Smart questions
Smart answers
Smart people

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.

Donate Today!

Do you enjoy these
technical forums?
Donate Today! Click Here

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.

Reducer on discharge side of cent. pumpHelpful Member! 

shahyar (Chemical)
21 Mar 07 1:21
I know it is not common to put reducer on discharge side of a centrifugal pump. But I like to know WHAT will happen if I put it? (centrfugal pump/blower).
The similar question could be "What if we put expander on suction side of a centrifugal pump?"

pennpiper (Mechanical)
21 Mar 07 9:24
Here is some advice on "Piping terminology"
-Reducer- A device used to change the size of a pipe line from one line size to another line size.  No matter which direction the commodity in the line is flowing (in process plant piping) the device is still called a "Reducer."  In piping fittings there is no such thing as an "Increaser"or "Expander".

Now, to addressr what I think is your question. (can you put a reducer in the discharge line of a pump?)
Yes! you can put a reducer in the discharge line of a pump.  It is done all the time, however it is normally required to change the discharge line from the (smaller) pump discharge size to a larger pipe line size.  

I have seen a few (not many) installations where the pump discharge flange is larger than the required line size.  In this case a reducer was included just downstream of the pump nozzle.
katmar (Chemical)
22 Mar 07 7:11
I agree with what pennpiper has said, but I would go further and say that it is so unusual to have a reducer installed on the discharge of a pump in such a way that the line size is smaller than the pump discharge that I would worry that either the pump has been wrongly selected or that it will run too far away from the BEP to be reliable.  There would have to be a very good reason for me to accept such a set up.  I'm not saying that it can't happen, but it is so unusual that it should be carefully checked.

Katmar Software
Engineering & Risk Analysis Software

rmw (Mechanical)
24 Mar 07 16:11
If an "expander" is put on the outlet of a fan or blower it is called an evase.

It is to convert the high velocity flow out of the fan discharge to static pressure.

BigInch (Petroleum)
25 Mar 07 4:54
It would be an indication that the system was poorly designed for its present duty, or that no other pump or pipe was available.

BigInchworm-born in the trenches.

Helpful Member!  25362 (Chemical)
25 Mar 07 5:32

Eccentric reducers attached to the suction flange are arranged with the flat side at the bottom when the supply comes from above the pump, and with the flat portion at the top when feeding the pump from below.
pennpiper (Mechanical)
25 Mar 07 8:42
The question above did not ask about the style of reducer (eccentric or concentric) used at pumps.  It just asked about the use of a reducer in both the suction and discharge.

Eccentric reducers (when required) should be and will be installed with the flat side up no matter where the suction comes from.
25362 (Chemical)
25 Mar 07 12:28


Your last paragraph:


Eccentric reducers (when required) should be and will be installed with the flat side up no matter where the suction comes from.

Igor J. Karassik's chapter 12 (Installation, operation, and maintenance), fig 9, in Karassik's et al. Pump Handbook, McGraw-Hill, tells a different story.

Kindly comment.
pennpiper (Mechanical)
25 Mar 07 16:44

It is the way I do it and that is the way I teach other pipers to do it.
Artisi (Mechanical)
25 Mar 07 16:49
Agreed - the flat side up no matter where the suction (inlet) comes from. This means from below the pump inlet or above the pump inlet or inline with the pump inlet.  
katmar (Chemical)
26 Mar 07 0:52
We are engineers here, and engineers like reasons for their rules. It does not help much to say "flat side up, no matter what". There have been many posts on this subject before, for those wanting to follow up on it.

My own rule is - flat side up if there is entrained air (to prevent air pockets), flat side down if there is silt or solids present (to encourage the solids to be flushed into the pump suction). But if you have a flooded suction with no air and no solids you can install it any way you want, or even use a concentric reducer.

Katmar Software
Engineering & Risk Analysis Software

25362 (Chemical)
26 Mar 07 1:25

Artisi and Pennpiper, it is well established that gas pockets (high spots) should be avoided and suction piping should be as short and direct as possible to reduce friction losses, therefore your comments on this subject are understood.

This is a shortened version of what Karassik had to say on the subject (second edition of the Pump Handbook):


Where a static suction head will exist the pump suction piping should slope continuously downward to the pump... If the source of supply is above the pump, the straight side of the reducer should be at the bottom (Fig. 9). Installing eccentric reducers with a change in diameter greater than 4 in (10 cm) could disturb the suction flow. If such a change is necessary, it is advisable to use properly vented concentric reducers.

The paragraph goes on explaining the arrangement of fittings to avoid flow disturbance.

What's wrong with Karassik's statement ?
Artisi (Mechanical)
26 Mar 07 1:35
" --  pump suction piping should slope continuously downward to the pump.." and if it doesn't and there is a horizontal section before the reducer then air could collect at the reducer with the flat on the bottom.

