×
INTELLIGENT WORK FORUMS
FOR ENGINEERING PROFESSIONALS

Log In

Come Join Us!

Are you an
Engineering professional?
Join Eng-Tips Forums!
  • 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!
  • Students Click Here

*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.

Students Click Here

Jobs

pump pressure variables
6

pump pressure variables

pump pressure variables

(OP)
Does anyone have ideas on how to reduce pump outlet pressure?  We recently installed a pair of pumps (primary & backup up operation) which are delivering 80 psig pressure where we were expecting about 50-60 psig. I'm not sure why it's building that much pressure, but need to know ways/ideas on how to reduce it.

The centrifugal pumps are 20 hp electric motor driven, sized for 325 gpm at 150 TDH.
 
What do you need to know to give some reasonable suggestions?

 

RE: pump pressure variables

1. Reduce the impeller diameter;- ask the supplier to do it for you or he can supply a new impeller and you can replace it, keep the old one spare. You also can machine it down to the required diamenter. However, that will also impinge on the duty point, with lower flow, if you can accept it.
2. Send it back to the supplier and tell him where to stick it. If your datasheet/specification was correct, he should replace the pump with a new, good one.
3. Buy a new one, this time with correct datasheet/specification.
Howzat?
gr2vessels
 

RE: pump pressure variables

The least expensive fix is to install an orifice plate in the discharge line.

Some other options:

Install a control valve.
Install a VFD.

The reason for the excess pressure is that your pump suction pressure is probably higher than anticipated.

RE: pump pressure variables

2
Hold on there. It is probably nothing wrong with the pump.

Outlet discharge pressure is not determined by the pump alone, so there may not be anything wrong with the pump at all.  Pumps supply a differential pressure (psi), which when added to the suction pressure, result in the outlet discharge pressure (psig) in any given system.  The outlet pressure depends on and equals = Suction pressure + pump differential pressure, so if suction pressure changed, there is your answer.

You also said nothing about where the pump is operating now in relation to where its BEP is and what flowrate you thought you would be running when you wanted to have that 50-60 psig discharge pressure.  If you reduced the flow now below the flow you expected to have at 50-60 psig output pressure, the pump will obviously give higher pressures than 50-60 (assuming suction pressure remains constant).  

So,

1) What flow did you want at 50-60 psig, and

2) what flow do you have now at 80 psig?

3)  What was your suction  pressure expectation when you ordered the pump and

4) What is your suction pressure now?


 

Only put off until tomorrow what you are willing to die having left undone. - Pablo Picasso

RE: pump pressure variables

Agree with BigInch.
It's the flow that is not what you expected for the system.

Ted

RE: pump pressure variables

(OP)
Thanks all.  I really appreciate the suggestions :)
gr2vessels, we could probably live with reduced flow if we go with a smaller impeller.  Is there a rule of thumb correlating flow with impeller diameter? or should I just go with the supplier's pump curve?

bimr, we have gate valves installed upstream & downstream of the pump. When I suggested pinching those valves off, no one seemed to think it was good idea.  How would you suggest they be operated to reduce the pressure?  My idea was to simply slowly & slightly start closing them until the downstream pressure was reduced.  Their concern was pump cavitation. I believe it can be avoided by a slow patient approach in balancing the closing of both valves.

Also if an orifice plate is installed how will that affect flow?

Thanks biginch! your insight is always welcome.  I can't really say what the pump's BEP is because when I wrote the message I didn't have the pump data with me nor could I recall that value.  The pump was supposed to be designed to deliver 325 gpm at around 40 - 50 psi.  We did not have a flow meter connected nor any easy way of measuring flowrate. So I can't tell you what we were actually seeing.

The suction pressure expected was simply the head on the pump coming from the discharge nozzle of a fin-fan unit.  I estimate about 5 psig (static head) at the suction.  We did not measure that pressure during the testing though. Thanks I will attach a gage there today & see what we get.

I have attached a diagram showing this piping. The vertical distance from the fin fan nozzle to pump suction centerline is ~10 ft.  We measured the 80psig right at the valve on one of the 3" branches (off the pump discharge header line).

