Smart questions
Smart answers
Smart people
Join Eng-Tips Forums
INTELLIGENT WORK FORUMS
FOR ENGINEERING PROFESSIONALS

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.

Minimum straight length requirements before suction nozzleHelpful Member!(5) 

arvi (Nuclear) (OP)
25 Nov 05 15:50
We have two identical centrifugal pumps (same model,size 4x3x10; 4" suction, 3" discharge)installed in our power plant. Suction line is 6". We intend to interconnect the suction lines (6") of the two pumps for redundancy in our operation. Because of the space restriction we are considering to join the two 6" suction lines with a 4" line. There won't be any problem from NPSH point of view (it is flooded suction and I have done the calculations).

Any advise on minimum straight length requirements before suction nozzle. In our present configuration, suction nozzle is app. 24.5" upto the valve in suction line (including valve length, valve is diaphragm type. There will be another 3"-4" upto the new branch connection.

Also, can there be any problem with increased velocity close to the suction because of reduction in size from 6" to 4".

Thanks in advance for your advise.   



    
jbmr (Petroleum)
26 Nov 05 3:24
Most companies set a maximum velocity at pump suction which is  generally between 6ft/s and 10ft/s for liquid at vapour/liquid equilibrium.

If you use high in line velocities you are more likely to have problems during pump start up and shut down phases because of transient phenomena such as hammer effect and cavitation pocket generation.
1969grad (Mechanical)
27 Nov 05 14:12
Ideally, use 5D of straight pipe before the pump to straighten out the flow pattern.
arvi (Nuclear) (OP)
28 Nov 05 9:23
Thanks for your suggestions. I know ideal configuration is 7-8D OR atleast 5 dia. This is again debatable.This 5D is 5 times the dia of nozzle or dia of suction pipe??? In most of the cases suction pipe is larger than suction nozzle.

And secondally in cases where 5D is not practical, what sort of calculations to do to justify?? Calculating the velocity??

Thanks
checman (Mechanical)
28 Nov 05 15:26
arvi It is all about the velocity and its effect on the flow profile. I would review the system using the velocity head. Turbulence in your suction piping can give you unlimited, unpredictable and unexplainable grief. I have seen similar attempts that worked ok and others that have compromised the system.

Regards checman   
Helpful Member!  JJPellin (Mechanical)
28 Nov 05 15:52
I work in a large oil refinery.  Our specifications require 8 times the pipe diameter straight run.  We don't have any specific calculations to justify this standard.  As noted in some of the replies above, the affect is unpredictable.  We constructed a coker unit in 1984 that had a row of end suction pumps with less than 1 pipe diameter straight run into each pump suction.  2 pumps out of about 20 are extremely unreliable.  These pumps have hydraulic problems that cause them to vibrate violently at one flow rate but run smoothly at another flow rate.  Straight pipe runs are expensive and take up lots of space.  But the more the straight run is shortened, the more likely you are to have this sort of problem.  And once the unit is built, it can be nearly impossible to correct a problem such as this.  In another unit, built in 1993, we adhered to the specifications in all of the pumps.  And the pumps in this unit are the most reliable in the refinery.  A group of a dozen pumps ran for almost 10 years before we had the first bearing failure.  
Helpful Member!  PUMPDESIGNER (Mechanical)
30 Nov 05 17:35
The required straight run of pipe connection to the suction nozzle of a pump can only be determined by modeling, which is not possible in most situations, therefore the rules of thumb trying to protect ourselves.

The concept is simple though, provide a smooth laminar flow with clean flow lines onto the impeller eye.

When in doubt, increase the straight run if possible before any other method such as flow straightening vanes.  Use vibration analyses before and after, while operating under continuous maximum flow conditions to see if you caused a problem.

PUMPDESIGNER

arvi (Nuclear) (OP)
1 Dec 05 9:50
Thanks friends for your input. Just to mention these two pumps are working fine for last twenty years. However we are now considering to connect the suction lines of these two pumps to add redundancy in our operation. In our present configuration we have ecc. reducer 6 x 4" just before the suction nozzle and then 6" diaphragm valve immediately before reducer (although butterfly valve was  better but we don't intend to change it for now).

Earlier I was thinking to connect these two lines with 4" line but now after thinking about the possible problems with change in velocity, I am considering it to be 6" by adding 6 x 6 tee. Because of space restraints this tee has to immediately before the valve.

