17 replies to this topic

[Prasanjit]

I have this configuration wherein the two Centrifugal pumps are in parallel.

Existing pump : 1000 m3/hr
Head : 162 m
Shut off head : 196.54 m
Motor Rating : 560 kW
RPM : 1800 

This existing pump will run in parallel to the new pump whose parameters i mentioned below:

New pump in parallel with above : 2200 m3/hr
Head : 107 m
Shut off head : 129 m
Motor Rating : 730 kw ( yet to be decided)
RPM : 1800 / 1200 (1 vendor quoted)

My question, is it possible to run this pump in parallel with the existing pumps since nothing is matching with the existing pump except the liquid service.
Nor the shut off head, nor the flow, neither the rpm for 1 vendor.

What do you think, is it still possible to connect those pumps in parallel ?

I mean there is no VFD that is envisaged here in the process ( I am a Mechanical Engineer )

As I am a mechanical engineer, could you please let me know what should I check in the New pump offers by Vendors looking from parallel operation perspective.

Thank you Sir; Hope my question is clear to you.

Thanks Again !!!

[breizh]

Prasanjit ,

 

To help you we need to know the full story , what are you supposed to achieve with these 2 pumps running in // ?

 You should tell  us :

             the operating point supposed as per calculation  ( Head vs Flowrate) and supply  the curves of the individual pump  to draw the pump curve for 2 pumps in //.

 

Consider the handbook attached , in particular page 101 for pump in // .

 

 

Note : the rpm of each pump has nothing to do with your current problem.

 

hope this helps ,

 

Breizh

 

 

 

[Prasanjit]

Dear Sir,

 

Please find the attached Curves for new and existing pumps.

 

Also the screen short of PFD attached herewith.

 

The pumps are attached in parallel so that the existing pump of 1000 m3/hr will add to the new pump flow of 2000 m3/hr and give me 3000 m3/hr.

 

Regards,

Prasanjit

Attached Files

•  New pump curve.pdf   233.82KB   52 downloads
•  Existing PumpCurve.pdf   227.83KB   38 downloads
•  PFD.JPG   97.48KB   4 downloads

[breizh]

Hi .

TDH should be about 115 m for 3000 m3/h based on your pumps' curves , will it be sufficient? 

Someone in your facility should have performed the hydraulic calculation of the system to ensure you will get the right equipment .It should be done by a process engineer . 

 

 

Hope this helps

Breizh

Edited by breizh, 29 November 2015 - 08:36 AM.

[Mahdi1980]

Prasanjit

To inspect more exactly if two pumps can give 3000 m3/hr you should have the curve of the system. In the other words, the resistance of pipelines, fittings and valves at pump's discharge to the end users determine if these two parallel pumps can handle the required capacity. It depends on resistance of system.

Anyway according to your data and performance curves it seems that they cannot deliver 3000 m3/hr.

[Prasanjit]

what should be done so that they can deliver 3000 m3/hr ?

Is there any way out, apart from VFD usage ?

[latexman]

Someone must calculate, or cause software to calculate, the system curve (pipe, fittings, etc.) before any reasonable decisions can be made on "what should be done".  Has the system curve been calculated?  What is the result?

[Prasanjit]

No Sir, not yet. The system curve is yet to be calculated.

I have asked my process person to do so already.

[Prasanjit]

Thank you Mr. Breizh for your inputs.

However, I want to know whether the pumps arranged this way*(attached in pictures) are going to meet my total flow requirement of 3000m3/hr together.

 

If this way is not going to meet, then what additional things I will have to see/check in the pump offer to ensure that the new pump in configuration with the existing pump can give me the required flow of 3000 m3/hr ? eg. shut off head of both the pumps etc.

[breizh]

Hi ,

Hydraulic calculation is the answer based on isometrics (not PID) .

 

Note : You can download free on Internet PSIM (pump system improvement modeling tool) a software to perform hydraulic calculation .You can learn from examples and build you own model based on Isometric.

 

Good luck ,

Breizh

Edited by breizh, 29 November 2015 - 11:27 PM.

[Prasanjit]

Thank you Sir for your inputs.

[latexman]

If the hydraulic calculations/system curve has not been done yet, how did a pump specification get made?  It seems out of order.

[Mahdi1980]

Prasanjit

 

Pleas let me know if you have bought the new pump or just ordered it?

Regardless of system curve, If your vendor has suggested the new pump, I should say that the new one is not compatible with the existing. Because of the big difference in their heads ( the new one has an shut off head of 129 m in comparison with the operating pressure of existing one that is 162 m of head) and if you look at the performance curve of the existing pump, you will recognize that at heads lower than 129 m the curve has been cut off. It means that  whenever two pumps will be parallel, at the head as low as 107 m (operating head of new pump) the existing pump will have a low efficiency and it will be damaged sooner than the normal conditions.

However, if the new pump has been purchased and you have to make it parallel with existing one, OK. we may find a solution.

For calculation of system curve, please inform me about pipeline diameter, length and number of valves and types of them and the total height difference between the level of the pumps and level of end user. In a brief word, Isometric as Breizh said.

 

Mahdi

Edited by Mahdi1980, 29 November 2015 - 09:36 AM.

[shawkat06]

Prasanjit,

 

This may help you.

 

www.sundyne.com

 

Regards

[fallah]

Prasanjit,

 

The main issue is two different pumps intended to work in parallel...

 

A tricky method...might lead to a reasonable estimation:

 

For each pump calculate the delta P along the pump discharge line from discharge nozzle up to connection point for several values of flow rates. Then by subtracting this delta P from corresponding pump's discharge pressure calculate the pressure at connection point for each flow rate. Draw the curve of pressure vs flow rate at connection point for each pump in the same graph. If the two curves cross each other they will both operate at the flow rates summation of which would be equal to the flow at crossing point, otherwise the pump with lower head might not be able to pump anything...

 

A simulator such as EPANET can be employed to get a better result...

Edited by fallah, 30 November 2015 - 04:29 AM.

[yadavrk103]

Prasanjit,

From the data it is very much clear that the parallel pump operation is not possible because of following reasons:

1. From existing pump rated differential head it seems that the required head for 3000 m3/hr will be much higher than 162 m

2. New Pump Shutoff head/ rated differential head is much lower than your existing pump differential head.

 

Thanks

[Padmakar Katre]

 

Hi,

 

What is the sizing basis of new pump, how you have arrived at diff head of 107 m when existing pump diff head is 162 m. If you are throttling the extra head from existing pump (162-107 or further lower) in a control valve and based on revised system resistance curve for total 3000 m³/h flow the revised head is 107 m only then yes you can operate this new pump in parallel to existing with few considerations:

 

1. protection of new pump in the event of no flow from the system the reason here is design pressure is lower than existing,

2. I presume the overall line sizes plus other instruments like flow orifice, control valve are being changed since the change in flow is 3 times the original design.

 

As you mentioned, you are a mechanical engineer, ask these questions to your process team who has specified these conditions in the pump datasheet which is shared with bidders/vendors. There might be some thoughts to request such offers from bidders where the conditions for new pump are not matching with existing.

[shan]

You should not conclude that your new parallel pumps system design rate is 3,000 m³/hr by simply adding the existing pump design rate (1,000 m³/hr) and new pump design rate (2,000 m³/hr).  Generally, you should generate a parallel pumps curve by adding flow rates at each corresponding head to integrate with your discharge piping system curve for determination of the operating point.  However, there is no common head point in your case.  The existing pump run-out head is 145 m, but the new pump shut-in head is 129 m.  There is no way the new pump can start up if the existing pump is in operation.

Edited by shan, 03 December 2015 - 10:55 AM.