Dear all,
As per API 610 "If parallel operation is specified, the head rise from rated point to shutoff shall be at least 10 %."
What is the reason behind this requirement and what will happen if it is 7%.
Regards
Lokesh
|
Posted 17 March 2016 - 09:21 AM
Dear all,
As per API 610 "If parallel operation is specified, the head rise from rated point to shutoff shall be at least 10 %."
What is the reason behind this requirement and what will happen if it is 7%.
Regards
Lokesh
Posted 17 March 2016 - 11:39 AM
As per API 610 "If parallel operation is specified, the head rise from rated point to shutoff shall be at least 10 %."
What is the reason behind this requirement and what will happen if it is 7%.
Hi,
Because in almost flat pump curve the total flow of parallel pumps will be highly decreased (in parallel operation of pumps the total flow in any case is lower than summation of pumps flowrates when are working individually); and if the head rise is lower than almost 10% the total flow reduction will be such that parallel operation cannot be logical...
Posted 17 March 2016 - 10:34 PM
Hi Fallah
Thanks for your reply.
I did not quite understand the explanation. Let me put this question with values.
Pump total flow rate- 1000 m3/hr, to meet flow rate 2 operating pumps are added with 500 m3/h capacity each.
Rated head- 10 m
Shut off head- 10.7 m.
As per API shut off head should be min, 11 m (10% above rated head). How the pump operation will impact if it is 10.7 m
Regards
Lokesh
Posted 17 March 2016 - 10:51 PM
Lokesh,
Have you seen a Flow v/s Head curve for a centrifugal pump ?
If yes, you will notice that the curve is little bit flatter at lower flowrate and starts to dip as the flowrate increases.
If your rated head is less than 10% of the shut-off head there are high chances that the operating point will be in the flatter trajectory of the curve, this can lead to pump operating at shut-off conditions (i.e. no flow condition) and that is not acceptable in a continuous process.
10% is a good engineering practice specified in order to keep the operating point away from the shut-off point.
Hope this clarifies your doubt.
Revert in case of any doubt.
Posted 17 March 2016 - 11:30 PM
Hi ,
Probably also good to draw the curves for pumps in // to understand the issue . Get the pumps curves and establish the curve with 2 pumps running // .
Breizh
Posted 20 March 2016 - 09:08 PM
My understanding is the characteristics of the system curve should also be studied carefully. In order to get 1000 m3 flow from two 500 m3 pump, a flat system curve is required which is very unusual. The best way to find out is to plot the system curve and the combined pump curve. Without that it is very hard to recommend anything.
http://www.pumpsands...allel-operation
Edited by 031, 20 March 2016 - 09:10 PM.
Posted 21 March 2016 - 12:44 AM
Dear Lokesh,
1. When you look at a pump curve, you have a reducing slope, non linear curve that starts at Zero flow (X Axis) and certain head (Y axis) and declines towards Zero Head at certain flow Q. It looks much like a hill that starts off flat and then declines downward towards a ground point.
2. Now the purpose of connecting 2 Pumps in Series is to increase the head produced for a certain amount of flow, and the purpose of running pumps in parallel is to produce a certain amount of head for increased amounts of flows. Speaking graphically, Series operation will increase the Y Axis (Head produced) of the curve while parallel operation keeps the same head, but extends the range of flow (X Axis).
3. If we now take a pump curve, then the head rise from the Rated point to the shutoff point is a governing parameter. It is used to check at what flow the pump will run.
So if the margin between the rated head and shutoff head is kept too low, what happens is that the pump will run at a Head condition which is very high, i.e. almost close to the shutoff head, but no flow.
4. The reason why we call it a shutoff head is because if the head produced is too high, the flow will be too low, close to almost zero. Running a pump at no flow makes no sense.
5. This is the reason that the pump is not run at heads that are too close to the shutoff head, which is why there needs to be a certain safety margin between the rated head and the shutoff head.