Hope this makes sense :)

RE: pump pressure variables

Since it is a closed system, I'll go out on a limb and predict that the initial hydraulic analysis for pressure drop around the loop was conservative and now you're getting less of a pressure drop that you thought and that results in more suction pressure remaining than you thought you would have.  If flow is also less than what you thought, you would tend to operate to the left of where you thought you would be on the pump curve, resulting in even a higher differential pressure than you thought to be added to that already higher suction pressure.

Could be one, the other, or both, but I will bet that there is nothing wrong with the pump in any and all cases.  RPM is at (or close to) normal operating speed right?

 

Only put off until tomorrow what you are willing to die having left undone. - Pablo Picasso

RE: pump pressure variables

I very rarely ever work with closed loops.  So, someone please correct me if I am wrong. But, it is not possible that the entire loop is pressuring up as it heats up because there is no accommodation made for thermal expansion.  In other words, if all you did was open a bleed valve and drop the pressure of the entire loop, the suction pressure to the pump would be reduced, which would reduce the discharge pressure.  

In a closed loop running at elevated temperature, I would expect to see some expansion chamber, vapor space in a vessel, bladder accumulator, etc.  I have a closed loop hot water boiler at my cabin.  If I blocked in the bladder accumulator with the system cold, and then fired up the boiler, I would expect the loop to pressure up and pop the relief valve

Determine the flow rate based on the pressures.  Read the suction pressure and discharge pressure.  Calculate the running head and plot it on the curve.  If the curve is reasonably accurate, this will tell you the approximate flow rate.  

If the entire loop is pressuring up because of expansion, cutting the impeller diameter may not solve the problem.  

Under no circumstance would I suggest pinching on a gate valve on the suction to the pump.  You can pinch down on the discharge valve as a test to see what affect an orifice is likely to have. But running that way long term will result in premature valve failure.   

Johnny Pellin

RE: pump pressure variables

Yes, W/O an expansion tank, or pressure relief, or thermal relief, pressure will increase with temperature, as long as the fluid expands faster than the pipe volume with temperature (most do).  This will be noted by a roughly equal increase in the pressure of the system at any point over the pressure at that same point when cold.  That could easily account for the 20 to 30 psi difference.

Only put off until tomorrow what you are willing to die having left undone. - Pablo Picasso

RE: pump pressure variables

I agree with BigInch that the loop head loss was probably conservatively estimated and the head loss around the loop is less than anticipated.

As JJPellin notes, you do not show how the loop is filled nor is a pressure control/relief.

As a short term fix, the discharge gate valves can be throttled back to provide some additional head loss. Throttling of the intake valves is not recommended. You should check the pump curve so that when you throttle back, your pump will still be operating in a good range on the pump curve.

Installing an orifice plate will be more or less the same as throttling the discharge valve.

If you can't live with the pressure, the best approach will be to reduce the impeller size.

Before spending any money, it would be wise to install some instrumentation such as pressure gauges and flow meters to see where the pump is operating on the pump curve.

RE: pump pressure variables

What is the highest elevation of your piping? In a closed loop system, the suction pressure will be the static height of the highest point of the piping system.
When you specified the pump to deliver 325 GPM at 150TDH, it is only the differential head generated by the pump and this would be the total friction losses of the system. Assuming the pump is delivering 325GPM as required, any positive static head available at the pump suction will add to the discharge head or pressure.

On the other hand, without knowing what flow you are getting and the pump curve, you could be operating at much reduce flow than you think.
 

RE: pump pressure variables

(OP)
I'm sorry, but the diagram does not indicate the expansion tank.  There is one in the system & it is vented. So it is not a truly closed loop.

I had also contacted the pump manufacturer yesterday & heard back from them today.  The problem was exactly as several had suggested - the impeller was too large.  The pumps were supposed to be sent with a 6.75" diam impeller, but instead were sent with 7.75" impellers. So hopefully this change will give the expected head downstream of the pump.

Although I believe this change will solve the mystery, we did take some measurements again this morning any way just to verify other parameters or guesses.  The suction pressure into the pump was about 6 to 8 psig with the pump off (static head).  I expected about 5 to 7 psig.  The suction pressure while running dropped to about 2-3 psig.  We also verified the motor speed was operating at the nameplate rating (3480 rpm).  These things onlfy confirmed that the pump was oversized.  

