I'm curious if anyone else has ever witnessed piping vibration in screw pump piping.

It is oil service, the pumps are progessive cavity type.

Downstream pressure pulses were measured around 9kPa, and the branch lines shake under certain operating modes. We don't know the upstream magnitudes.

It appears from the vibration study that mechanical natural frequencies, acoustic natural frequencies, and the pressure pulse frequencies are close to each other (within 1.1 to 1.2x) and is causing the shaking.

Given this is the first time I've experienced vibration problems in screw pump piping I'm wondering if anyone else has experienced this and if perhaps it would be prudent/justified to perform acoustic studies on future screw pump piping systems. Or perhaps establish a method to determine which systems require vibration analysis.

I would never have believed the pressure pulsation from a screw pump would be enough to initiate vibration.

Paul,

The 9 kPa pressure pulse equates to 1-1/2 psig which does not seem to be much of a cause for vibration. The progrssive cavity types of pumps are usually small, so the piping size would be small. If the pump is run at variable speeds, then there could be resonance of piping and pump speeds.

Is the small diameter of piping so flexible that it is resulting in the observed vibration? I would think adding some more restraints could take care of the vibration problem. A Modal piping analysis could show where restraints could have benefit and reduce the vibration. Add enough restraints to get the piping first Modal frequency above 5 to 6 Hz.

It could be similar to seismic restaints on piping where each run of pipe has at lease one guide and one line stop restraint.

Thanks Richard for your answer..

The allowable loads on the pumps were much lower than API610 so the piping is somewhat 'flimsy', however the MNF was 7 Hz (in a PSV branch, the suction and discharge are above 14 Hz) so it should have been okay.

The piping is 8 NPS-10 NPS, it's a mainline booster station. We suspect there are other issues, so I'm just trying to get a census on others experience with these type of pumps. A sanity check!

The apparent resonate frequencies are in the 35 to 40 Hz range which caused some doubt since even piping on a 1200 rpm compressor would only be designed for MNF>24 Hz.

Paul,

The progressive cavity pumps that come to mind are small Moyno with Nps-1 or Nps-2 size connections. Is the pump that you are working with at size Nps-10 like a Tuthill?

The cross section area of Nps-10 will have force of 120. lbs at the elbows and changes of direction due to the 1-1/2 psig pressure pulse. It is more significant but not huge at the 120. lbs. If the pulse pressure is being read from a gage indicator, then the 35 Hz pulses won't show peak pressure which could be several multiples of the 1-1/2 psig. Also, the pump discharge flow could have a negative pressure as flow halts between the pressure pulses of flow. The force at albows could actually be +120. lbs and then -120. lbs alternating along with higher peak values.

Pulsation dampener air bladder tanks might be effective in reducing the pressure pulse peak amplitudes. Some vendors are Blacoh Sentry, Flexicraft Hydro, Kemlon, and Nixco. The piping could have suitable locations for air pipe-stands for expansion volume to take pulses.

There were some Braun-Luebbe pumps causing piping vibrations from their pulsations on outlet flow. It was caused by the timing gears which put output flow peak nearly same from each lobe. After the timing gears were corrected the synchronized outlet flow was much smoother with less pulsation. Is there anything similar on your progressive cavity pump that could be checked for timing synchronization?

Just to give you some support in the field of pulsations: I don't know how this level has been determined, but in general one cannot speak about a global pulsation level in the piping. That would require measuring points at regular distances, and especially at certain locations like each closed valve. In practice this is not possible, and the only way to find out how pulsation levels are distributed acrosss the pipe system is through a pulsation analysis (simulation). So I assume that the pulsation level was recorded at one point only.
It is good to realize that when pulsations exist you probably also have to deal with acoustic resonances, which will amplify pulsations. It is well possible that at one spot you record 9kPa but at an another spot you would have recorded 20*9=180 kPa, which is already significant.

Further, pulsations itself do not cause pipe vibrations, but only do dynamic pressure differences (think of standing wave patterns).

Hope this helps you in understanding how to deal with 'pulsations' in general.

Thanks Jan, much appreciated...we did do a pulsation simulation, and given the known downstream measured value, the simulation was calibrated to produce the same level. Upstream values were much higher according to the simulation.

The challenge isn't understanding the concepts of pulsation, I am trying to get a consensus or gathering of experience with this particular style of pump. To me this shouldn't normally be an issue.

I will also share some other observations from my non-stress colleagues. I have a process engineering friend that explained that if the discharge pressure is too high backflow could result in excessive pressure pulsations. I don't completely understand how this works but it did shed some light on the issue as we suspected there are some issues related to how the pumps were being operated versus the design intent. He also said that if these type of pumps are sized and operated properly they run very smoothly.