Linear pump/reciprocating pump

[howearthmademe]

Hi All,

I am designing a test rig for a reciprocating pump and motor assembly to obtain the system curves etc...

The problem is how am i going to increase the inlet pressure for this pump (manufacturer suggested an inlet pressure around 10psi-100psi)?

If i add a pump at the inlet say a centrifugal pump to reach the required head, what effect would that have on the performance of the reciprocating pump?

Any suggestions would be greatly appreciated!!

Regards

 

[ports394]

This sounds similar to something I needed to do with two centrifugal pumps. I needed to achieve 40 GPM at 400-600 PSI.

The pump supplier said that 1 centrifugal pump, even with a 100 HP could not accomplish it because it was so far to the left on the performance curve. So what they setup was a smaller, high pressure pump to feed into the big pump. The small pump got something like 50-100 psi at 40 GPM. Then the big pump stepped it up to the 400-600 psi as we needed.

They just added the pressures.

 

[micalbrch]

What kind of reciprocating pump do you have, plunger or piston? Duplex or triplex?

To use a centrifugal pump as a booster pump is common and will not have any negative effect to the performance of the reciprocating pump. But the centrifugal pump must provide higher flow than the reciprocating pump. How much depends upon your answers to above mentioned questions.

 

[curtis26]

Charge Pump Sizing

1) Flowrate: Size the flowrate of the centrifugal charge pumps for 150% of the reciprocating pump’s flowrate. For example if the reciprocating pump's flowrate is 200 GPM then the CP pump should be sized for 300 GPM.

2) Impellor Size & Motor Speed: Charge pumps are not a constant flow application. The discharge flowrate of the charge pump must accelerate and decelerate to match the variation in flow of the reciprocating pump that it feeds. Small impellors operating at 1750 RPM tend to slip through the fluid when acceleration is needed. (Similar to car tires slipping on wet pavement.) Although the pump curves may show that enough head will be available at 1750 RPM with a small impellor shock loading may occur. To avoid this charge pumps should have larger impellors running at either 1150 RPM or 1450 RPM. For example a 6 x 5 x 14 with a 13.5” impellor (14” max. on this model) would work well, where a 10 or an 11” impellor would not be recommended.

3) Head: Make sure that you take any loses in the suction piping or losses for high vapor pressure into consideration, and be sure that charge pumps is sized to produce a minimum of 150% of the NPSHr.

4) Location of Charge Pump:
A charge pump should be placed as close to the supply tank or source as possible. For this reason many charge pumps are a long ways away from the location of the reciprocating pump so they can be closer to the tank.

5) Drive for on Charge Pump: If you place the charge pump next to the reciprocating pump you can belt drive a speed increaser to the non-drive side of a double extended crankshaft. If you don’t have a double extended crankshaft, you can direct couple an electric motor to the charge pumps. If electricity is not available, you can belt drive the charge pump from the PTO of the Gas or Diesel Engine that drives the reciprocating pump. Be sure to add the charge pump BHP to the HP requirement when sizing the main drive if you are belt driving it to the PTO or the opposite end of the pump crankshaft.

 

[curtis26]

The effects on performance...

The purpose of the charge pump to a reciprocating pump is simple to ensure that the pump has adequate suction to avoid cavitation. Request and NPSHr curve from the manufacture to establish how much head is required at the suction of the pump. Since Reciprocating pumps don't produce head, the pressure at the suction does not change the flowrate or discharge pressure of the reciprocating pump.

 

[howearthmademe]

Hi all,

thank you for your response!

The NPSHr is not specified by the manufacturer because it is a bespoke design for a particular application. Hence the supplier did not perform any tests on similar pumps before and the supplier is relying on us to test the pump. The pump only has 2check valves for openening and closing and a piston in rod config...

I am not sure if i understood the concept properly regarding the flow rate of the charge pump... to my understanding if i choose the flow rate too high the motor has to overcome the force generated by the inrush flow rate hence the power requirements on the motor would increase!!