6. When we consider parallel pumps, the purpose as explained before is to divide the flow but produce a certain head. This arrangement here itself is already reducing the flow.
Additionally if you run the pump at close to shutoff head, the flow will be even lower, almost near zero.
7. To avoid this, a certain margin is considered as 10% of the shutoff head for good engineering practice so as to allow for a minimum flow for the pump to run rather than lead to no flow. You can run the pump at lower margins too but that would be purposeless because the flow will be too low.
8. Also take a look at the curves here for easier visualization
http://www.engineeri...rial-d_636.html
http://www.kelairpum...umpClinic35.pdf
http://www.mcnallyin...8-html/18-1.htm
Regards,
Shantanu Kallakuri
Edited by shantanuk100, 21 March 2016 - 01:00 AM.
Posted 29 March 2016 - 05:43 AM
Dear All
Thanks for your replies.
I agree that it is a good engineering practice to have 10% margin on rated head to shut off head. But for parallel operating pump operation we have to provide this as per code requirement. For single operating pump it is good to have kind of thing. Please see below the excerpt from API 610.
" Pumps that have stable head /flowrate curves (continuous head rise to shut off) are preferred for all applications and are required if parallel operation is specified. if parallel operation is specified, the head rise from rated point to shutoff shall be atleast 10%."
In revamp of a plant, To handle the revamp capacity a new pump is added in parallel (now 2 operating 1 spare) - same model no. and same head curve as existing pump. With the new flow for each pump ( revamp flow/2) the head curve shifted to left side and from rated point to shut off head margin is now 7%. I believe which is not acceptable as per code and we have to delete both existing pumps and add 2 pumps for this. Therefore just want to know the implication from where this requirement came, if it is only good engineering practice then 2 operating + 1 spare pump can be considered with client approval or it has any other reason ( operational impact etc.) to change both pumps by new pumps
Let me know your comments on this.
Regards
Lokesh
Posted 29 March 2016 - 06:52 AM
Please see below the excerpt from API 610.
" Pumps that have stable head /flowrate curves (continuous head rise to shut off) are preferred for all applications and are required if parallel operation is specified. if parallel operation is specified, the head rise from rated point to shutoff shall be atleast 10%."
In revamp of a plant, To handle the revamp capacity a new pump is added in parallel (now 2 operating 1 spare) - same model no. and same head curve as existing pump. With the new flow for each pump ( revamp flow/2) the head curve shifted to left side and from rated point to shut off head margin is now 7%.
Hi,
If before revamp the head rise from rated point to shut off for single pump has been 10%; it couldn't be changed for each of two identical pumps (new+old) due to revamp...means after revamp each pump in parallel working pumps would be operated at a flow rate lower than the rated one for each individual working pump...but still the head rise from rated point to shut off for each pump is 10%...
Edited by fallah, 29 March 2016 - 06:55 AM.
Posted 29 March 2016 - 12:00 PM
Two centrifugal pumps operated in parallel produce less than twice the flow of a single pump. The combined pump curve is different from the single pump curve. When both pumps are running, the resulting head may be higher than the single pump would see. For stable operation, each pump has to be able to start up alone and also start when the other pump is running. This requires that the pump be sized to allow for stable operation in both circumstances.
![]() Hysys Expander/compressor Input Using Dimensionless CurveStarted by Guest_Alfreedo_* , 21 Apr 2025 |
|
![]() |
||
Centrifugal Pump In Recycle Mode: Constant Power Consumption But FluctStarted by Guest_Kakka_* , 23 Jan 2025 |
|
![]() |
||
Need Help Understanding Increased Power Consumption In Injection PumpStarted by Guest_Kakka_* , 21 Feb 2025 |
|
![]() |
||
Improving Safety By Re-Routing Priming Vent Line In Centrifugal Pump HStarted by Guest_Srinivas Agastheeswaran_* , 31 Jan 2025 |
|
![]() |
||
![]() Centrifugal PumpsStarted by Guest_panoska_* , 01 Feb 2025 |
|
![]() |