Merits of condition monitoring (CM):

• Maintenance is based on actual condition, so components are serviced just before failure, avoiding both unnecessary overhaul and unexpected breakdown.
• Early warning of developing faults, reducing the risk of catastrophic failure and consequential damage.
• Extends machinery life and reduces downtime and maintenance cost.
• Improves safety and reliability.
• Allows maintenance to be planned at convenient times (e.g. in port).
• Optimises spare-part usage and reduces inventory.
• Provides a record/ trend of machinery condition for survey and management.

Demerits of CM:

• Requires investment in monitoring equipment (vibration analysers, sensors, software) and training.
• Requires skilled personnel to interpret the data correctly.
• A fault may develop rapidly between monitoring intervals and be missed.
• The monitoring system itself requires maintenance and calibration.
• Initial set-up (baselines, measuring points) is time-consuming.
• May not detect all fault types (e.g. sudden failures).

Comparison with other regimes:

• Breakdown (reactive) maintenance: cheaper to set up but risks unexpected failure, downtime and damage; CM avoids this.
• Planned/ preventive (time-based) maintenance: services components at fixed intervals regardless of condition, so components may be overhauled unnecessarily (waste) or fail before the interval; CM is more efficient but requires more data/ skill.
• CM is a predictive/ condition-based approach that combines the reliability of preventive maintenance with the efficiency of only servicing what needs it.

• Gathering: Vibration data is collected using accelerometers/ vibration transducers mounted at designated measuring points (bearing housings, machine feet). Data is collected either by a portable data collector/ analyser (taken to each point) or by permanently installed sensors feeding a monitoring system. The data includes overall vibration level and the vibration spectrum (frequency content).
• Storage: The collected data is stored in a computer-based system (a vibration database/ software), organised by machine, measuring point, and date. Baseline (reference) spectra and historical trends are stored for each point.
• Evaluation: The software analyses the data by:
• Comparing the overall vibration level with alarm/ limit values (ISO/ classified standards).
• Performing frequency analysis (FFT) to identify the frequency components and relate them to specific faults (e.g. unbalance at 1x rpm, misalignment at 1x/2x rpm, bearing faults at high frequencies, gear faults at gear-mesh frequency).
• Trending the vibration levels and specific frequency components over time to detect the onset and rate of deterioration.
• Comparing the current spectrum with the baseline to identify new/ growing components.
• Generating reports/ alarms to alert the engineer to developing faults, so maintenance can be planned.