Smart questions
Smart answers
Smart people
Join Eng-Tips Forums

Member Login

Remember Me
Forgot Password?
Join Us!

Come Join Us!

Are you an
Engineering professional?
Join Eng-Tips now!
  • Talk With Other Members
  • Be Notified Of Responses
    To Your Posts
  • Keyword Search
  • One-Click Access To Your
    Favorite Forums
  • Automated Signatures
    On Your Posts
  • Best Of All, It's Free!

Join Eng-Tips
*Eng-Tips's functionality depends on members receiving e-mail. By joining you are opting in to receive e-mail.

Posting Guidelines

Promoting, selling, recruiting, coursework and thesis posting is forbidden.
Jobs from Indeed

Link To This Forum!

Partner Button
Add Stickiness To Your Site By Linking To This Professionally Managed Technical Forum.
Just copy and paste the
code below into your site.

why pump cavitates despite enough NPSHHelpful Member!(2) 

kalamarhss (Mechanical) (OP)
19 Mar 12 13:26
Hi to all fellow engineers.
We have a pump that cavitates despite there is enough NPSH.
The working fluid is water at 45 Celcius. Pressure just before the pump is 0.2 bar that means NPSH=12m, well above the NPSHr.
What are the possible reasons for this except wrong curve from the manufacturer?
1gibson (Mechanical)
19 Mar 12 14:46
Did manufacturer perform an NPSH test to confirm it met published values?
Is it a semi-open impeller, is lift set too high?
Is the vane eye tip speed too high? (ask manufacturer to confirm)
Is min submergence met, is it in a sump, do you have vortices?
Is there suction blockage?
kalamarhss (Mechanical) (OP)
19 Mar 12 15:02
1gibson thanks for the interest.
No the manufacturer did not do a test.
It is a centrifugal pump single stage closed impeler.
How do u find the tip speed? The pump runs at 2900 1/min.
The pump sucks from a tank which is 10 meters higher. The tank is closed and there is 0.2 bar pressure in the tank.
There is a flow straightener installed before the pump.
There is a filter and a buterfly valve before the pump.
But as I said the manometer shows 0.2 bars before suction when NPSHr is 4-5 meters.
bimr (Civil/Environmental)
19 Mar 12 15:58
Actually 0.2 bar = 2.04 meter of water head. How did you come up with 12m?

Do you actually have a filter or a strainer on the inlet side of the pump. There is probably not enough head available for a filter.

Note that pump cavitation starts gradually. There is not a sharp break where cavitation starts to occure at 4-5 meters.
kalamarhss (Mechanical) (OP)
19 Mar 12 16:03
I think that manometres show deferential pressure.
but when calculating NPSH we need absolute pressure.
Am I wrong?
You are right its a strainer.
Excuse my poor English.
1gibson (Mechanical)
19 Mar 12 16:13
If it is an open sump, and that is 0.2barg, then it's about 7 ft head plus 32 ft (depending on elevation) for atmospheric pressure, which is about 12m. This is what I assumed, because I assumed it was a vertical pump.

Bimr, you asssumed it was a horizontal, closed system, so yes about 2 m.

Now more info, seems like it is leaning towards being a horizontal pump, with tank and 10m static head??? I'm out of guesses for today.

To find the tip speed, as I indicated, you ask the manufacturer. Sometimes 1800 rpm pumps are misapplied at 3000 rpm and impeller design is not appropriate. Is that pump available at 3600 RPM? If not, look into this furhter. If it is available at 3600, then nevermind. To estimate, take the impeller eye diameter (should be available in some for or another in the literature, if not then just take the impeller eye side wear ring diameter (check your spare parts list) and reduce it by about 1/2".) With diameter and speed, convert it to a velocity. Typically do not want to exceed 85 ft/s but this is more of a vertical pump rule.
bimr (Civil/Environmental)
19 Mar 12 16:33
What do you mean by the "The tank is closed"?

You should have enough NPSH if the tank is open to the atmosphere. If the tank is closed, you will be pumping from a vacuum and you will not have enough NPSH.

