Booster Pump feed Positive Displacement Pump in Series 2

[bmw318be]

Hi Sir,

Here's the situation: We retrofit a fuel system, We use centrifugal pumps that are sitting directly at the storage tanks to pump to existing day tank which is all at level 1, pressure is less than 2 bar maximum.

Now, our end user is build a new building with the day tank on the top level 4, 50 m vertical height. Pressure losses around 6 bar.

Assuming, distance between New Pump room of PD pumps is too far away, we would need to tie in to the existing centrifugal pump in order for the PD to have sufficient NPSHa as distance is a concern and it is not economical to build a new line.

Centrifugal flow is 600 LPM at shut off head of 50 m

Centrifugal discharge line is having a calculated frictional loss up to PD pump inlet is 22 m

PD pump rated flow is 200 LPM at 5 bar

Question 1:
If the centrifugal feed the PD pump which has a lower flow of 200 LPM, would the system pressure of centrifugal follow to 200 LPM since the maximum displacement of the PD is only 200 LPM ?

Question 2:

Assuming the centrifugal is able to operate at 200 LPM curve, can I say that the centrifugal pump able to feed well to the PD pump ?

Question 3:
Would the Inlet pressure before the PD pump inlet is 22 m, which is the frictional loss or it is close to 0 barg since PD pump would withdraw 200 LPM from 400 LPM at no pressure:

Question 4:
i see some discussion, for series, we need to have a balance flow, would the centrifugal on the inlet considered balance as it would adapt to the flow the PD pump displacing

Question 5:
Can I considered the inlet pressure as 0 barg so the NPSHa shall be around 10 m absolute. Assuming the inlet is packed


Question 6:

Is it recommend to install low pressure switch at 1 bar on the inlet of the PD pump so it would protect the PD pump in case no flow.

https://1drv.ms/b/s!AshU1On9Vy6fiasvViVIgVb4EWlrvA?e=tR84SG

 

[LittleInch]

If ever a post needs a drawing to explain this, this one does.

Please add a drawing or sketch of the system showing elevation and distances, pipe sizes, etc. [pre][/pre]

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[bmw318be]

Hi LittleInch,

Thanks for reading this post, the sketch is attached.

https://1drv.ms/b/s!AshU1On9Vy6fiasvViVIgVb4EWlrvA?e=tR84SG

 

[LittleInch]

Here it is for ewveryone else

OK
Question 1:
If the centrifugal feed the PD pump which has a lower flow of 200 LPM, would the system pressure of centrifugal follow to 200 LPM since the maximum displacement of the PD is only 200 LPM ?
The flow would be a total of about 200LPM. The pressure at the inlet to your PD pump could be about 20-30m head

Question 2:

Assuming the centrifugal is able to operate at 200 LPM curve, can I say that the centrifugal pump able to feed well to the PD pump ?
Yes, but you need to check the pump curve of your pump. Only knowing the shut off head isn't enough. You need to know the head at 200, 400 and 600 LPM as well

Question 3:
Would the Inlet pressure before the PD pump inlet is 22 m, which is the frictional loss or it is close to 0 barg since PD pump would withdraw 200 LPM from 400 LPM at no pressure: It would be appox 20-30m becasue the centrifugasl pump delivers a more or less constant head but allows flow to vary. So take off your 22m head loss for friction and elevation and you're left with about 20-30m which is what the pressure at the inlet to the PD pump will be.

Question 4:
i see some discussion, for series, we need to have a balance flow, would the centrifugal on the inlet considered balance as it would adapt to the flow the PD pump displacing
Yes. Centrifugal pumps are able to cope with a wide flow range at more or less (~20%) the same head.

Question 5:
Can I considered the inlet pressure as 0 barg so the NPSHa shall be around 10 m absolute. Assuming the inlet is packed
inlet pressure to your PD pump will be a lot higher than 0barg - 20-30m head is my guess.

Question 6:

Is it recommend to install low pressure switch at 1 bar on the inlet of the PD pump so it would protect the PD pump in case no flow.
No as this would be tripping all the time. You can get flow through a PD pump if there is a differential pressure, but it probably won't be harmful.

Does this help?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[bmw318be]

Hi Littleinch,

thanks for your valuable response, just clarrifications for me to better understand the theory.

Question 1:
If the centrifugal feed the PD pump which has a lower flow of 200 LPM, would the system pressure of centrifugal follow to 200 LPM since the maximum displacement of the PD is only 200 LPM ?
The flow would be a total of about 200LPM. The pressure at the inlet to your PD pump could be about 20-30m head

Q: my understanding that the PD pump would move the liquid from low pressurre zone to high which is at discharge of Pd pump at 50 m head. The PD suction pressue, Ps1 would be < Pd1, shown in the illustration, I would thinm Pd1 should br around 20 m which is approximately frictional losses of centrifugal pump at maximum displacement of pd pump could run.

