×
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

Pumps in Parallel
2

Pumps in Parallel

Pumps in Parallel

(OP)
Good day,

My Boss threw an idea at me which while sounds correct in theory, I'm not sure how accurate it is practically. We have 6 vertical pumps in parallel and there's a discharge gauge on each. We have no individual flowmeters, just a total flow of all 6 pumps from the header. My boss was suggesting to check the power required by the pump and use this value on the power curve to obtain a flow value and then plot this on the performance curve with the pressure value from the gauge. Is using the power to obtain a flow value accurate?

Thanks

RE: Pumps in Parallel

2
Probably not.  It depends on the power curve. We have used exactly the technique that you describe for some big cooling tower pumps where it is impossible to get individual flow measurements.  And, in general, it has worked to help us understand which pump is the weakest of the bunch. But, we don't use it to do detailed analysis of each pump condition. It is more a screening tool.  If the total flow drops below some limit, we use the motor amps to decide which pump to pull for overhaul.  But, there are a number of potential problems.  Some big vertical turbine pumps have a power curve with a hump in the middle. There could be two points with the same power.  Some of these pumps have a horsepower curve that increases to shut-off.  Some of them have a hump plus a rise to shut-off. There could be as many as three points with the same power.  So, in order to know where you are on the curve, you need to trend the motor power on an ongoing basis.  If you have a baseline reading on a freshly overhauled pump, and trend it monthly, you can distinguish where you probably are on the curve as the amps decrease and then increase again.  But, if you wait until you have a low total flow problem and then go out to take amps, you may not be able to tell.  Of course, if your power curve is continuously rising as flow increases, then it is easier.  If the power curve has a long flat section, it may be difficult to distinguish where you are at.  In any case, beware of inaccuracy.  The power curve is based on a pump in new condition, running over a range of flows.  You are looking at a pump that may be in degraded condition with extensive internal recirculation through open clearances. You cannot get an accurate measurement of pump performance.  You can only get a relative measurement of which pumps are probably stronger and which ones are probably weaker.   

Johnny Pellin

RE: Pumps in Parallel

Yes JJPellin is correct, but you can get a good indicator of the flow if you take the motor power, and put that onto the pump curve, and then plot the head on the curve also. this will give a good indication on the flow.

Hopefully you have got a performance test curve for each pump. if so then you are better getting the head generated from each pump using a pressure gauge at the pump discharge, and having a pressure gauge at the pump suction.

take the suction pressure from the discharge pressure and you have the pump generated head. plot this onto the pump perf curve and this should give you the flow.

this is better than trying to use the power to get the flow.

RE: Pumps in Parallel

There is one problem with the use of discharge pressure to estimate flow.  If the system is set up the way ours are, the discharge lines are in a common manifold with no significant pressure drop between them.  In other words, by definition, you will always measure the exact same discharge pressure for all running pumps.  This provides no usefull information to diferential between a strong pump and a weak one.  

Johnny Pellin

RE: Pumps in Parallel

You will also want to verify that you don't have impeller problems (wear, wrong size, wrong trim, clearance, etc.) by making sure that each pump's dead-head head is about what you would expect it to be.  You also need to be pretty sure of your efficiency curves for both the pump and the motor before the power draw itself becomes a meaningful indicator of flow.

I have used portable Polysonics Doppler meters before to get flow measurements (circa 1990s) when I otherwise had no device to rely on.  I rented them typically for about $100.00 per day.  There is probably even better technology today.

Regards,

SNORGY.

RE: Pumps in Parallel

If you have a check valve in the outlet of the parallel pumps, try taking an x-ray (sideways, like you do for a pipe weld) of the six check valves.   

Whichever is shut (or only partially open)  is attached to a failing/failed pump.    

RE: Pumps in Parallel

Most of us don't have radioactive material laying around. smile

**********************
"The problem isn't finding the solution, its trying to get to the real question." BigInch
http://virtualpipeline.spaces.live.com/

RE: Pumps in Parallel

Even if you were able to do this in spite of the obstacles mentioned above, in order to obtain the power you will need to know the power factor accurately at the pump motors.  Do you have the capability to determine that?

