Plunger pump vibrations

[Murec]

We have a plunger pump operating at 335 rpm.
Some time ago a backup was added. The backup pump is smaller and operates at 786 rpm. None of the pumps has a pulsation dampener.
The flowmeter on the common discharge works OK with the larger pump, but shows erratic readings with the smaller pump.
Adding a pulsation dampener was suggested, however I suppose that it is the larger and slower pump that would cause harder flow pulsations and not the smaller pump.
There could also be a matter of mechanical vibrations transmitted through the piping, as the backup pump is installed on a steel construction.
Any ideas are welcome
Thanks
M

 

[BigInch]

Are the two pumps operated separately? Otherwise there may some addition of pulsation waves, since one pump is roughly twice the operating speed of the other.

You might check pressure pulsations against the naturally resonant frequency of the length of pipe between pump and flowmeter. Higher resonance frequencies in the range of 800 to 1200 may be more likely than those in the range of 300 to 400.

"The top of the organisation doesn't listen sufficiently to what the bottom is saying." Tony Hayward X-CEO BP
"Being GREEN isn't easy." Kermit

http://www.youtube.com/watch?v=hpiIWMWWVco

http://virtualpipeline.spaces.live.com

 

[FMJalink]

Murec,
Maybe are the acceleration forces in your suction system causing the problem. At every suction stroke of the pump the total suction line liquid has to be accelerated following the stroke of the piston of the pump. The accelerating plunger of the pump sucks on the suction line and wants the liquid in the suction line to accelerate on the same rate. The locsl suction pressure can never be larger than the vapor pressure or absolute vacuum.

If the pump piston movement caused a vapor (or vacuum) bubble to be generated at the pump inlet, temporary no liquid enters the pump. There after the suction liquid gets accelerated and enters the pump again.

Just to imagine the process indicatively, you could see the pump flowrate as Flowrate = Volume_cylinder x (sin( rotational_speed x time) + 1) . Differentiating this would give d(Flowrate)/dt = Volume_cylinder x rotational_speed x cos ( rotational_speed x time). When the product of Volume_cylinder x rotation_speed is for the smaller pump above that for the larger, you have your answer.

Kind regards,
Frank Jalink

 

[FMJalink]

Murec,
I have to correct myself. A nice glass of beer at the end of the day makes one less focused and talking less clear.

If you imagine a single plunger of the pump to move approximately in accordance with a sinus, the plunger head location could be represented as x(time) = Half_stroke x sin (rotational_speed x time). The liquid in the suction line has to follow this plunger movement during the suction stroke. In the pressure stroke no liquid enters the pump at this plunger and is decelerated.

When the plunger is again in a suction stroke it pulls again on the suction line column back to the suction vessel. The pressure can locally not become lower than absolute vacuum. If the pressure locally becomes below the vapor pressure a vapor bubble will appear. The pulling force is limited by the vapor pressure. When the plunger is decelerating or in pressure stroke the liquid in the suction line catches up and will bump on the pump.

Differentiating the indicative equation of above, the acceleration at the head of the plunger becomes: a(t) = - Half_stroke x rotational_speed ^ 2 x sin( rotational_speed x time). In order not to have a vapor (or vacuum) bubble will the liquid at the plunger (and the entire suction inline) with mass m have to follow this acceleration. The required force to accelerate this mass m is F =m x a. This force F divided by the cross section A of the suction pipe gives you an indication of the local pressure required to achieve the acceleration.

In order to reduce the risk of vapor (or vacuum) bubble creation at the suction of the pump:
• The suction mass to accelerate is reduced by placing a pulsation dampener (only the suction line part up to the pulsation dampener has to follow the plunger head), and/or
• The suction vessel pressure is increased.

The factor: Half_stroke x rotational_speed ^2 could be compared between both pumps in order to give you an indication of the difference in pulling and suction pressure. The above is only indicative, as might be clear to you.

Kind regards,
Frank Jalink

 

[micalbrch]

Murec,

You are right, the bigger pump should cause more pulsations than the smaller as its stroke volume is bigger.

Anti-vibrations pads between pump base frame and steel construction or foundation are recommended as they reduce vibrations. For the same reason a suction side expansion joint (compensator) is a nice thing to add. But the vibrations do not affect the flow meter reading or the pressure reading on the gauge.

I think Frank is right with his suggestions although some of his assumptions might not be correct. I beleive your pumps are triplex plunger pumps, correct? If so, there is flow all the time into the pump because at any time at least one of the plungers performs a suction stroke. But the suction side of the bak-up pump is much smaller if it has the same flow as the bigger pump. Not only the suction inlet but also the valves are much smaller. That means the velocity inside the pump and the accelerations are huge. Most probably the back-up pump cavitates. This would explain the erratic flow meter readings on the discharge side, too.

What can you do? I think not much. You can enlarge the suction pipe of the back-up pump but this will only help if the existing suction pipe is rather long and small in size. But as the this will not change the velocity inside the pump (in the suction manifold and through the suction valves) the effect might only be small. What was the reason for the smaller back-up up pump (smaller in size, not in flow). Let me guess: price?

 

[micalbrch]

One thing I forgot to recommend. Install a vacuum pressure gauge at the back-up pump's suction and have a look what the gauge shows when the pump operates.

 

[Murec]

Thanks to everybody for your help
M