the original premise. Is there a way of measuring the frictional resistance of a heating system using gauges / flow meters / electronic sensors?

I'm afraid I can,t give a really meaningful answer to the above, apart from saying that it is physically possible to fit differential pressure and flow gauges to every loop in the system.

 

[Riley]

It’s just overkill though chaps. You are going to get joe public looking at this and thinking it’s rocket science to change a pump when it’s just not. Talk about baffling with science

 

[John.g]

The joe publics that change their own pumps will quickly learn that these pumps have the three traditional "fixed speed" settings, some, who are interested, will play around with the other settings. I suppose the vast majority of pumps are replaced by plumbers who will certainly be aware of these traditional settings and as time is money won't waste too much of their time playing around with the other settings.

 

[fixitflav]

I think the Alpha 3 displays both the pump head and the flow rate which can be very handy.

It's hard to see how you could measure flow at all accurately within the pump, unless Grundfos have come up with something clever. Easy enough to measure the pressure rise across the pump and use that to predict flow based on the pump curve, but that only works while the pump is performing as new.
For several years I had a minor system noise problem, but didn't suspect the pump. Then the pump itself started vibrating noisily, so I changed it, and the system noise went away. The old pump was badly clogged inside and behind the impeller, so performance down. A genuine flow reading would have indicated a problem, but one based on reduced measured DP and (new) pump curve would indicate high flow. A reading of DP would be useful, as along with estimated system curve would give an idea of actual pump performance.

 

[fixitflav]

Imagine you had a pressure / flow gauge on your outside tap as soon as it comes out the wall. Then imagine you have one on the end of the hose pipe after its run 20m down the garden. Surely the gauges would read different values?

The pressure at the end of the hose would be lower (if there is some flow) due to pipe friction, but the flowrate is the same for both (conservation of mass and incompressible fluid).

 

[John.g]

It was a grundfos 25-80, beast of a pump, think it was 20 litres a min on his laptop but the pump just had the usual 3 settings, it didn’t display watts.

Also I suppose it’s not really the performance of the pump I’m interested in. I’m interested in the performance of the system.

So how much force is the pump generating and how much is lost going round the various pipe runs.

I realise we could work this out on paper if we knew every bend and pipe run but in reality it’s either impossible or just too much grief!

So back to the original premise. Is there a way of measuring the frictional resistance of a heating system using gauges / flow meters / electronic sensors?

Bit of info here, page 36, not quite what you are after, but interesting all the same.
https://www.caleffi.com/sites/default/files/coll_attach_file/idronics_16_na_0.pdf

 

[Jones82]

An update on this; I recently installed an S plan at home. I fitted a 0-2 bar pressure gauge straight after the pump and another on the return near the boiler.

You can indeed see the pressure difference on the gauges when the central heating is running.

As the gauges are in the loft and the filling loop gauge downstairs;

I'm getting 29psi with the pump on setting 3 and 20 psi on the return, you can see the pump performance change as I change the speed setting and also the 9 psi loss across the heating circuit.

Interesting!

So if your on a job where you want to see the performance of the pump or the loss across the heating circuit, I'd recommend 0-2 bar gauges.

This also leads me on to the stupid 0-4 or 0-6 gauges that plumb centre sell. If normal operating pressure is 1-2 bar wth is with 6 bar gauges!!

Ps I got mine off ebay

Pressure Gauge 50mm 1/4 BSPT Vertical 0/15,30,60.100,150,300 PSI & Bar. | eBay

Optional Hose Tail adaptors to suit 1/4 BSPT Female x 6mm Tube Brass (Not included in Gauge Price). 50mm DIAL PRESSURE GAUGES.

www.ebay.co.uk

 

[John.g]

Have you allowed for the static height (if any) difference between two gauges as 9 psi or 6.2M seems to be a very high system loss??

 

[Ric2013]

Its not "extra water" or "less water" but loss of pressure or flow.

For example

Imagine you had a pressure / flow gauge on your outside tap as soon as it comes out the wall. Then imagine you have one on the end of the hose pipe after its run 20m down the garden. Surely the gauges would read different values?

Only if the hose is inflating like a balloon or you remove the hose.

Thought experiment: you have a tap with incorporated flow gauge running at 15l/m and then you add a hose. Flow drops to 10l/m. A gauge at the end of the hose must read the same.

Otherwise you have 15l/m going into the hose and only 10l/m leaving the hose, leaving 5 litres of water every minute that enters the hose but does not leave it. So where is that 5 litres of 'extra' water going? If it is going in one end, it must go out the other end.

 

[Jones82]

your measuring mas

Only if the hose is inflating like a balloon or you remove the hose.

Thought experiment: you have a tap with incorporated flow gauge running at 15l/m and then you add a hose. Flow drops to 10l/m. A gauge at the end of the hose must read the same.

Otherwise you have 15l/m going into the hose and only 10l/m leaving the hose, leaving 5 litres of water every minute that enters the hose but does not leave it. So where is that 5 litres of 'extra' water going? If it is going in one end, it must go out the other end.
[/QU

Only if the hose is inflating like a balloon or you remove the hose.

Thought experiment: you have a tap with incorporated flow gauge running at 15l/m and then you add a hose. Flow drops to 10l/m. A gauge at the end of the hose must read the same.

Otherwise you have 15l/m going into the hose and only 10l/m leaving the hose, leaving 5 litres of water every minute that enters the hose but does not leave it. So where is that 5 litres of 'extra' water going? If it is going in one end, it must go out the other end.

In litres per min we’re measuring mass over time, the 5 litres hasn’t disappeared. The water is moving slower due to increased frictional resistance.... I suppose it’s still in the pipe?

How about if a hose pipe is 50m long? Would you still get the same performance as if it was 1m long? No, the pressure and flow rate would drop the greater the length, or frictional resistance of the pipe. A gauge immediately out of the tap would give different readings than one on the end of the run.

For example. Yorkshire water have a pumping station, the pump is new and the pipes are new, the gauge near the pump reads 3 bar and a flow cup gives 20 litres a min. Joe blogs has a house 20 miles away. His mains water pipes are ancient and all furred up, he’s getting 8 litres per min from his kitchen tap due to the restriction in the pipes..... the water is there it’s just moving slower

The two gauges on my pipe work are close to the boiler, literally on the flow and return, on a heat only sealed system. The flow gauge is 420mm higher than the return gauge and is located on the positive side of the pump. I turned the pump down to setting 2 as it was a little noisy. I’m getting around a 5-6psi difference across all the pipe work, rads and unvented cylinder. It’s a 2 bed semi, boiler in the loft.

rads are

1000x600 k1 x2
600x600 k1 x2
1600x600 p+
1800x500 k2

 

[Ric2013]

In litres per min we’re measuring mass over time, the 5 litres hasn’t disappeared. The water is moving slower due to increased frictional resistance.. I suppose it’s still in the pipe?

Okay, so after 2 minutes, we have an extra 10 litres in the pipe; after 10 minutes, another 100 litres in the pipe? In my thought experiment I can't see it happening.
If you've actually carried out the experiment in real life with two flow gauges and I'm wrong, then please let me know.

 

[Jones82]

Back to the original premise!

Lets say you’ve just installed a new boiler. Is the pump working? Is it circulating around the byepass or the whole heating circuit? With a gauge on the flow and return you can see the real time pressure loss across the circuit.If it is pumping around the byepass or just the coil then the difference between the gauges will be next to nil. If it’s pumping around the entire circuit you will see a larger difference