Honeywell v4073a valve noise

[DirksPlumbing]

Who said anything about squeezing water through pipe? The purpose of balancing is to get the correct flow through every radiator. The fact that each radiator is fed by a separate pipe helps but a 500W rad still needs half the flow rate of a 1kW rad.

I doubt if Drayton and Honeywell would agree with you.

Water has roughly 1.163 Wh/kg/K and 1kg/l.

At DeltaT 10K that is 11.63 Wh/l.

That gives you 43 l/h.

DN6 has 0.028 l/m. Resulting in 0.42 m/s.

But to KISS it: 500W / (11Wh/l *3600) = 0.0126 l/s.

My flow resistance table goes up to 0.025 kg/s for DN6.

Meaning your comparably tiny rad of 6x6 K2 will result in > 0.566 m head per meter.
At a 5m distance from your manifold this is > 5.66m head.

Or in other words, to use this radiator as per design it needs a separate Alpha2L 15/50 for itself.

I am not saying it does not get heat as the resistance per meter drops quickly. Are you sure that Drayton and Honeywell will tell me that physics are not valid for plumbers?

As for the balancing I have to apologise. Grundfos Ecademy does not provide these courses for UK.

 

[DirksPlumbing]

It was Danfoss E-Learning. http://learning.danfoss.com/NR/rdon...earningHeatingCourseCatalogue_UK_05082014.pdf

 

[DirksPlumbing]

As for the correct flow on a microbore system you will end up with a great mixture of temperature spreads as you cannot get more than the pipes will deliver. How that could be correct, I do not know. But it definitely will not be balanced.

Glow-worm specifies a maximum spread of 20K in some manuals. On lower temperature specs (higher efficiencies) you usually design the 10K only anyway.

Gone are the days of 90/70/20 for most gas appliances.

 

[Bennachie]

So, Dirks, what setting should my 'intelligent' pump be on? The boiler requires 780 l/hr min if I remember correctly. This is all doing my head in.
Dodgy valve emits a dull, damped, rubber thunk even when no pump running and just the spring return to DHW only.........

 

[DirksPlumbing]

I have noticed I had mixed up DN6 calculations and 6mm copper pipe resistances.

What the pump should be set to? Honestly I do not know. But typically the variable pressure modes do not work all to well on microbore systems. Is your pump wired into the boiler pump port? There are so many things I do not know about this system, how could I be able to say this or that?

It is indeed mysterious how this valve can cause so much noise even when the pump does not even run. Change the valve, fit the automatic bypass and a balancing valve and lets see.

 

[DirksPlumbing]

Grundfos Product Center

This is roughly the same curve as on the Alpha 2L 15/60.

 

[DirksPlumbing]

As you can see, the high constant pressure setting limits to 4.5m head and the lower constant pressure setting to 3m head.

 

[Bennachie]

Thanks for your time....and sage words. It is appreciated.
I shall go back to basics and work everything out.
The valve will get changed soon and the bypass will go in.
I'll let you know.

It is weird that it thunks even just on spring pressure. I have a theory about it but it is just that, concerning the seating of the ball within the valve. I think the newer ones contact all of the seat immediately whereas the older ones contact the periphery initially then settle onto the seat, thus giving a 'gentler' shut off.

Trawling the net has brought up a number of similar threads concerning Honeywell valves in particular making this noise, so..........................hey ho.

Constant pressure setting is ok for an open vent system I assume
Thanks again.

 

[doitmyself]

I have noticed I had mixed up DN6 calculations and 6mm copper pipe resistances.

Can't see what relevance DN6 and 6mm has? The OP has 8mm pipe -presumably copper considering the age of the installation - approx 28 years.

A 500W rad with a 10K differential will require 0.012 kg/sec so 8mm pipe will have a resistance of approx 0.033 metres/metre length, i.e, 0.3m for a rad 5m from the manifold. The water velocity would be only 0.3 metres/sec, which is just about acceptable if sludge deposits are to be avoided.

 

[DirksPlumbing]

Can't see what relevance DN6 and 6mm has? The OP has 8mm pipe -presumably copper considering the age of the installation - approx 28 years.

