Problem Statement (Challenge)

      • What (type of machine): Four Large Variable Speed Vertical Pumps in Filter Influent Pump Station (FIPS) service: Figure 1.
      • Where: Northeast USA
      • Why (problem/reason): The newto pump systems did not meet the specified vibration acceptance criteria according the ANSI/HI 9.6.4 Rotodynamic Pumps for Vibration Measurements and Allowable Values and lacked adequate separation margin for the main structural natural frequencies: Figure 2.
Figure1A-1

Figure 1A

Figure 1B

Figure 1C

Figure 1A: Measurement location and directions specified by the Hydraulic Institute Figure 1B: MSI’s VibVue® Motion Magnification Video (MMV) Figure 1C: Operating Deflection Shape (ODS) test animation based on test data, not FEA. The video and animation depict the worst-case vibration scenario: 11 mils peak-to-peak in displacement, exceeding the specified maximum of 7.2 mils pk-pk, and 0.20 in/s RMS in velocity, surpassing the specified maximum of 0.17 in/s RMS.

 

Figure2A-1

Figure 2: Amplification factor versus frequency showing an example of 10% Separation Margin for variable speed machinery. The green Nf mode is unacceptable as it is within the running speed range.  The modes indicated by blue and red meet or exceed a 10% separation margin criteria.  Learn More.

Work Performed
  • Testing Methods: Experimental Modal Analysis (EMA), Continuous Monitoring (CM) of vibration, Operating Deflection Shape (ODS), Figure 1A), and VibVue® Motion Magnified Video: Figure 1B.
  • Analysis Methods: ANSYS and Creo Simulate Finite Element Analysis (FEA) computer programs with “as built” pump system models calibrated using test data.

Background and Timing (MSI’s involvement was initiated in early 2023):

  • Circa 2013:The initial specifications were produced without including an independent dynamics assessment of the pump system, which would have aimed at reducing vibration risks.
  • Late 2015: Completion of 60% of the plant design was achieved leading to the purchase of six constant speed and four variable speed pumps.
  • Spring 2021 to Summer 2022: Despite significant efforts to modify them, the four variable speed pumps were operating but continued to face challenges in meeting the specified vibration- related requirements. It was believed by some of those involved at the time that major modifications to the discharge piping or the pump foundations might be necessary to meet the vibration and separation margin requirements. 

  • January 2023: MSI contacted by the plant owner. MSI proposed and carried out our standard specialized testing and analysis process focused on characterizing and addressing the issues of high vibration and inadequate separation margin. 

  • March 2024: The plant owner reported that the issues related to vibration and separation margin had been successfully resolved.


    MSI Test Results indicated the following:

    1. Two out of four pumps exhibited vibration in velocity that exceeded the specified vibration levels: Figure 3.
    2. Three out of the four pumps exceeded the allowable overall vibration displacement: Figure 4.
    3. All four variable speed pumps demonstrated a separation margin of less than 10%, indicating that resonance was either the primary cause or a significant contributor to the high vibration issues. Without correction, it was also assessed that vibration issues would likely worsen as the plant aged.
    4. In all cases, the identified problematic natural frequency was the classic above-ground 1st bending mode (i.e., reed frequency), which was too close to the operating speed frequency or within the running speed range: Figure 5.
    5. The discharge head mounting flange was found to be excessively flexible, compromising the stability and efficiency of the pump operation, as evidenced by the video and animation: Figure 1.
    6. The motor mounting flange exhibited signs of lifting off or sliding in relation to the motor support flange, likely due to potential looseness at this connection: Figure 1.
    7. The Dresser Coupling was not acting as a flexible joint: Figure 1.
Figure3A

Figure 3: The overall vibration levels, indicated by the red line and measured with accelerometers, surpassed the allowable limits as specified by ANSI/HI 9.6.4 for Rotodynamic Pumps in Vibration Measurements and Allowable Values, which are represented by the purple line measured at the top of the discharge head. The black line denotes the pump speed in RPM.

 

Figure4AFigure 4: Because the maximum rotational speed was 600 rpm or less (black line), the overall vibration in velocity was integrated into displacement to compare with the amplitude specified ANSI/HI 9.6.4 Rotodynamic Pumps for Vibration Measurements and Allowable Values also in displacement (purple line).  Excessive vibration leads to shorter operating times between repairs.

 

Figure5AFigure 5 – A Frequency Response Function (FRF) plot is a crucial result of a properly conducted experimental modal analysis (EMA) or an impact test that includes the use of an instrumented impact hammer. The plot displays a lightly damped natural frequency at 8.5 Hz, which is superimposed on the running speed range of 8.1 Hz to 10 Hz.

Recommended Solutions: 

MSI recommended a series of solutions based on Finite Element Analysis (FEA) calibrated data from field tests. These solutions involved evaluating various “what if” scenarios to identify the most effective fixes for the pump system's vibrational issues. The most highly rated potential solutions scenarios were:

  1. Add an MSI specified weight amount and reduced stiffness to lower the natural frequency (not chosen by the customer).
  2. Repaired previous cuts in the discharge elbow and added struts specified by MSI to the wall to raise the natural frequency (this was chosen by the customer and struts were added to the top of the motor).

The resolution to the vibration issues were based on a thorough FEA evaluation that provided options to either shift the offending natural frequencies below or above the running speed range to mitigate resonance effects (without floor or discharge piping modifications). In both cases, the key recommendation from the analysis suggested that most or all of the previously attempted trial and error fixes be reversed. Following this, additional FEA modeling conducted by the pump manufacturer, in collaboration with the plant owner, led to the decision to shift the offending natural frequencies above maximum running speed with a strut solution approximated: Figure 6.


Figure6A

Figure 6: This FEA model illustrates a successfully implemented solution. Struts were added from the top of the motor to the ceiling above the pump, reversing previous unsuccessful attempts, such as cuts in the discharge head. The model comprehensively includes adjacent piping and floor details, reflecting MSI's standard practices. It also accounts for variations in water levels and the pump's operational speed range. Notably, FEA models of vertical pumps that omit considerations of surrounding floor and piping are, at best, misleading.

 

MSI Is A Water And Wastewater Industry Long-Term Partner

 

REAL-WORLD EXAMPLES AND CASE STUDIES

MSI In Action

Case Study

A Cure for the Common Cold – in Turbines

A steam turbine at a waste-to-energy facility in the Northeast underwent a scheduled repair, at which time the turbine bearings were reworked, after which the turbine periodically experienced very high amplitude shaft vibrations.

Case Study

BFP Fluid Drive Troubleshooting

A boiler feed pump (BFP) driven off of the main steam turbine via a fluid drive was experiencing high vibration levels leading to frequent replacement of the fluid drive bearings.

Case Study

Aeroderivative Gas Turbine Modifications to Solve a Problem

MSI was contracted to understand and help solve a high vibration problem on an aero-derivative gas turbine driven generator.