See the diagrams in this link:

clay87 (Mechanical)
19 Mar 12 16:57
Are you operating at a flow rate that is considerably lower than design?
kalamarhss (Mechanical) (OP)
19 Mar 12 17:04
Ok sorry for the misunderstanding.
We have a closed tank wih water at 45 Celcius. The tank is elevated 10 meters. It is in -0.8bar vacuum.
In the suction line there is a strainer and a buterfly valve.
A pressure gauge at the suction of the pump shows 0.2 bar.
Dos this mean I have NPSHa = 2 - vapor pressure?
Or NPSHa = 12 - vapor pressure?
To make things more clear. I think NPSH is absolute pressure. But manometers show diferential.
If you place a manometer in the suction of a pump that pumps from a n open tank at 1m elevation it will show 0.1 bar.
But the NPSHa will be 11m.
kalamarhss (Mechanical) (OP)
19 Mar 12 17:06
Yes clay but according to the manufacturers curve we should be OK.
TD2K (Chemical)
19 Mar 12 17:59
NPSH is the additional pressure at the suction of the pump in excess of the vapor pressure or o.2 bar(a) minus the vapor pressure at 45C. I'm sitting here in a hospital waiting room so I have limited resources but you don't have close to 12m NPSHA. What is the NPSHR?
TD2K (Chemical)
19 Mar 12 18:13
Please specify your pressures as gauge or absolute.  I assumed from your various posts the pressure at the pump suction was absolute but in looking at your numbers again it could be gauge and that makes a huge difference.

Don't be vague in your units.  
kalamarhss (Mechanical) (OP)
19 Mar 12 18:23
Ok I will open a new thread about the NPSH.
Lets asume that I am sure that NPSHa > NPSHr.
What else should I check if there is cavitation?
kalamarhss (Mechanical) (OP)
19 Mar 12 18:26
Sorry TD2K.
Its just my English. I will be more careful.
TD2K (Chemical)
19 Mar 12 18:49
Okay but it's a key piece of information to your problem.  What is the suction pressure?  Are you measuring that on the pump side of the suction strainer?
bimr (Civil/Environmental)
19 Mar 12 21:40
NPSH calculation:

Atmospheric pressure = 10.2 m
 Static head = 10 meters
 Tank Gauge pressure = -0.8 bar = -8.2 m
 Estimate pf vapor pressure of water =  0.18 m
NPSHA (net positive suction head available) = 10.2 + 10  - 8.2 - 0.18 = 11.8 m.

It looks like you have adequate NPSH, but this NPSH calculation does not include the headloss through the suction piping. The pipe headloss should be subtracted.

What is the diameter of your piping? You should be using 200 mm dia. pipe for this water flow.

If you only have 0.2 bar (2 meter water head) at the suction of the pump, you may have too much headloss in the suction piping. Is the strainer clogged?
stanier (Mechanical)
19 Mar 12 22:53
Is the pump running in the right direction?

IS it cavitation or perhaps vibration? Vibration can occur if you are running too ar left of BEP.

Is the pump running at duty point? NPSHr increases exponentially past BEP.

"Sharing knowledge is the way to immortality"
His Holiness the Dalai Lama.

clay87 (Mechanical)
19 Mar 12 22:55
How do you know you are 'cavitating'?
Is it noise?  If so, what exactly does it sound like?  

Is it vibes?  If so, what does the vibration signature look like?

Is it pressure/flow instability?  If so, please explain.

Is it impeller pitting damage?  If so, where on the impeller?

BigInch (Petroleum)
19 Mar 12 23:08
Straight suction piping?    

From "BigInch's Extremely simple theory of everything."

kalamarhss (Mechanical) (OP)
20 Mar 12 2:33
TD2K  I am measuring after the strainer before the pump suction. The gauge pressure is 0.2 bar.
bimr The suction line is DN 450. does the matter that i am measuring gauge pressure change anything?
stanier yes the pump is running at the right direction. It is definitely cavitation due to the damaged impeller at the pressure side of the vanes. The pump runs at half the design flow .
clay87 It sound like there are rocks in the pump. Lote of vibration. What do u mean vibration signature. What is the vibratio signature of cavitation? There is no great instability. A little fluctuation. As I said the impeller is damaged at the pressure side of the vanes. There is pitting of the impeller.
Big inch. After the strainer there are  more than5X D of straight suction piping meaning aproximately 5m.
I am pretty sure it is cavitation but the question is why when the curve says that there shouldnt be any. The NPSHr at that point is 4-5 m.
Thanks everyone for the interest.

bimr (Civil/Environmental)
20 Mar 12 9:03
Is the pump running at half the design flow because of poor pump system performance or by operational requirements?

