I need a hard cite for # straight diameters upstream of suction 2

[KernOily]

OK I am having a major pi##ing match with my piping designer. I need a hard cite/reference to a reference book for the required number of straight pipe diameters upstream of the suction flange/suction reducer. I checked all my references (Karassik, Marks, Dufour, Rip Weaver, etc.) but no hard numbers. Everybody just says "smooth flow into the suction...".

So I did a search here and read all the posts re: straight suction lengths (all 60-plus posts) but nobody gave any cites.

Pumps are horizontal split-case multistage, BFW service, 4x6-10, 3600 rpm, 3900' TDH, 1160 gpm, pumping temp 180° F, 12" suction line size to get velocity down to 2+ ft/s.

Ever have one of those days when you can't get any respect from nobody for nuthin'??? Arghhhh... where's my fishing pole...


Thanks!
Pete

 

[ccor]

I checked several of my references and the only one close was Cameron Hydraulic Data, which states at least 6D to an intake screen. Hope it helps.

 

[KernOily]

ccor - thanks, what page did you find that on??



Thanks!
Pete

 

[MikeHalloran]

I looked in Kent's "Mechanical Engineers Pocket Book" ca. 1901, and there was no trace of what you seek.





Mike Halloran
Pembroke Pines, FL, USA

 

[Artisi]

I don't believe that you will find what you are looking for as I don't see a requirement to have a straight run of pipe after a reducer at a pump inlet. What is important is straight, smooth, and uniform flow as already sighted in your research. However if you had a 90 degree elbow before the inlet, it might be necessary, depending on its orientation to the inlet flange, to have a straight run of pipe to the inlet.

ccor's comment re 6D from the pump inlet (Page 1-26 Cameron 16th Edition) is for intake screens prior to a pump inlet, an entirely different scenario.

A posting in the piping and fluids forum might generate some interest for you.

Naresuan University
Phitsanulok
Thailand

 

[CRG]

Have you contacted your pump manufacturer and ask them if there needs to be any minimum length of straight pipe on suction side for them to guarantee the performance indicated in their pump curve.

“Standard practice normally provides for a straight run of pipe of several pipe diameters (2 to 5) at the suction and the discharge of pumps,” Reference: WRC 449, Guidelines for the Design and Installation of Pump Piping Systems, by Vince A. Carucci & James Ft. Payne.

 

[25362]

Looking from another angle (repeating discussions on previous threads) a centrifugal pump may be subjected to cavitation during starting, a time when the liquid in the suction line must be accelerated.

The liquid is accelerated by a force resulting from the difference in total head at the inlet to the suction pipe and at the impeller inlet.

This force must be able to accelerate the liquid to its full velocity within a (short) time equal to the period required for the pump to attain its full operational speed.

From that we may infer that to allow the inertia forces to elude cavitation on starting one could act as follows:

[•] Reduce the mass of liquid in the suction line, ie, for a given diameter, reduce the length.
[•] Increase the difference between NPSH available at the inlet to the suction line and the NPSHR at the impeller eye.
[•] Start the pump (not the propeller type) against a closed valve which should be opened slowly to allow enough time to accelerate the liquid in the suction line. Long discharge lines causing resistance to flow may also act as a partially closed valve. When the liquid in the discharge line attains full flow before the liquid in the suction line does, impeller cavitation occurs.
[•] Lengthen the time needed by the pump rotating parts to reach full speed. A greater time interval, would enable more time to accelerate the liquid in the suction line to avoid cavitation.

Sam Yedidiah's Centrifugal Pump User's Guidebook - Problems and Solutions (Chapman and Hall) has a chapter on Pump System Layout, that may interest you.

 

[electricpete]

Know and Understand Centrifugal Pumps by Bachus and Custodio Elsevier Copyright 2003

Page 245 "You should respect 10 pipe diameters before the first elbow in the suction piping (Figure 17-16). Example: If the pump has a 6 inch suction nozzle, you should have 60 inches of straight pipe before the first elbow."

The figure shows a run of pipe which is clearly on the pump suction and identified as 10 pipe diameters long.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[electricpete]

"Pumping Station Design" Copyyright 1998 Butterworth Heineman (A member of the Reed Elsevier group)

Page 645 (Section 22.2 - Avoiding Vibration Problems) - "Pump Suction Piping - If possible, use straight suction piping up to the pump suction nozzle, and keep any required bends in a single plane. As a minimum, avoid piping reducers or bends within 5 to 10 suction pipe diameters of the pump inlet, unless the pump has been specially designed to accommodate such fittings closer to the pump (as in some vertical nonclog pumps)"

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[electricpete]

"Pumping Station Design" Copyyright 1998 Butterworth Heineman (A member of the Reed Elsevier group)

Page 645 (Section 22.2 - Avoiding Vibration Problems) - "Pump Suction Piping - If possible, use straight suction piping up to the pump suction nozzle, and keep any required bends in a single plane. As a minimum, avoid piping reducers or bends within 5 to 10 suction pipe diameters of the pump inlet, unless the pump has been specially designed to accommodate such fittings closer to the pump (as in some vertical nonclog pumps). If in doubt, seek the manufacturer's advice."