The flow will now have app. 28" of (sort of) straight portion after sharp turn of 90 deg. in tee. 28" includes reducer length and valve legth. Theoritically speaking this 28" is not straight portion because there is flow pattern disturbance in valve and reducer. But is much better situation as compared to immediately after any bend. And 28" is app. 7 times the suction nozzle dia OR app. 4.5 times the suction pipe dia.   

Any thoughts/suggestion???

Thanks in advance for your inputs.

Helpful Member!(2)  Zapster (Electrical)
1 Dec 05 14:34
The concept is simple though, provide a smooth laminar flow with clean flow lines onto the impeller eye. ?????

Pumpdesigner, this is neither possible nor desirable for most applications.  Most piping is designed for turbulent flow, including pump inlets.  At best, one should try for an concentrically uniform, when mathematically averaged, velocity profile.    
PUMPDESIGNER (Mechanical)
1 Dec 05 18:37
Zapster,
I am aware of the field problems of applications, and I certainly do not condemn those who struggle with this issue due to lack of space, retrofitting, etc.
I helped a guy last year with a 150hp fire engine pump problem.
No way could they get anything close to a flow run onto the impeller eye, looked like a sardine cannery put that pump in.  Tight 90-degree ell right up against the suction nozzle.  I could not have made it better, no room on the truck.  There are many situations where you are correct.

But, we get the job done in our equipment frequently. We get our long term customers to have their architects right at the very front end of the job give us ample room, instead of shoving us into a closet. We also use a lot of turbines, which we package in our own flow tubes, and the architects just have to give us what we need.

The payoff is huge, so we never stop trying because it is a real pleasure to have a pump operate silky smooth from low to high flow, no vibration, no bearing problems, no motor problems, for many many years.

PUMPDESIGNER

Helpful Member!  Zapster (Electrical)
1 Dec 05 21:09
Pumpdesigner,

I think it is important to get the terminology correct regarding laminar flow. The historic definition of laminar flow exists in fluid with an associated Reynolds number less than or equal to 2000.  I would not want to see anyone trying to resolve a pump problem for most fluids by trying to get a Reynolds number less than 2000. To put this in perspective, a 4” sch 40 pipe would have an upper velocity limit of less than one inch per second to be considered laminar flow for water at room temperature.  This would be less than 3 gallons per minute for the 4” pipe.  This low velocity is neither practical nor desirable for most piping applications.  

For most water applications the design velocity is somewhere between 3 to 8 feet per second.  A four inch pipe flowing water with a Reynolds number at 2000 is 36 times slower than the typical lower limit of a design velocity of 3 feet per second.  Do not try to fix pump problems by limiting flow to the laminar flow velocities.  
Artisi (Mechanical)
2 Dec 05 2:30
Pumpdesigner and Zapter

I think what Pumpdesiner is saying is - straightened flow-ie, have enough straight pipe after bends, elbows, valves etc to at least allow the flow to get straighten out a bit before the pump inlet.  

In the case where you cannot afford or don't have the luxury of a straight run of pipe, then flow straightening at the pump inlet must be used.

Naresuan University
Phitsanulok
Thailand

PUMPDESIGNER (Mechanical)
7 Dec 05 17:41
Zapster,
You are correct about the term laminar flow, I misused the term, thank you for pointing that out.

So the correct statement would be something like,
Obtain the lowest possible practical velocity and clean flow lines into the impeller.

PUMPDESIGNER

vanstoja (Mechanical)
12 Dec 05 11:29
For further discussions of of the criticality of centrifugal pump and adjacent component suction and discharge pipe spacing see vanstoja and others replies in the following related threads:
  thread407-106384
  thread378-78272
  thread407-129352
  thread378-109206
mw1009 (Petroleum)
6 Apr 06 18:47
What is the maximum Suction Specific Speed for Hydrocarbons on a Vertical Canned 4 Stage Pump-Suction Nozzle 30" in diameter.

Artisi (Mechanical)
6 Apr 06 22:15
"mw1009 (Petroleum) 6 Apr 06 18:47  
What is the maximum Suction Specific Speed for Hydrocarbons on a Vertical Canned 4 Stage Pump-Suction Nozzle 30" in diameter."
 
You should start a new post for this - not add it to this posting.

Naresuan University
Phitsanulok
Thailand

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