You may have air ingestion. Air ingestion is not really cavitation, but acts like it.

Air ingestion has an adverse affect on the pump. Air ingestion seldom causes damage to the impeller or casing. The main effect of air ingestion is loss of capacity.

Artisi (Mechanical)
20 Mar 12 9:43
Running at half its design flow is probably where the problem stems from, poor entry conditions of the flow onto the impeller blade - this usually results in noise which can sound like cavitation.   

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)  

nfinit (Mechanical)
20 Mar 12 10:07
just a question, have you verified that the suction pressure gauge is reading the correct pressure?
BigInch (Petroleum)
20 Mar 12 10:13
Perhaps for a large diameter pipe, you may have prerotation eddies that do not disipate in such a large space within only 5 diameters.  Remember that 5D is an absolute minimum.

If you are controlling the flow to 1/2 BEP with the butterfly valve, I would suspect the problem therein is, especially if the filter and valve are between the straightner and the pump.

From "BigInch's Extremely simple theory of everything."

BigInch (Petroleum)
20 Mar 12 10:16
It also seems that (I don't pump hot water myself) hot water can require even more NPSH than what would be indicated from adjusting the cold water curve for hot water's vapor pressure alone, but I'll leave that topic open for those hot water pumpers out there.  I'll bet on the partially closed butterfly valve.

From "BigInch's Extremely simple theory of everything."

TD2K (Chemical)
20 Mar 12 10:21
"TD2K  I am measuring after the strainer before the pump suction. The gauge pressure is 0.2 bar."

Thanks.  If you have 0.2 barg at the pump suction and you are pumping 45C water, you have about 11.3m NPSHA per my calc, pretty close to your estimate, that's a fair bit.  What is the vendor saying the pump requirement should be for NPSHR?

Where is the operational point sitting on the pump curve?  Is the NPSHR rapidly increasing at this flow?  Can you reduce the flow and see what happens?

Lots of other good suggestions also from other people.
TD2K (Chemical)
20 Mar 12 10:51
Sorry, I missed seeing your earlier post where you provided the NPSHR, you should have ample.

I looked at a troubleshooting boot by Norm Leiberman, his comments are:

1. The pump discharge pressure is erratic, it varies from normal to low
2. the pump flow is erratic
3. The pump suction pressure is drawn down slowly but evenly and then jumps back up
4. The pump amps are erratically low
5. The pump may (but not always) make a rattling sound

He then says "Unfortunately reality is often more complex than this conventional definition of cavitation"

If you raise or lower the level in the suction vessel is there any change in what you are seeing?

Have you looked at the manufacturer's standard curves for your style of pump and seen if the NPSHR for those impellers are in the same area as your impeller?

Maybe you've provided this and I'm missing it also but what is the operational flow rate and design head.  Is the pump's current operational point agree with the pump curve?
clay87 (Mechanical)
20 Mar 12 11:14
I hate to keep pestering but I don't quite understand:

When you say 'pressure side' does that mean the damage is on the underside of the vanes at the suction (in other words, do you need a mirror to see the pitting)?  If so, suction recirculation as Artisi suggests sounds like your problem.  This is consistant with low flow rate, pumping rocks noise, and high vibrations.  

Out of curiosity, is this a double suction impeller?

bimr (Civil/Environmental)
20 Mar 12 11:44
I believe you may be experiencing recirculation cavitation because you are operating the pump at far below the flow range of the pump.

In centrifugal pumps, there is typically a small flow from the impeller discharge back to the suction through the clearance between the impeller hub and front wear ring. When some pumps run at flows well below the design range, water can recirculate within the pump.