A 500W rad with a 10K differential will require 0.012 kg/sec so 8mm pipe will have a resistance of approx 0.033 metres/metre length, i.e, 0.3m for a rad 5m from the manifold. The water velocity would be only 0.3 metres/sec, which is just about acceptable if sludge deposits are to be avoided.

Okay, given a property of 28 years with 3 bedrooms, one bath, one living room, a hall and a utility. That's 7 to 8 rads. How many of them are likely to be 6x4 P+?

Thinking about 18kW (the boiler) then that would be an interesting sight of ~30 radiators of 500W throughout the house 😉.

DN6 is "diamètre nominal" and 8mm copper pipe has 6.4 mm and is very close to DN6. By latest once you got an 0.2mm oxide layer inside it definitely is.
And the 10mm plastic pipes are playing in the same league if at all.

On balancing your worst radiator determines the settings for all others (logically as you can only throttle). This may be your 6x16 K2 at ~2800W with 0.068 l/sec and roughly 12m microbore 8mm. That is a gently 2.1 m/sec and therefore a good speed well under 2.5 m/sec.
Looking in the table the pipe resistance is: oops , we do not know as it's outwith the specs in accordance to the table. But lets assume 0.4 meter head per meter x 12 its 4.8m head.
Now the boiler requires a minimum flow of 773.86 l/min. Adding 1.4 meter head for the boiler as per MI and another meter for the rest of the pipework and fittings we are now at 7.2 meter head loss at ~774 l/min. You need a Magna 25/80 to allow that system to run half decent once it's balanced.

I am sure the customer will be totally happy about the savings he is going to acquire after the pump and installation cost (~1k?) and the rise in electricity costs. Grundfos estimates "only" 68.2 W for the required setting.

And after a while the customer will not even notice the valve noises anymore as he got used to it. Hopefully. Would a pump like this push 22mm 2port or diverter valves open? But hey you won't need an automatic bypass then.

 

[DirksPlumbing]

[snip]
Constant pressure setting is ok for an open vent system I assume
Thanks again.

When fitting an automatic bypass Grundfos used to insist in using the constant pressure settings. The variable settings alter the pressure and make it almost impossible to find a suitable setting for the bypass.

I find this "new" requirement to fit an automatic bypass, no matter what the system design is, contradictive to the attempt to save energy as it prevents the use of the most energy saving settings on the pump. Though in many cases it can help to lower the return temperature and therefore increase the efficiency.

 

[DirksPlumbing]

Regarding the design quality of the newer Honeywell gear I made similar observations. Before 2013 I had barely any replacements. Within the last year I had to replace a lot more and all were 2013 and after.
I started looking at the Horstmann gear but can not say to much about the longevity. Though their (2/)3/4 channel timers are without alternative and I had no trouble with them yet.

 

[Bennachie]

Thanks Dirksplumbing. All very helpful, informative and interesting.

Some I do not understand though. How does one arrive at the velocity figure for the speed of water in the pipe?
I understand the resistance calculations but the speed escapes me.
Also when working out resistance do you add figures for fittings and rad valves? I assume they were left out of the last couple of posts for simplicity.

Thanks again for your time.

 

[DirksPlumbing]

You have to apologise if I am not using the best terminology as I am not a native english speaker.

That is simple maths. Pipes are round (or are supposed to be) and consist of some type of material or mixture of materials. This material has a certain thickness (the wall thickness). Lets take a 15mm copper pipe. It has 1mm wall thickness all the way round. This makes the internal diameter 13mm.

Millimeter is not very handy for calculating volumes as we are typically using liters. Decimeter makes more sense as 1dm * 1dm * 1dm = 1dm^3 = 1l.
One decimeter has 100 millimeter. Our 15mm copper pipe has an internal diameter of 0.13dm.

The surface of a circle is pi / 4 * diameter^2. Thus gives us roughly 0.01327dm^2 for our copper pipe. 1 meter has 10 decimeter.
That gives us 0.1327dm^3 or 0.1327 liter / meter pipe.

If I want to pass 0.1327 liters per second through this pipe then the water has moved exactly 1 meter and as this happened in one second we know now that the water has been moving that one meter in one second (1 meter per second).