Unfortunately we have no flow meter & no real/practical way to measure it.  Jpellin, how do I calculate the flowrate given suction & discharge pressures?

thanks again everyone :)

RE: pump pressure variables

You need the pump curve to estimate flow rate. Get your differential pressure (discharge pressure - suction pressure) and convert to feet of head (psi x 2.31) and then locate that differential head on the pump performance curve for the correct impeller diameter. The corresponding flow rate from the performance curve should be fairly accurate.

RE: pump pressure variables

(OP)
Thanks Dubmac.  Based on the dP measured & the manufacturer's curve seems we were getting close to 200 feet of head and over 400 gpm!  wow!  We're in the process of having the impellers trimmed. The supplier didn't have any in the size needed in stock.

Hopefully reducing the impeller diameter from 7.75" to 6.75" will give the head (discharge pressure) expected.  I'll keep everyone posted on how it turns out.

Thanks to all. You were very helpful & immediately when needed :)

RE: pump pressure variables

Oh well, so much going out on limb.  What's the chance of getting an impeller from the mfgr with the wrong size.  It seems to be a lot bigger than I thought.

Only put off until tomorrow what you are willing to die having left undone. - Pablo Picasso

RE: pump pressure variables

It was only an 1" so near enough hairpull3
but 6.75" will give you 60 PSI - that's assuming the 80 psi was a correct measurement.

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.)  

RE: pump pressure variables

I am confused. You stated that the pump is designed for 325 gpm at 150 feet of head.  Then, later, you said you measured the pressures and plotted the head on the curve and got 400 gpm at 200 feet of head.  These two points can't be on the same curve.  It was my understanding that you have no way to make a direct measurement of flow.  How is it that you determined that the flow is higher when running at a higher head?  For a typical pump, as the head is increased, the flow is decreased.  Please elaborate on your methods.  

Johnny Pellin

RE: pump pressure variables

he thought it was designed for 325 & 150, but it is really 400 gpm at 200'.  Apparently it came with a bigger motor too.

Only put off until tomorrow what you are willing to die having left undone. - Pablo Picasso

RE: pump pressure variables

JJPellin, increasing flow to a given system increases head.  They did not get the pump they designed for, it was providing too much flow.  Hence, higher pressure.

Ted

RE: pump pressure variables

(OP)
Well the 6.75" impeller did not give the expected result. The pressure at the 3" branch was 72 psig.  (This is the same location that we took all previous discharge pressures, by the way).

Johnny, as Ted suggested I thought increasing the head increases the pump flow. I believe it's a proportional relationship:

 (Q2^2/Q1^2)=h2/h1

I tend to agree with BigInch about the motor being oversized, but I think the pump might be more oversized than the motor.  

We also measured motor data during this activity.  It is rated for 23 FLA.  When the pump was producing 80 psig with the 7.75" impeller, the current draw was 36-37 amps. At 72 psig with the 6.75" impeller the current draw was at its full rated capacity 23 amps. Wouldn't this indicate that the pump is too big for the motor?  The pump supplier did indicate we could safely reduce the impeller to 6" which I'm sure will reduce both the discharge pressure. However I'm not positive that doing so will give the desired pressure (40-50 psi) at the 3" branch.
 

RE: pump pressure variables

I guess am not being specific enough with my terminology. I am a pump guy.  A centrifugal pump will tend to perform to a curve of flow versus head.  As the head is reduced, the flow will increase. As the head is increased, the flow will be reduced.  In these statements, the head is the total developed head of the pump.  This is determined by subtracting the suction pressure from the discharge pressure and converting from units of pressure to units of head.  

You may be using the term head to describe the head loss in the system.  That is based on the system curve. For a system curve, as the flow is increased, the head is increased and vice-versa.  The operating point for a given pump in a given system is the intersection of the pump curve with the system curve.  It is impossible for a typical centrifugal pump to have a curve such that 325 gpm at 150 feet and 400 gpm at 200 feet are both on the same curve.  

It sounds like you decided to use a different curve based on the fact that you were getting more pressure at one point than you expected.  But, if you can't measure the flow directly, you can't make that assumption.  What is the suction pressure right at the suction of the pump?  What is the discharge pressure right at the discharge of the pump? What flow rate would this correspond to on the original curve for the installed impeller and motor speed?  What would cause you to believe that the flow rate is higher than this value?