A characteristic damage from suction recirculation is that most of the damage occurs on the high pressure side of the vane. The noise from cavitation also tends to be higher than that cause by low NPSH cavitation.

Helpful Member!(2)  bimr (Civil/Environmental)
20 Mar 12 11:59
kalamarhss (Mechanical) (OP)
20 Mar 12 16:36
bimr  the pump runs at half the flow due to operational reasons. Air ingestion is a possibility but there is severe damage of the impeler. The damage actually is in both sides but it is more serious in the obvious side ( the one you dont need a mirror to see). Recirculation is one of the possibilities. I have to check with the diagramms you uploaded. Very useful thanks.
Big inch the valve is completely open. Hot water pumpers come out please. And the distance is more than 10X 0.4 which is the ID of the pipe.
TD2K thanks for the input. I wouldnt say the pressure is "erratic". It fluctuates from 28-30 bar.
clay as I said the damage is in both sides but more severe in the obvious side.
If it is not cavitation shouldnt the manufacturer point out the operational limits of the pump and to prohibit operation under a certain flow?
Thanks everyone for the valuable info and your time.
bimr (Civil/Environmental)
20 Mar 12 17:25
Operation of pumps at less than optimum conditions is a major problem in industry. Most of the pump vendors that I have worked with provide minimal input in the pump selection process.

Here is a link to some reprints from Igor Karassik articles discussing the problems associated with off BEP operation.
clay87 (Mechanical)
20 Mar 12 21:52
Good luck, kalamar.  Maybe there is some way to increase flow rate, perhaps you can open a recirc line?  I suppose this would make the condition worse if you are experiencing NPSH cavitation but at least you'd know :).
stanier (Mechanical)
21 Mar 12 0:45
Perhaps you should consider reducing the diamter of the impeller so that it is on point.

"Sharing knowledge is the way to immortality"
His Holiness the Dalai Lama.

Artisi (Mechanical)
21 Mar 12 3:31
we all working in the dark, therefore our crystal balls don't function all that well - how about some meaningful operating and pump data, this might go a long way to help in solving the problem.

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)  

BigInch (Petroleum)
21 Mar 12 4:33
And tell the truth.  You're throttling with the butterfly on suction arn't you.

From "BigInch's Extremely simple theory of everything."

kalamarhss (Mechanical) (OP)
21 Mar 12 13:38
bimr the manufacturer gives a min operational Q of 90 qmph. But I guess this is for cold water . Is there a way to translate this to my operational temp?
clay it is a possibility
stanier there would not be enpugh head
Artisi you are right i will come back with moore data I promise.
Big inch I wish I was (sigh)things would be much simpler. :)
BigInch (Petroleum)
21 Mar 12 15:47
If you're not throttling with that butterfly, then that should not be a butterfly valve.  Remove it and replace it with a fully opening ball valve at your earliest opportunity and leave it fully open unless the pump is off.

From "BigInch's Extremely simple theory of everything."

bimr (Civil/Environmental)
21 Mar 12 17:00
My opinion based on the information provided is that you have suction recirculation. Have you tried operating the pump at close to design flow rate and observed the cavitation?

Regarding "the manufacturer gives a min operational Q of 90 qmph. But I guess this is for cold water. Is there a way to translate this to my operational temp?"

Is that 90 cubic meters per hour? That would represent about 25% of the design flow if 2900 liters per minute is half.

Have you contacted the manufacturer about this problem and given the manufacturer your as-built operating parameters?

BigInch (Petroleum)
21 Mar 12 22:55
I agree.  Just trying to emphisize that the butterfly valve there isn't helping things.

From "BigInch's Extremely simple theory of everything."

ports394 (Mechanical)
3 Apr 12 12:59
Is the impeller damaged from cavitation? Check that to verify it is happening.

Reply To This Thread

Posting in the Eng-Tips forums is a member-only feature.

Click Here to join Eng-Tips and talk with other members!

Back To Forum

Close Box

Join Eng-Tips® Today!

Join your peers on the Internet's largest technical engineering professional community.
It's easy to join and it's free.

Here's Why Members Love Eng-Tips Forums:

Register now while it's still free!

Already a member? Close this window and log in.

Join Us             Close