Do I want to pass 0.2654 liters per second (double figure) through the pipe then the speed doubles as well (2 meter per second).

x liter/second / y liter/meter = v in meter/second

Yes in the previous calculation I just threw a meter of resistance in for the rest of the fittings and pipework. As you would have needed to adjust the lock shield (no manufacturer I know manufactures presettable valves that would allow this kind of adjustment) hence you can forget about the lockshield as it is to be set to what ever additional resistance will be needed.
Except of course for your worst radiator, there it needs calculated as fully open.

 

[Bennachie]

Thanks. No apologies require re native language. Your English grammar and sentence construction are fine; and significantly better than my flimsy grasp of 'foreign' languages, which is limited to spoken (poorly spoken) French and a little Italian and the ability to read French.

 

[doitmyself]

On balancing your worst radiator determines the settings for all others (logically as you can only throttle). This may be your 6x16 K2 at ~2800W with 0.068 l/sec and roughly 12m microbore 8mm. That is a gently 2.1 m/sec and therefore a good speed well under 2.5 m/sec.

But even the index rad may need to be throttled to give the required temperature drop if the pump can't be set to do this.

2.5 m/s seems excessive, most books suggest a maximum of 1.5 m/s and preferably only 1 m/s.

Presumably the calculation used to find the index rad for a manifold based system is exactly the same as for a non manifold system.

 

[DirksPlumbing]

But even the index rad may need to be throttled to give the required temperature drop if the pump can't be set to do this.

2.5 m/s seems excessive, most books suggest a maximum of 1.5 m/s and preferably only 1 m/s.

Presumably the calculation used to find the index rad for a manifold based system is exactly the same as for a non manifold system.

In terms of a heating system you are correct. 2.5 m/s is excessive.
Copper is usually specified for 0.5 - 2.5 m/s. But we are looking at this layout at a full load situation.
Once some radiators start shutting down the pressure may increase depending on the pump / bypass valve settings.
Then the flow rates go up and your previous 2+ m/s quickly gets beyond the 2.5 m/s. Noises, water hammers ...
Though I think that these DN6 and below pipes are usually somewhat restricting you never get much over 2.5 m/s anyway.

Yes there can be situations where you want to throttle the index rad. But with the availability of different pipe systems and various pumps with a wide range of settings this becomes rare in domestic situations. In commercial applications you would rather limit the stream via differential pressure control valves very granular.

Mind that most presettable valves only have 4 - 8 settings anyway. And apparently that seems quite sufficient. I still stand to my opinion that in a microbore system I rather live with the limitations instead of even further reducing the flow and fitting a commercial pump. That does not stand for mixed installations.

I do not fit microbore systems but I admit the use of 10mm pipe in existing microbore installations otherwise you bring the system in quite an imbalance.

Manifold systems still have their place (underfloor heating for example). If there were half decent priced normal bore manifolds available at the merchants I possibly would do normal bore manifold systems with MLCP. No joints under the floor. Especially with pipes run in concrete floorings within a corrugated conduit you just pull the old pipe out and the new one in (if not to long or to sharp bends or crushed -> workmanship). Though you need a place for the manifold in a cupboard or so.

 

[DirksPlumbing]

Ah, "rather rare" is maybe not so right. One thing that drives me mad is that there are still a lot heat only boilers sold that do not have the ability to differentiate between a CH demand and a HW demand. Then you have to set up the flow temperature in order to satisfy the cylinder rather the heating system.

I wonder how that could slip through ErP regulations.

 

[Bennachie]

For those interested I have refitted my 28 year old HEAD unit to my new valve.....(interim as I am busy and do not want to drain down in the cold weather) ............. and the noise whilst not totally erdicated IS much reduced. So for all us conspiracy theorists out there Honeywell HAVE hanged something ( or their tolerances are not what they once were.......)

 

[doitmyself]

In terms of a heating system you are correct. 2.5 m/s is excessive.
Copper is usually specified for 0.5 - 2.5 m/s. But we are looking at this layout at a full load situation.
Once some radiators start shutting down the pressure may increase depending on the pump / bypass valve settings.
Then the flow rates go up and your previous 2+ m/s quickly gets beyond the 2.5 m/s.

But, with most pumps, if the pressure increases then the flow rate, and consequently the velocity, reduces.