I get the feeling that we have wondered off into the swamp on this one and we are getting bogged down.  Something seems to be drastically wrong. Are you looking at 50 Hz curves but running with a 60 Hz motor?  Are you looking at curves for a different size, speed or hydraulic design?  I don't see enough information anywhere in this chain to understand what is going on.   

Johnny Pellin

RE: pump pressure variables

Johnny

The OP specified a pump to produce 325 GPM with 150 diff. head and according to the pump curve should be fitted with a impeller of 6.75"

But the pump vendor supplied him a pump with a impeller 1 biginch over size at 7.75" and gave a flow of 400GPM at 400 ft diff  head
on him system.


 

RE: pump pressure variables

(OP)
Johnny, I make no profession to be a "pump guy" & I respect your knowledge in the field.  What I shared before was simply what I observed in the field and what I could interpolate from the suppliers pump curve based on field results. I do agree that something is not right.  I have attached the pump curve and a table summarizing the field results. hopefully this will make results easier to follow for everyone including me! :)

RE: pump pressure variables

(OP)
Oops.  Here is the other file summarizing the measured parameters. I guess I can only load one file per message?

Anyway we still have slightly more head than anticipated. Since we can't safely reduce the impeller size any less, we will install either an orifice plate in that 3" branch or open a bypass line put in between the 6" suction & discharge headers to reduce the overall line pressure.

RE: pump pressure variables

"Since it is a closed system, I'll go out on a limb and predict that the initial hydraulic analysis for pressure drop around the loop was conservative and now you're getting less of a pressure drop that you thought and that results in more suction pressure remaining than you thought you would have."

From a design engineer's standpoint, I'd have to say BigInch nailed it.

RE: pump pressure variables

Sorry, missed the more recent post, oversized impeller from vendor.   

RE: pump pressure variables

I should know by now not to assume ANYTHING in these forums.  The seem to be affected by Murphy's_Laws^2

Only put off until tomorrow what you are willing to die having left undone. - Pablo Picasso

RE: pump pressure variables

(OP)
"I should know by now not to assume ANYTHING in these forums.  The seem to be affected by Murphy's_Laws^2 "  lol!

Or assume that you got incomplete info  :(  Sorry about that BigInch, but do appreciate your input as always - very insightful.

On a related note with regards to the pressure concerns, looks like we are good with the pressures in the 3" branch which leads to nozzles.  There was an orifice further downstream in that branch which reduced the pressure to an acceptable level.  We were not measuring at that location last week when we first started testing things out.  Yesterday we did.   :)

BTW, anyone have any idea why most previous posts did not include the files I uploaded?  Is there a limit on file uploads to a thread?
 

RE: pump pressure variables

There was some "reworking" of the website going on over the weekend and quite a number of things went bad in the process.  Probably had something to do with that.   

Only put off until tomorrow what you are willing to die having left undone. - Pablo Picasso

RE: pump pressure variables

kahlilj, no problem.  Sorry I might have mislead you, but I thought I was starting with the highest probability items first.  As always when using probabilities, sometimes you win, sometimes you lose.  From my perspective, when I get a question like that, its not often that I actually find something wrong with a pump.  Often? No. In fact, I never have.  It's always been a conservative pressure drop resulting in a pump selected with the wrong BEP for the job at hand, or visa versa. Faulty engineering, is more common than a faulty pump.  That vendor is going to start making the good engineers look bad and the bad engineers look good.  Not a lot of future in that.

Only put off until tomorrow what you are willing to die having left undone. - Pablo Picasso

RE: pump pressure variables

There is still a number of questions hanging above;- how did the wrong pump ended installed and started operating? It obviously was a long string of engineering failure. We haven't heard if the purchase specification was correct or not, how did the supplier end selecting/recommending a wrong pump, how did the wrong pump offer ending accepted (by whom?), how could a wrong pump be installed without proper engineering review and approval and then operated without proper commissioning/testing?
Just curious, I think we all can learn something of this.
Cheers,
gr2vessels

Red Flag This Post

Please let us know here why this post is inappropriate. Reasons such as off-topic, duplicates, flames, illegal, vulgar, or students posting their homework.

Red Flag Submitted

Thank you for helping keep Eng-Tips Forums free from inappropriate posts.
The Eng-Tips staff will check this out and take appropriate action.

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!


Resources