However another thought, since pd pump is self priming, once it is packed, would there is a tendency booster pump forced spin the screw rotor faster, in the end it would adjust thr curve to run out condition which of much bigger tha 200 LPM flow.


Question 2:

Assuming the centrifugal is able to operate at 200 LPM curve, can I say that the centrifugal pump able to feed well to the PD pump ?
Yes, but you need to check the pump curve of your pump. Only knowing the shut off head isn't enough. You need to know the head at 200, 400 and 600 LPM as well

q: you mean from the flow head curve of my booster pump ? . or tbe curve would change when we rub in series with pd pump .

Question 3:
Would the Inlet pressure before the PD pump inlet is 22 m, which is the frictional loss or it is close to 0 barg since PD pump would withdraw 200 LPM from 400 LPM at no pressure: It would be appox 20-30m becasue the centrifugasl pump delivers a more or less constant head but allows flow to vary. So take off your 22m head loss for friction and elevation and you're left with about 20-30m which is what the pressure at the inlet to the PD pump will be.


q: 22 m when we assumed that the discharge length is up to the pd pump inlet, however there is tendency which i am not certain, assuming the centrifugal pump able to pushed morr than the displacement of the pd pump, and it able to feed spinning the rotor faster, would the booster pushed to the run out condition since displacement of pd would be then higher. this Just some possible assumption.

Question 4:
i see some discussion, for series, we need to have a balance flow, would the centrifugal on the inlet considered balance as it would adapt to the flow the PD pump displacing
Yes. Centrifugal pumps are able to cope with a wide flow range at more or less (~20%) the same head.


Question 5:
Can I considered the inlet pressure as 0 barg so the NPSHa shall be around 10 m absolute. Assuming the inlet is packed
inlet pressure to your PD pump will be a lot higher than 0barg - 20-30m head is my guess.

Question 6:

Is it recommend to install low pressure switch at 1 bar on the inlet of the PD pump so it would protect the PD pump in case no flow.
No as this would be tripping all the time. You can get flow through a PD pump if there is a differential pressure, but it probably won't be harmful.


q: there is scenario where the ground existing tank call for fuel that all the isolation valve would be opened,once this opend the discharge line of booster would then be very low about 10 m,pd pump would then automatically run at let say 50 LPM, as fuel call at 2nd floor called for.

Would the booster pump still able to reach inlet of pd pump when scenario is run ? . if it is, I would like to ensure thr minimum pressure is reached to protect the pd pump from dry run.


Theoretical Performance curve

https://1drv.ms/b/s!AshU1On9Vy6fias3O7kzArcF21rxIQ?e=j5IPCh

 

[LittleInch]

OK,

Your pump curve you posted is not what I would expect a normal centrifugal pump curve to look like. There are some that have that sort of drop off, but from 50m at 0 flow to 10m at 600 in straight line is quite strange. Is that from the vendor? why call it "theoretical"??

1) PD will only pump the flow within about 10% regardless of the inlet pressure. In inlet pressure is high and outlet low, then all that happens is the power consumption goes down. It doesn't overspeed.
2) I mean the flow / differential head curve of the booster pump
3) No
4) -
5) -
6) Sorry I misread your question. If you mean a low pressure trip, i.e. the pressure needs to be > say 10m to make the PD pump work, then yes, this is a good idea otherwise the PD pump will continue to pump a more or less fixed amount and will lower the inlet pressure in an attempt to do this which could result in cavitation
Difficult to say what would happen unless you calcualte all the pressure drops and flows to see what happens. But you need a proper pump curve from the booster pump..


Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[bmw318be]

Hi Littleinch,

thanks for your response, the curve is the only curve from the manufacturer, it is very old pump, like turbine pump.

I am curious, is the pd pump acted as back pressure valve when it run together as booster pump, so at discharge of pd pump pressurre would be higher so the boaater can not feed than the displacement of the pd pump.

isnt that the differential pressure of pd pump, (discharge Pressure - suction pressure), The lowwest pressure is right before PD Pump inlet, high pressure would be at right before booster discharge run. The total back pressure of booster pump is in the function of total length, the furtherest from booster is the lesser the pressure as it was pumped out by the pd pump.

 

[LittleInch]

Ah. Turbine or axial flow pumps work very differently to centrifugal pumps so please be careful not to confuse the two in future.

In extreme circumstances the of pump might act that way but not here.

Don't understand your last point.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[bimr]

These fuel systems are normally arranged in a loop arrangement with the loop discharging back into the tank. A PRV is fitted prior to the tank discharge to control the pressure in the loop.

