here's a good one for you!!!!!

Belgiumboy - can you give some more info on the pump model - the UPS 200 covers a number of different pumps. If you can give the actual one I will look up the curves.

This does sound like cavitation as the others have pointed out. This can really damage a pump in fairly short time. To get around it there are a few things you can try:
1.) try reducing temperature although there are limits - if your boilers aren't designed to run in condensing mode they won't last long at low temperature operation.
2.) Reduce pump speed - all UPS have a three speed selector, make sure yours is running at the slowest.
3.) reduce resistance at the pump negative pressure side - make sure the inlet valves are fully open etc,
4.) increase resistance on the pump delivery side - daft as it may seem, if you can partially close a valve downstream of the pump the inlet absolute pressure will rise and may stop the cavitation.
5.) The expensive option - fit an inverter to slow the pump down.
6.) Change the pump - to cope with the full available output of the two 53 kW boilers the pump will need to move about 2.3l/s against your total system resistance.

Hope this is some help.

Oh one other thing if it's proved to be cavitation whats the cure if any???

PTH: some info for you:

Pump UPS40 - 60/2F Model C
Pump is set on slowest speed
Not a condensing boiler they're about 20 years old
all inlet valves are fully open size 1.5" bsp
will try to partically close the valve on the posative side of the pump later today
only running one boiler as this is quite enough for the system, when originally designed they went right over the top on spec

Although a bit on the big side, the pump could be about right for your system, but only if both boilers are running - the 2.3l/s would match a system resistance of about 240 kPa, which is quite a lot (260 kPa if the pump is 3 phase).

You really need to work out the system resistance to select the right pump, typically for a simple arrangement without too many control valves or miles, sorry kilometres of pipework, I would expect something like 80 - 100 kPa., which would put your pump well into cavitation territory.

If you are successful in stopping the cavitation by throttling the downstream valves, then you may be trading one noise for another - it's possible that you will get cavitation at the valve, which could destroy it's innards. Valves are cheaper than pumps though.

The best, but most expensive solution is to change the system - with two boilers, if you don't want to run them together, ever, you could simply change the pump for one better suited to the system resistance and the boiler rating (in your case 1.15 l/s). If you may want to run both boilers, each should have it's own shunt pump onto a primary circuit with your distribution circuit driven by a further pump connected to it.

Good luck.

PTH:

Just spoken to Grunfos and they say the pump is too big for just running a 53Kw boiler, they are recommending the UPS32-55 pump giving 1.15L/s as opposed to the one i have now which is 2.7L/s, would like you thoughts on this before i go running out to buy a new pump

Thanks for your help so far

I don't have the 32-55 on my Grundfoss software but that may be because I'm a couple of years out of date. I do have the 32-40 and the 32-60, so I assume that the one they are recommending is between the two.

The flow rate is certainly right, and the developed pressure against resistance is much more in the league of simple systems.

You will have to modify the pipework slightly as your existing pump has 40mm connections and the new one has 32mm, shouldn't be too much of a pain though.

If you have the boiler manual you should be able to check the minimum flow rate that they can withstand, if this is more than 0.575 l/s you will only be able to run one boiler safely so you may like to add in an electrical changeover switch to prevent both being switched on by mistake.

Only running one boiler, the other is disconnected from gas and electric and the rest of the system, haven't got the boiler manuals as they are quite old, just seen the price of a new pump ehhhhhhhh, so i need to know that there is a fair chance of this being the problem before replacing the pump, seems logical to me but then i'm no heating engineer. so just to recap:

53Kw boiler
15 old style cast iron radiators
NO hot water heating
all rads have TRV's
main curcuit 2"bsp running to 1" bsp to rads
system pressure 1.2bar
house has 8 bedrooms 3 receptions + hallway and bathroom

Whats your thoughtsnow
much appreciated

belgiumboy said:

Onetap: what you say makes sense to me but there are some differences, layout: Boiler, pump, cold feed (top up), expansion tank.

Click to expand...

As I understand it, you have a sealed system. The "cold feed (top up)" is the connection for the fill point from the cold water mains, which would have the valve closed in normal use; the "expansion tank" is a diaphragm expansion vessel, a cylindrical red (probably) type thing.
Is that right?

The best, quick fix would be to connect the expansion vessel before the pump; before the boiler would be better than where it is now. There's only a small flow rate to/from the expansion vessel, so 15mm (or whatever the Belgians use) is plenty & convenient.

Reasons are as follows. The pump generates a pressure difference (dP), low pressure at inlet and high pressure at outlet. PTH says about 100kPa, = 1 bar, which sounds likely & is a nice round number. The actual flow rate and differential pressure depende on the pump and the system.

The actual static pressure ( what a pressure gauge would show) depends on where the expansion vessel is. So you have about 1.2 bar at pump outlet; pump dP= 1 bar, pressure at pump inlet is 0.2 bar gauge. Not good.

You put expansion vessel connection at pump inlet. You have about 1.2 bar at pump inlet; pump dP= 1 bar, pressure at pump outlet is 2.2 bar gauge. Much better.

Is the pressure gauge still good? Does it return to zero if disconnected or the system is drained? Cheap ones are not reliable.

If the problem is cavitation the new pump will fix it.