I once had to use the technique you want to use in a power station to corroborate the flow data I was getting from the condenser DP and from the TDH measurements across the pumps and fortunately they knew their station PF, so the technique worked and the flow measurements checked out.

rmw

RE: Pumps in Parallel

rmw makes an excellent point. If you hope to have useful accuracy, you will need to know the motor efficiency, power factor and actual operating voltage.   

Johnny Pellin

RE: Pumps in Parallel

(OP)
Well we have a 701 relay that allows us to read the power straight off it so we don't have to calculate it.

Another question on this topic, what's the theory behind lowering of the water level in the basin, and the apparent increase in the horsepower required by the pump. When the level is raised, the power required is less. Why is that?

RE: Pumps in Parallel


The Brake Horse Power, that is the work done by the pump is

BHP = (Ht * Q *  SG)/(3960 * Eff)

Where:

Q = flow rate
SG = liquid specific gravity
Eff = pump efficiency

Increasing the suction head Hs (reservoir above centerline of the impeller) the total differential head Ht decreases:

Ht = Hd – Hs (being Hd the total discharge head)

So decreasing Ht, BHP decreases as well.
 

RE: Pumps in Parallel

(OP)
I was considering that same calculation you performed above, however, in order to find the total head, Ht, won't your suction head have to be referred to the same datum as the point you're taking your discharge pressure. In this case, since they're vertical pumps as stated above, wont it be like a suction lift? since you'll have to then add the distance of the water level below the datum to the point you're taking you're discharge from?

RE: Pumps in Parallel

For vertical pumps the datum point is the entrance eye of the impeller. The formula for the BHP stands still and so lowering the water level in the basin you do reduce suction head and increase BHP.

RE: Pumps in Parallel

For a vertical radial flow impeller, lowering the sump level will probably lower the power as the head will increase, flow will reduce as the pump moves left along its curve, of course this is relative to the pumps efficiency as the duty point moves to the left.
If a very mixed or axial flow impeller, power will probably increase as the sump level reduces and head increase due to a rising power curve as head reduces for high specific speed pumps.

To say that power will increase for radial flow impellers as head increases assumes that flow remains the same which is unlikely.  

RE: Pumps in Parallel

Artisi is absolutely correct, but the key is in the formula provided by Ione.  In order for BHP to increase, you cannot just increase head and leave all other variables constant.  Based on the formula, the product of head times flow divided by efficiency must increase for BHP to increase.  This requires a particular shape to the curves.  Many mixed flow vertical turbine pumps have a tendency to increase BPH as the flow decreases.  Most radial flow vertical or horizontal pumps have a tendency to decrease BHP as the flow decreases.   

Johnny Pellin

RE: Pumps in Parallel

impeller,

In your statement you mentioned you don't have individual flow meters... do you have one flow meter on the common header, or a better way to measure total flow?

If all the pumps are identical, and all are reading approx. the same discharge pressures, individual flow of each pump should be directly proportional (ie. 4 pumps in parallel, indiv. flow is 1/4 total flow.)

Did you know that 76.4% of all statistics are made up...

RE: Pumps in Parallel

dabluffrat,
If all pumps are connected to the same header and operating then all pumps will have the same identical discharge pressure irrespective of individual flowrates.

RE: Pumps in Parallel

if i have two pumps in series and a recycle line comes after the last pump, how am i going to go about the shutting down procedure?  

RE: Pumps in Parallel

renin,
Start a new thread for your question there is no point to tagging it onto this thread.  

RE: Pumps in Parallel

Artisi,

And if all pumps are identical and have a good head rise to shut off, then they'd be approx the same flows for that given pressure.

Did you know that 76.4% of all statistics are made up...

RE: Pumps in Parallel

the answer is in your pump curve, power curve etc.
have a good look at the shape of the curves and you should be able to tell if measuring the power can help you identify the weakest pump. This is usually difficult and can be inaccurate since pumps discharging on a common header influence each others operating point.
the most accurate method would be to use an ultrasonic flow meter; provided you have a good straight run of piping on the discharge of each pump.  

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