Page 646 - "Net Positive Suction Head - To minimize hydraulic forces for all types of pumps, it
is important to operate the pump with sufficient NPSHA
(see Section 10-4). In addition to causing cavitation,
inadequate NPSHA can excite unexpectedly high vibrations
at various natural frequencies. From Bernoulli's
equation, any increase in local velocity decreases static
pressure and thus increases the cavitation potential, so
it is important to keep the pump inlet flow velocities
low (preferably below about 2 m/s or 6 ft/s) and evenly
distributed. Hence, a large suction pipe (at least equal to
the pump suction flange ID) should be used. Pipe
reducers should preferably be at least five pipe diameters
upstream of the pump flange. Maintaining this distance
from more aggressive suction disturbances such
as valves and elbows is even more important. Besides
avoiding cavitation, this practice also ensures well-distributed flow velocity and pressure at the inlet and
throughout the pump, and thus minimizes the hydraulic
excitations in general. High velocity in the discharge is not usually important because the pump increases the
static pressure well above the cavitation point."

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[jet1749]

McNally advises 10 X suction pipe diameter for sraight pipe length from closest fitting. Other sources advise similar (although 5 X diameter for larger sizes)

 

[KernOily]

Guys thanks for the replies and the cites. Exactly what I needed.

I found an article by Bill Krutzsch (he who sitteth at the right hand, along with Igor Karassik ) entitled "Hydraulic Design Considerations for Pump Suction Piping" from Power and Fluids Magazine, Volume 8, Numbers 3 and 4. (Awesome stuff if you're a pump nerd. Which I know you all are.) His main concern in the article is the elimination of 'swirl' into the suction flange. In the article, he provides much hard information on the location and orientation of reducers, elbows, and tees.

The pump manufacturer (Sulzer-Bingham) has recommended 5 to 7 diameters of straight pipe of the same size as the suction flange. But in the next sentence, he says it's OK to place an elbow AND a reducer right at the suction flange, as long as I provide the 5 to 7 diameters of straight pipe upstream of the elbow/reducer. Huh??? To me this is a contradiction.

My current proposed suction piping layout has a 12" suction lateral to get the velocity down to about 3.3 ft/sec. Then I have a 12" 90 down, a 12" 90 horizontal, then a 12x8 concentric, then 7 diameters of straight 8" into the suction flange. The pump is a 4x8-10, 9 stage, with double suction first stage.

I am using a 12x8 concentric because the suction line falls continuously from the suction tank to the suction flange and the literature says that a concentric induces less swirl than an eccentric. Since the line falls continuously from the source, I shouldn't have the issue with vapor pockets forming in the upset part of the reducer, because any gas breaking out of solution in the suction line should have vented back upstream.

I'd like to hear any ideas/thoughts/snide comments/cheap shots you guys might have on this proposed configuration.



Thanks!
Pete

 

[Artisi]

Although the 2 x 90's sounds like a pretty nasty arangement and will result in some awfully disturbed flow the 7 x pipe diameters should go a long way to straightening it out prior to the inlet flange - and bear in mind that you are using HS-C pumps which usually have the advantage of an fairly long inlet approach to the impeller allowing for more flow straightening.



Naresuan University
Phitsanulok
Thailand

 

[bimr]

Page B.82 of Piping Handbook by Nayyar:

Minimum of 3-4 diameters of horizontal straight pipe between pump suction and first elbow. The eccentric reducer may be included in this straight section.

 

[pennpiper]

Pete,
I commend you for the way you asked your question. You are one of the very few (maybe the only one) who has given us some specific information about your pump. (Pumps are horizontal split-case multistage, BFW service, 4x6-10, 3600 rpm, 3900' TDH, 1160 gpm, pumping temp 180° F, 12" suction line size to get velocity down to 2+ ft/s). Most of the people who ask questions do not give us the information (tools) we need to give qualified answers.

Following the average question, almost all of the people who give answers, give answers that are too general and too vague. The answer to most questions like the question really should be very specific. The hazard here is not so much with the original questioner but with all the novice pipers and others out there who read an inappropriate general answer to a vague question and then accept it as gospel and apply it to all similar sounding situations.

The problem here is that there are so many different types, styles and configurations for pumps that the “popular rule” (10 diameter before the suction flange) makes sense for maybe only 5% of the total family of pumps and no sense for the other 95%. We could also bring up all the different types of pumps that we know about now and we would still miss the new latest state-of-the-art pump.

So, in my opinion the “10D” rule does not apply to all pumps.

I can understand the basic logic for using the “10D” rule on the typical straight in, end suction API style pump. However, even with this type of pump we have to ask if an “inducer” is included on the pump impeller is the desired smooth flow destroyed?

I do not believe it applies to any of the many types of positive displacement pumps. It is just not an issue.

I also do not believe it applies to the vertical “inline” centrifugal pumps or the typical “API”, top suction, top discharge pump. I also do not believe that it applies to your specific pump.

My reasoning, based on my more than 45 years experience in the piping field is this: Why go to all the trouble and extra cost to provide the “10D” configuration on one side of the pump flange gasket and then on the other side the pump case contorts into a complex reverse curve before the liquid comes into contact with the impeller.

You gave us the specifics about your pump but not all that is important. I suspect that is a side suction pump. This suction nozzle is well below the centerline of the pump. A cut-away of the pump would no doubt show that the inlet path is a very complex shape.

You also told us that the commodity is Boiler Feed Water (BFW) which is a tough fluid to handle do to the potential to boil and create bubble cavitation problems. I would guess that the Deaireator Drum is located directly or near directly 25 to 30 feet above the BFW pumps. The pump suction line should drop down to the pumps in the most direct and minimum length possible with the following exception. The suction line will be operating at a very high temperature and therefore needs to have proper flexibility built into the configuration. The last drop should be minimum distance from the suction nozzle. There are two potential options in my book. There should only be the elbow the reducer and then the pipe flange (fitting-to-fitting). The alternate is a reducing elbow then the pipe flange. Under the elbow (same in either case) there should be a properly designed (secondary) pipe support.

Pete, I would be interested in your thoughts on this.