For me the best rule is flat on the top - then you are safe from air collection - however, if there is good reason to change the "rule" to the flat on bottom it becomes an engineering decision where a factors can be reviewed.
katmar (Chemical)
26 Mar 07 3:32
I don't know the Karassik book, but I would not slope piping down towards the pump if I was concerned about air entrainment. I once came across an installation where this was done and it seemed that the capacity was being restricted by bubbles trying to rise against the flow of liquid. I felt it would be better to have the bouyancy and flow forces pushing the bubbles in the same direction. Unfortunately I was not able to persuade the user to change the piping, so I cannot be sure that this was the problem.

If possible I would prefer to have the vertical section sized to cope with self venting flow, and then the horizontal section either truly horizontal or sloping slightly up to the pump. If there had to be a reducer I would put the flat side at the top. Does Karassik give a reason for wanting it flat side down? Of course, if solids are present then sloping down to the pump would make sense.

Katmar Software
Engineering & Risk Analysis Software

25362 (Chemical)
26 Mar 07 5:27

Excluding considerations on air pockets created by pipejoint steps, or by projecting gaskets, it appears that when the size of the suction pipe is larger than the bore of the pump inlet air pockets can be avoided precisely by providing eccentric reducers with the flat side on top.

Sam Yedidiah (Centrifugal Pump User's Guidebook), tells us that when the pipe diameter is more than one size or two larger than the bore of the suction nozzle, a straight pipe section of equal diameter as the pump inlet should be installed between the reducer and the pump because the eccentric reducer may deflect the liquid stream toward one side of the impeller eye, reducing output and efficiency.

From the picture in fig. 9 (in Karassik's manual, chapter 12) one can see that (at the end of the vertical pipe) there is a flanged long-radius elbow, followed by a horizontal eccentric reducer, with its flat portion at the bottom, connecting to the pumps' suction flange (itself an elbow). The arrangement shows a smooth flow profile with no air pocket. Contrarywise, with the flat portion on top, a slope would appear at the bottom, giving rise to more turbulence.


Artisi (Mechanical)
26 Mar 07 6:50
What is interesting with this posting is it started out with eccentric reducers on the discharge side and has now done a complete 180 to eccentric reducers on the inlet side, and we haven't really touched on the second part of the original question of expanders (eccentric or conical) on the inlet side which could prove to be an interesting discussion.
katmar (Chemical)
26 Mar 07 7:20
Artisi, I don't think there is much more to say about "expanders" in the suction line. It would be an even more extreme case of what I wrote in my post of 22 Mar 07 7:11, or what BigInch wrote on 25 Mar 07 4:54.  Except that pump manufacturers might find such a setup useful for determining their products' resiliance in a cavitating environment.

The example given by 25362 explains how Karassik justifies putting the flat side down when the feed is from above, or the flat side up when the feed is from below.  Does this mean that if the line feeding the pump suction is horizontal and comes from the east that the flat side should be on the west?

Katmar Software
Engineering & Risk Analysis Software

SeanB (Chemical)
26 Mar 07 7:32
After reading the posts above I had to throw my 2 cents in here and I pasted in the below reference that I have - which would agree with what 25362 had posted.

Jacques Chaurette p. eng., Fluide Design Inc.
August 2005

Always use an eccentric reducer at the pump suction when a pipe size transition is
required. Put the flat on top when the fluid is coming from below or straight (see next
Figure) and the flat on the bottom when the fluid is coming from the top. This will avoid
an air pocket at the pump suction and allow air to be evacuated.
Artisi (Mechanical)
26 Mar 07 7:34
East or west - flat on top :)

My reading of original posting re expander on the suction side was from small to large against the inlet or close to the pump inlet which could be problematic.
JJPellin (Mechanical)
26 Mar 07 7:41
I have never understood the concern about an air pocket trapped on the suction side of a pump.  I have never heard how it can cause any problem.  Anyone working in an oil refinery with a jet water pump has seen an example that seems to disprove the problem.  Every jet water pump I have ever seen was a barrel style with top suction and top discharge.  Every one I have ever seen was pumping from an atmospheric tank sitting on the ground.  That means that all of these pumps have a high point in the suction piping where it turns to go into the top suction. In a typical refinery, this jet water pump will be the highest pressure, highest energy pump in the plant.  If an air pocket at the suction is such a terrible problem, why do we not have a problem with jet water pumps.  The only problem I have seen with the suction to a jet water pump was when someone tried to prevent the air pocket.  They installed a suction pot, open to the atmosphere on the top so there would be no air pocket.  This suction pot caused several wrecks of the jet pump when the level controller plugged off with coke fines and the pump was allowed to run dry.  Simply because someone states that a certain pipe fitting has to be installed a certain way, I will not do so unless there is some basic engineering principle to describe the need or solid test data to prove the benefit.  Does anyone have such hard test data?

