A rotating asset rarely fails without warning. Bearings degrade, shafts misalign, and rotors lose balance over weeks or months, and each of these mechanical faults writes its signature into the machine's vibration pattern long before a catastrophic breakdown occurs. The question reliability engineers in Indonesia's power plants, oil & gas, mines, and manufacturing lines face is not whether vibration analysis works, but whether it is the right maintenance strategy for a specific asset, and how rigorously it needs to be applied.

1. What Vibration Analysis Actually Tells You

Vibration analysis measures amplitude, frequency, and phase characteristics of a machine's mechanical motion, then compares that signature against a healthy baseline. An accelerometer mounted on the bearing housing converts physical movement into a voltage signal. That signal is captured as a time waveform, converted through Fast Fourier Transform into a frequency spectrum, and interpreted against known fault frequencies for bearings, gears, and rotating elements.

The value of this data depends entirely on what question is being asked. A raw vibration reading confirms only that a machine is running. A spectral analysis, read by a trained analyst, can distinguish between imbalance, misalignment, looseness, bearing wear, and lubrication failure, each of which demands a different corrective action and carries a different urgency.

A study on the HP oil pump motor at PLTU Barru, operated under PT Indonesia Power, illustrates this precisely. Overall vibration velocity alone did not isolate the fault. Only after spectral analysis at the motor's non-drive end and drive end, benchmarked against ISO velocity severity standards, could engineers confirm the abnormal reading and determine whether the motor remained within acceptable operating limits [1].

2. When RCM Logic Says Vibration Monitoring Is the Right Call

Reliability Centered Maintenance (RCM) does not prescribe vibration analysis as a default. It prescribes a maintenance strategy based on failure mode, failure consequence, and the detectability of a P-F interval, the window between when a fault becomes detectable and when it results in functional failure.

Vibration monitoring earns its place when three conditions align.

Failure mode characteristics. The asset exhibits a dominant failure mode that develops gradually and produces a measurable vibration signature well before functional failure. Rolling element bearing degradation, rotor imbalance, and coupling misalignment fit this profile. A failure mode driven by sudden fatigue fracture or electrical fault often does not.

Consequence severity. The failure has safety, environmental, or high-cost operational consequences that justify the cost of continuous or periodic monitoring.

P-F interval width. The gap between detectable onset and functional failure must be wide enough that a monitoring interval can realistically catch it. A P-F interval measured in weeks supports monthly route-based monitoring. A P-F interval measured in days requires continuous online monitoring, and a P-F interval measured in hours may render vibration analysis impractical regardless of technology.

A diesel power plant operated by PT PLN in Masohi applied exactly this discrimination logic. Components were classified by failure frequency, with the highest-risk category assigned scheduled preventive intervals of five months, mid and lower categories left on repair-on-failure, and vibration and tribology analysis reserved specifically for the predictive layer of the program rather than spread uniformly across the fleet [2].

Where these three conditions are not met, RCM logic points toward alternative strategies: time-based overhaul for assets with predictable wear patterns, run-to-failure for low-consequence non-critical equipment, or redesign where the failure mode itself should be eliminated rather than monitored.

3. Building the Monitoring Program Around Criticality, Not Convenience

A common mistake in Indonesian industrial facilities is applying uniform monitoring frequency across all rotating equipment, regardless of criticality. This produces two failure patterns simultaneously: over-monitoring of low-risk assets that consumes analyst time without proportional benefit, and under-monitoring of critical assets where a monthly route misses a fast-developing fault. A predictive maintenance study on the compressor and vacuum emulsifier homogenizer at PT Kosmetika Global Indonesia built its monitoring schedule around exactly this logic, concentrating vibration checks on the two rotating assets whose failure would stop the production line rather than spreading equal attention across the plant's full equipment list [3].

The correct approach ties monitoring intensity to the criticality ranking established during the RCM analysis, then selects a technique that matches the diagnostic requirement:

• 🔒Contact us to learn about the methodology!
• 🔒Contact us to learn about the methodology!
• 🔒Contact us to learn about the methodology!
• 🔒Contact us to learn about the methodology!

The analyst's competency matters as much as the hardware, and a bearing failure investigation at PT Freeport Indonesia shows why. A conveyor drive pulley bearing failed suddenly in 2009 despite being on a vibration monitoring route. The data showed radial and axial amplitude declining steadily in the months before failure, a trend a less experienced analyst would read as improving condition. The actual cause was friction between the inner race and the shaft, which damped the vibration signal instead of eliminating the underlying defect, so the bearing kept deteriorating while the readings quietly went down [4]. A declining trend is not automatically a healthy trend, and reading it correctly requires understanding the physical mechanism behind the number, not just the number itself.

4. Bridging the Gap Between Technical Knowledge and Structured Decision-Making

The technical case for vibration analysis is well established in Indonesian industry, but most facilities stall at the step before the data collector ever gets used, when nobody has formally worked out which assets deserve monitoring, which failure modes justify the investment, and which components can be left on run-to-failure without argument.

An RCM and FMECA exercise on CNC milling machines at a manufacturing company found that once critical components were properly ranked and their maintenance intervals reset on reliability calculations instead of habit, the company saved Rp 51,800,331 in maintenance cost, a figure that reflects how much is normally spent maintaining the wrong things at the wrong frequency [5].

That gap between technical knowledge and structured decision-making is precisely where Cliste Rekayasa Indonesia works with reliability teams. We sit down with plant and reliability engineers to run the FMEA and criticality analysis, define P-F intervals per failure mode, and translate that into a vibration monitoring scope that matches consequence to cost instead of applying blanket coverage.

If your team is weighing whether a vibration program is worth building or whether an existing one is actually earning its keep, that is a conversation worth having before the next budget cycle closes.

Let’s Build a More Reliable Future.

Contact us now!

Author: Jen Megah Bremanda Sembiring (Reliability Engineer)

References

1. Analisis Predictive Maintenance pada Motor High Pressure Oil Pump dengan Monitoring Vibrasi di PLTU Barru, Jurnal Media Elektrik, Vol. 20 No. 2, April 2023.
2. Penjadwalan Perawatan Mesin dengan Metode Preventive Maintenance dan Predictive Maintenance, Studi Kasus di PLTD Kota Masohi, Jurnal Tekstil: Jurnal Keilmuan dan Aplikasi Bidang Tekstil dan Manajemen Industri, Vol. 7 No. 1, 2024.
3. Predictive Maintenance Berbasis Analisis Vibrasi pada Kompresor dan Vacuum Emulsifier Homogenizer PT Kosmetika Global Indonesia, Universitas Kristen Petra, 2023.
4. Hutahaean, R. Y., Identifikasi Kerusakkan Mesin Berdasarkan Sinyal Getaran dalam Domain Frekwensi, Studi Kasus di PT Freeport Indonesia, Jurnal Teknik Mesin Machine, Vol. 7 No. 2, Oktober 2021.
5. Penerapan Metode Reliability Centered Maintenance pada Mesin CNC Milling, Jurnal PASTI, Vol. XIII No. 3, Desember 2019.
6. PT Pertamina Geothermal Energy, implementasi digital twin untuk turbine monitoring, laporan peningkatan availability factor 2024 sampai 2026.