I was advised not to use a centrifugal pump because this may have an adverse effect on the check valves...

i feel more confused than ever before... can anyone point me in the right direction please.... maybe i am not understanding the performance of the pump correctly..

Thank you all for the support..

 

[micalbrch]

You have a single acting piston pump with only one piston and one suction and one discharge side check valve? If yes, what check valves - ball checks or spring loaded cone or plate valves?

I'm not sure why you need a charge pump at all if you have a single acting piston or plunger pump. These pumps usually have an excellent suction capability and do not necessarily require a positive inlet pressure (= positive gauge pressure). Do you tell us the pump brand and perhaps model? I think that is important to understand the problem.

 

[Compositepro]

Use an elevated or pressurized feed tank and recirculate the pump flow back to the tank. 10 psi is 20 ft of head.

 

[howearthmademe]

Hi,

Just double checked the pump, the pump has a long plunger in a rod and has two check valves.

The manufacturer said we need pressure at the inlet and the outlet is set to 4kpsi. i cannot raise the feed tank higher than 2 m above ground level because i simply dont have space in workshop...

the pump is running at high speeds and it is a learning curve for manufactuerer as well.

i was thinking in the lines of curtis26 a pump at the inlet should do the trick but other engineers reckon that the pump performance is compromised due to flow rate generated by centrigal pump...

Thanks all for replying..

 

[chicopee]

I don't know how much water you need to develop the reciprocting pump performance curve, however, you may want to look at a pressurized water reservoir such as a pressurized propane tank fully filled with water from a municipal water source such as a hydrant connected to tank with a fire hose at around 50 psig.

 

[curtis26]

If you size the charge pump for the maximum instantaneous flowrate (for a single acting piston pump this would be twice the flowrate, since 50% of the time the piston is moving backwards not producing flow) then you will maintain a constant pressure on the piston pump inlet and will not have any “inrush”. The addition flow the CP is rated for will recycle around the casing of the CP (charge pumps with concentric impellers in the casing work best for this application). Duplex, triples, quintuplex pumps have variations in flowrate that are all less than 50% which is where the 150% flowrate guide line comes from. You will need a CP rated for 200% since you have a simplex pump otherwise you starve the pump during the forward stroke which will cause cavitation (but shouldn't hurt the motor).
The CP's will not generate flowrate. It will produce a constant head at the suction side of the pump. A piston pump is a positive displacement devise, so as long as its NPSHr is met it is going to have the same discharge flowrate at 10 PSI suction as it does at 100 PSI suction (and the same if its NPSHr is achieved with a boosted pump, elevated tank, etc).
If it makes any difference, I have been to a few multiplex pump manufactures’ test facilities, and all of them had CP charge pumps to feed their test stand for performance testing.

 

[Pumpsonly]

You never mention the capacity of the recip pump which is important to decide the set up. Without this info you will get over kill porposal.
10-100PSI suction pressure is a very wide range. Pump manufacturer should be able to estimate more closely the NPSP for the pump.
You can install a pulsation dampener / accumulator as close as possible to the recip pump suction so that you do not need to oversize the charge pump flow rate and also less pulsation in the suction flow.

 

[micalbrch]

I still do not understand what kind of single acting reciprocating pump needs a booster pump on the suction side - unless it is running that fast that it would cavitate without booster pump or sufficient inlet pressure. It would be good to know a little bit more about the pump (photo?).

 

[DubMac]

WOuld think the simple elevated tank is preferable to CP, since it would be constant inlet pressure for all flows.

If you insist on CP, then make sure the curve is as flat as possible to avoid as much head change over flow range. Centrifugal with a steep curve will apply head pressures that vary with flow. A performance curve that is essentially flat throughout the flow range will deliver essentially contant head.

Of course the CP will have some sort of NPSH requirements as well.