Johnny Pellin

SeanB (Chemical)
26 Mar 07 11:41
Just curious, what type of pump is a jet water pump?
JJPellin (Mechanical)
26 Mar 07 12:08
A jet water pump is a multistage pump used to provide the cutting water for cutting coke out of the drum.  One of ours may be typical. It is 10 stages, barrel style.  The impellers are all aligned with a very large balance piston at one end.  The pump produces 1000 gpm at about 3600 psi pumping water that is recycled so it tends to be dirty.  The pump starts and stops several times per day.  It only runs when they need to cut a drum which could be on a cycle of 12 to 24 hours with one jet pump serving 2 to 6 drums.  So, in combination it might run twice per day or as much as 6 times per day for 2 to 4 hours each time.

Johnny Pellin

BenThayer (Chemical)
26 Mar 07 13:40
i have to admit that we routinely put the flat side of ecentric reducers on the bottom at pump inlets if we think we might have a cleanability issue in our pharma plant.

drainability and being able to prove it can be important to the FDA, etc.  and if you do it right, it is not an issue for the operation and reliability of the pump (and even if it was, the FDA and Q would not care).
insult2injury (Mechanical)
26 Mar 07 22:38

A jet pump and a multistage barrel pump are very different.  A jet pump refers to an ejector, eductor, etc...  The pump you are talking about is just a multistage centrifugal pump with double casing same as a boiler feed pump.  In boiler feed applications (and condensate pumps for that matter) where the pumped liquid is at or very near saturation.  Ask any pump manufacturer, and they will agree with the advice of Karassik.  In fact with barrel pumps the manufacturer's preference would be to have a completely vertical drop from the source tank to the suction nozzle.


quark (Mechanical)
27 Mar 07 0:19

You can have reducers in the discharge(one size, as a thumb rule) but pump manufacturers suggest not to reduce the suction piping below the given suction flange size. So, an expander(I am comfortable with the term expander, technically and semantically, and the term has been in use) is avoidable in the suction line.

JJPellin (Mechanical)
27 Mar 07 7:46
When I refer to a jet water pump I am referring to the service. It is like saying a Naphtha Reflux pump or a Fuel Oil Product pump.  Jet water pumps could be many configurations, but I have only ever seen multi-stage barrel style pumps in this service.  I make this an example, because jet water service in a coker will tend to be the largest, highest pressure, highest suction energy pump in a typical refinery. If there was ever going to be a problem associated with a high point vapor trap in the suction line of a pump, this should be the service where it would be most extreme.  But I have never seen or heard of a jet water pump that did not have a very large vapor trap in the suction line.  

Johnny Pellin

djack77494 (Chemical)
29 Mar 07 17:37
As much as we may have our preferences and our good and bad experiences with doing things one way or another, there is no substitute for engineering judgement. Generalities may apply in the majority of situations, but there are always the exceptions.

Moving on to the original question and adding something to satisfy the engineer's desire to have reasons for our rules, let's think about the discharge of a pump. Exiting the pump is a highly turbulent, undeveloped liquid stream. The pump vendor is striving to offer a product that economically meets the specification, and a larger size discharge nozzle = more cost to him. The added cost does not equate to added value beyond the point where excessive wear or reduced product life is avoided. Once out of the pump, however, the engineer seeks the optimum size pipe, which typically would result in lower velocities and pressure drops. Thus the piping is typically larger than the pump discharge nozzle.

Late katmar, I'd almost say that this is ALWAYS the case, but if I did so, I'd violate the points I was making in my first paragraph. ;-)
pumpking (Chemical)
21 Apr 07 3:58
Many subjects being discussed here, but coming back to the original question, have to agree with BigInch, if your putting a reducer on the pump discharge then it is worth looking at the whole system design, as friction losses of a smaller discharge pipework system will be much higher, therefore, it will more than likely be the case that you can install a much smaller pump to do exactly the same job !!  It will also make running costs and energy costs much less over a period of time !!
TenPenny (Mechanical)
21 Apr 07 13:31
I have seen reducers on the discharge.  All you are doing is adding restriction and therefore more head to overcome, which then will change the performance of the pump.  You'd have to analyze the whole system and the pump curve to see the effects.

Step back a minute and think - you're paying for energy to move a liquid at a certain pressure, then by putting in a pipe restriction, you are wasting energy.  It's like opening a window in the middle of winter - you're just wasting energy.

There are times when this actually helps, but in general, it's not a logical idea.
Artisi (Mechanical)
21 Apr 07 17:14
It is very posssible that a reducer on the discharge side is valid - take for instance an installation where a large pump operating a slow speed has been selected for a hydraulic consideration ie, very low NPSHa. It is therefore valid for the discharge pipeline from a pure economic perspective would be of a smaller diameter than the pump discharge flange.

Plus there is no law from the pump police to say that the discharge pipeline must be the same size as the pump discharge flange- pipeline diameter is based on a friction head / economic consideration.

As TenPenny has pointed out "There are times when this actually helps, but in general, it's not a logical idea."
Scipio (Mechanical)
23 Apr 07 12:12
Agreed with Artisi, I've seen reducers on discharge twice for valid reasons, once it was a larger progressing cavity pump run at a very low speed for erosion concerns, the second was a slurry service where the pipe diameter was reduced to keep the slurry above it's settling velocity.

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