 

[bmw318be]

Little Inch,

Thanks, sorry let me rephrase the last question.

isn't that the differential pressure of pd pump, (discharge Pressure - suction pressure), The lowest pressure is right before PD Pump inlet, high pressure would be at right before booster discharge run. The total back pressure of booster pump is in the function of total length, the furtheest from booster is the lesser the pressure as it was pumped out by the pd pump.

Because you mentioned before the pd pump, the pressure is 20-30 m head, however I am trying to understand, just right before the pd pump, isn't it shall be lowest ? . This allow me to understand what low pressure switch to set and size.

 

[LittleInch]

I think you don't quite understand the physics going on here.

For the same flowrate, yes the inlet pressure into a PB pump will reduce the further you go.

The pressure you get out of the PD pump is dependant on the resistance to the flow or the static head. The pressure difference will impact the power the pd pump takes.

If you're setting a low pressure switch, just set it to the minimum that that particular pump requires, so probably just above atmospheric pressure.

Does that answer the question?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[bmw318be]

Little Inch,

Let me use illustration to ask my doubts.

PD pump sucking the liquid that feed by booster, the pressure at the Inlet of pd pump should be approximately 0.

https://1drv.ms/u/s!AshU1On9Vy6fiatJkTcwyb7Vn_xB9w?e=FjXZ3d

 

[LittleInch]

bmw,

The pressure at the inlet will vary depending on the total fuel flow from the axial pump.

If nothing else is using fuel, it could be in the range 20-30m.

If lots of other users have flow at the same time, then it could get close to zero pressure. It doesn't help that your start pump is an axial flow / turbine pump which has much lower differential head at higher flows.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[bmw318be]

Hi littleinch,

thanks, i just curious is it across the pipe it is within 20 to 30 m ? assuming no other ask for fuel.

can you hel me elaborate the physic behind this constany pressure

 

[LittleInch]

Your pump according to your curve puts out about 40m head at 200 gpm.

According to your data at 200 gpm your friction losses from main pump to PD booster pump is 22 m.

so that leaves 18-20m as the head at the inlet to your PD pump assuming it is the only pump working and you're flowing at 200 GPM.

It is simply coincidence that your friction losses equal the inlet head into the PD pump.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Nutzman]

Hi bmw318be,

If I read your system correctly, you have a centrifugal pump, pumping into a PD pump?
The centrifugal pump will only supply as much flow as the system resistance will allow. In essence, if you had an open ended pipe, the pump would pump as much fluid, as it could until it met system resistance. So if we fitted a valve at the end of the line, by opening/closing the valve, we could control the flow.
In this scenario the PD pump is the "valve".
Although the centrifugal pump could pump 600LPM through the pipe (providing the PD pump were not there), with the PD pump it is restricting the flow to 200LPM. This means that the centrifugal pump is pushed left on it's curve to the point of 200LPM (increased head).

Looking at the curve, for a flow of 200LPM, the head pressure is ~41m
You mention the friction loss's of 22m.
That means the inlet pressure at the PD pump is 41-22 = 19m
The PD pump can generate 200LPM @ 5bar.
That means the discharge pressure on the PD pump will be 5 + ~1.9 = ~6.9 bar.
If you restrict the discharge of the PD, it will keep pumping until something goes Pop!!
The increased pressure and "flow" of the centrifugal pump, will not cause the PD pump to pump more flow. NIMHO.

 

[LittleInch]

"That means the discharge pressure on the PD pump will be 5 + ~1.9 = ~6.9 bar."

Errr, well the PD pump is set up to pump a certain amount of product. The outlet pressure will be , within reason, that required to overcome static and frictional losses in order to pump the set flow.

If the end point only need 3 bar ( or even 1 bar) discharge pressure to pump 200 lpm then that's what it will do. If it needs 10 bar to pump 200 lpm then that's what it will do unless there are pressure relief valve preventing it or the motor blows up.

but other than that yes, that's a good summation. IMHO.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[bmw318be]

Nutzman,

i think this term not correct

That means the discharge pressure on the PD pump will be 5 + ~1.9 = ~6.9 bar.

pd pumo is based on the differential pressure = discharge - suction pressure= 5- 1.9 = 3.1 bar

if my frictional loss of pd on dischsrge is 5 barg, then the differential is 3.1 bar, so can we say we would see 3.1 bar on dischsrge.


i still do not get the inlet is means the inlet pressure at the PD pump is 41-22 = 19m

ao npsha is 19 m ? . i expect the inlet should be approximately to 0 barg, positive as there is dynamic flow keep moving on the inlet of pd pump,

 

[bmw318be]

I managed to use this illustration of pressurr drop in series to confirm pressure at end is very low and near the pump is high

https://photos.app.goo.gl/DsMMWmHqiF9eNyUo8

https://youtu.be/F4VM_Xlp-SU

 

[LittleInch]

Well it isn't quite what you have, but if it helps you understand then good.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.