You have to be sure that you have eliminated all the other possible causes of noise. If you are confident of that then just to prove it try the throttling technique - you may need to close a downstream valve almost fully (but not completely) to get much effect, but you should hear the noise from the pump diminish and more noise starting to come from the valve you are adjusting, this will prove cavitation.

Cavitation noise is usually quite distinctive, it can be almost metalic; as if there were a handfull of ball bearing rattling around in the pump body. It is also quite destructive, when you take the pump out you will probably see lots of very small pits around the outer edge of the impeller, in extreme cases it will look like mice have been eating the outer edge. Your pump shouldn't be too bad as it has a stainless steel impeller, but there will be some damage, even if you need a magnifying glass to see it.

Keep us informed.

Just throttled back the pump outlet valve to halfway if i close it anymore the noise gets louder, the metallic noise has quietened but not gone, i have reduced the boiler temp to 65c hoping that will help slightly, pressure gauge is good as it drops when bleeding and rises when filling so i presume it's good.

Well, you've proved that there was cavitation if the metallic noise reduces. The rest of the noise is general water noise - 2.7 l/s into a 40mm pipe will be in excess of 2m/s - I would normally design to 1.25m/s max.

You could also look at how the pump is supported as this will have a substantial effect on noise. It should be supported independantly of the pipework and ideally through some anti vibration mountings. There should be a flexible connection either side of the pump to reduce the transmission of vibration into the pipework. A sure sign of this will be that the noise will reduce if you lift the pump body while it is running.

When you put your new pump in make sure that it is properly bracketed - there will be a diagram in the installlation guide.

--------------

Onetap suggests moving the expansion vessel but there will be limited benefit from this. The vessel only acts to exert a static pressure on the whole system and to allow the water to expand as it heats up.

It cannot be compared to the location of the cold feed and open vent connections in an atmospheric system

The position that it connects to the system is not important although a connection too close to the discharge of the pump will result in pump hydraulic pressure being absorbed by the diaphragm. By all means try moving its connection before buying a new pump but I don't hold out much hope for a significant reduction.

You could look at other manufacturers if the Grundfoss is too expensive, you won't get much cheaper but you could try Wilo or Lowara, both sell in Belgium.

PTH said:

The position that it connects to the system is not important although a connection too close to the discharge of the pump will result in pump hydraulic pressure being absorbed by the diaphragm.

Click to expand...

Wrong and wrong.

If any of you think the location of the expansion tank connection isn't important, then I'd suggest you read this book.

http://www.heatinghelp.com/shopcart/product.cfm?category=2-16

It is called 'Pumping Away', and it mostly about pumping away from the PONPC, the expansion tank connection. It's in US units (deg F, USgpm, etc, ) and is written in mostly non-technical terms for home owners. It's a very good book.

OK guys the new pump is on order, takes 2 days to travel 25km, but thats Belgium for you. 308 Euros about the same price as i could get it for in the UK.
So there'll be no more news until Weds afternoon when it should be fitted.

Onetap: take on board what you say about the pressure vessel and it will be moved at the same time as fitting the pump, easy enough job (don't mind those) arranging the iron pipework to fit the new pump dimensions is another matter should be fun

Thanks for all your help, let you know the out come

Onetap, I don't want to argue with you. Between us all we have hopefully managed to solve someones problem and I don't feel a need to defend myself but this is what I mean:

If you have a sealed loop of pipework which is, for arguments sake 10m from it's lowest to it's highest point then the static gauge pressure at the top will be zero and at the bottom will be roughly 1 bar (please, nobody start being all pedantic about gravity and relative water densities).

If you add an expansion vessel at the top of the system and push in 0.5 bar of air the entire system pressure is raised by that value ie the gauge at thee top will now read 0.5 bar and the one at the bottom will read 1.5.
If you now move the expansion vessel to the bottom of the circuit exactly the same pressures will be found throughout the system.

This is all before any pumps start.

Now, I agree that the best location for the vessel connection is between boiler and pump if pump is on the flow connection but it doesn't matter much if the other system components ie. pump, are correctly selected for the job in hand.

I understand "pumping away" but it's not that important for a sealed system unless the pump is near the top of the system or the system in not effectively filled. There isn't the same risk of air draw-in or pumping over that exists in an atmospheric arrangement - that is what I meant by it not being important.

My guess is that in this case the pump is next to the boilers at ground floor or even in the basement so will always have a reasonable static pressure to meet it's minimum NPSH requirements (Net Positive Suction Head). The cause of the cavitation in this instant is simply a case of a pump being selected at the wrong point of its operating curve. The hydraulic circulating pressure which it develops will be the same whether the expansion vessel is up or downstream. Moving the connection will, as you suggest help, but I don't believe that it will be enough to cure the noise problem.

Oops, sorry for the ramble. Onetap, if you don't agree - that's fine, I don't mind. Can we just agree to differ?

Ok guys just had a look at the pipework where the new pump will go, got to change the flanges and as it's all iron pipe gonna be difficult getting the right lengths to fit.

Any reason why I can't use copper pipe, say from the 1.5" bsp to 28mm copper or better still what size would you recommend using with the new pump UPS 32-55 and also do you know if this pump is supplied with the unions and what size pipe do they noramlly attach to.

Sorry to be a pain but better to ask now than to ask after

Cheers for your help