This project examines the root cause of increased vibrations observed in the control valve 670FC005-VC at Shell Pernis following its replacement with a new valve from a different manufacturer. The resulting vibrations led to a critical mechanical failure at the connection between the valve stem and the actuator.
The objectives of the analysis were to systematically determine the origin of the vibrations, assess potential failure mechanisms, and recommend effective mitigation strategies. The study was conducted by Dynaflow Research Group and focused exclusively on field measurements, comparative valve analysis, and dynamic response evaluation under actual operating conditions.
Problem Description
Prior to its replacement the control valve 670FC005-VC operated for over 25 years without any vibration-related incidents. Immediately after commissioning the new valve, significant vibrations were detected, leading to the loosening and eventual disconnect of the bolted actuator stem-plug connection. This mechanical failure necessitated a comprehensive investigation to prevent recurrence and restore reliable operation.
Analysis Approach
A methodical, multi-step procedure was implemented to identify the root cause of the vibrations and subsequent failure:
- Process and Flow Condition Assessment: Evaluate the potential for cavitation or other flow-induced phenomena at the observed process conditions.
- Valve Comparison: Analyze geometric and mechanical differences between the legacy and replacement valves, including flow direction, dimensions, and internal features.
- Onsite Vibration Measurements: Carry out accelerometer-based vibration measurements at multiple locations on and around the valve and actuator under representative operating scenarios.
- Data Analysis: Use frequency domain analysis to correlate measured vibration spectra with possible mechanical eigenmodes and flow-induced excitation sources.
Cavitation Assessment
A methodical, multi-step procedure was implemented to identify the root cause of the vibrations and subsequent failure:
- Process and Flow Condition Assessment: Evaluate the potential for cavitation or other flow-induced phenomena at the observed process conditions.
- Valve Comparison: Analyze geometric and mechanical differences between the legacy and replacement valves, including flow direction, dimensions, and internal features.
- Onsite Vibration Measurements: Carry out accelerometer-based vibration measurements at multiple locations on and around the valve and actuator under representative operating scenarios.
- Data Analysis: Use frequency domain analysis to correlate measured vibration spectra with possible mechanical eigenmodes and flow-induced excitation sources.
Comparative Valve Assessment
A detailed comparison revealed significant differences between the old and new valves. The prior valve had a plug length of 686 mm, trim size of 89 mm, and an estimated plug stem diameter of 33 mm, with a specified Cv of 175. By contrast, the new valve featured a shorter plug (320–420 mm), a marginally larger trim (97 mm), a much thinner plug stem (16 mm), and a lower specified Cv (161, with even lower values recorded in prior documentation). Crucially, the direction of flow through the valve was reversed with the new installation: the old valve operated with upward flow, while the new valve used downward flow.
The change in flow direction altered how fluid interacted with the valve internals. In the original design, upward flow entered the cage downstream of the plug, which provided a stabilizing effect and suppressed vibration. With downward flow in the new valve, the cage is upstream, offering little to no stabilization, potentially amplifying oscillatory flow phenomena that were previously mitigated.
Vibration Measurement Methodology
Vibration measurements were performed at six distinct locations: bottom, rings at plug attachment, middle, and top of the actuator, plus two positions on pipe supports near and away from the valve. Accelerations were logged in three orthogonal directions (vertical, lateral, and axial relative to piping) and analyzed for two valve openings (38% and 34%), simulating altered operating conditions by varying upstream pressure while maintaining constant flow rate.
Frequency spectra were generated using Fourier Transform techniques to identify dominant vibration modes and their spatial distribution across the valve assembly.
Vibration Measurement Results
The analysis of the vibration data indicated the following:
- The highest acceleration amplitudes were observed at the plug attachment ring (location 2), with a pronounced peak at approximately 525 Hz. This peak was also detected at the valve bottom (location 1) but with much lower amplitude, indicating the vibration was highly localized and resonant within the valve body.
- At frequencies below 120 Hz, acceleration peaks were consistently observed across all valve locations, suggesting the presence of a local excitation phenomenon within the valve. These lower-frequency peaks shifted upwards when the valve was closed further (from 38% to 34% open), consistent with increased flow velocity through the reduced flow area.
- The amplitude of vibrations increased as the valve opening decreased, correlating with higher local fluid velocities.
- Vibration measurements on pipe supports (locations 5 and 6) were negligible, indicating the vibrations did not propagate significantly into the wider piping system, nor did large-scale system eigenmodes contribute meaningfully to the observed behavior.
Dynamic and Modal Analysis
A finite element analysis (FEA) conducted by the valve supplier identified the second natural frequency (eigenmode) of the plug at 534 Hz, matching the frequency of the localized high-amplitude vibration observed in the measurements. The spatial distribution and frequency of this vibration indicated it was a mechanical resonance of the plug itself, excited by flow-induced phenomena such as vortex shedding and turbulent oscillations within the valve body.
At lower frequencies (44 Hz at 38% open, 54 Hz at 34% open), the shift in peak frequencies was in line with theoretical expectations for flow-induced vibration scaling with velocity. These lower-frequency oscillations were likely the result of flow instabilities, but did not appear to coincide with actuator or system natural frequencies.
Root Cause and Failure Mechanism
The investigation established that the increased vibrations were not caused by cavitation, actuator resonance, or piping eigenmodes. Instead, the root cause was a combination of:
- Valve Internal Flow Dynamics: The change in flow direction and the absence of effective flow stabilization from the cage in the new valve design created conditions favorable for localized, high-frequency flow-induced excitation.
- Mechanical Resonance: The frequency of this excitation matched the plug’s natural frequency, leading to resonance and high local vibration amplitudes.
- Fatigue and Loosening: Sustained resonance and vibration at the actuator-plug connection led to gradual loosening and eventual disconnect of the bolted joint.
Recommendations
Two principal mitigation strategies were identified:
- Flow Stabilization: The valve manufacturer should confirm the flow direction and assess the presence and position of flow-stabilizing features, such as the cage. If possible, reversing the flow direction or modifying internal features to reintroduce downstream flow stabilization could attenuate vibration.
- Operational Adjustment: Operating the valve at higher openings by adjusting upstream/downstream pressures will reduce flow velocity through the valve bore, diminishing excitation amplitudes.
- Mechanical Modification: To disrupt the high-frequency plug mode, an additional guide bearing could be installed mid-span between existing supports, modifying the plug’s mode shape and increasing resistance to resonance.
Conclusions
The vibration and failure of control valve 670FC005-VC after replacement was conclusively linked to the interaction of altered internal flow dynamics and mechanical resonance of the plug. The absence of effective flow stabilization in the new valve configuration and the match between excitation and plug natural frequencies were critical contributors. Future replacements or modifications should prioritize careful alignment of valve internal geometry and flow direction with proven, vibration-suppressing design principles, and mechanical support structures should be reviewed to prevent resonance amplification at critical components.