Time-based maintenance is easy to plan but shows weaknesses when load profiles vary significantly. An engine running predominantly at part load ages differently from one under full load. Yet the PMS frequently prescribes identical intervals – for example, a cylinder inspection every 8,000 running hours regardless of actual wear.
The consequence: either maintenance happens too early, generating unnecessary costs and spare parts consumption, or damage occurs between intervals because the maintenance frequency is insufficient for the actual load profile. This is particularly problematic for vessels with frequently changing routes and loading conditions – bulk carriers on spot market charters are a typical example.
Another fundamental problem: time-based maintenance cannot capture damage triggered by extraordinary events – poor fuel quality, overload, contamination in the lube oil system. These events follow no calendar. They require condition-based detection.
Condition-based monitoring is especially valuable for rotating machinery and bearings. Vibration measurements on main bearings, turbochargers, generators and large pumps provide early indications of imbalance, misalignment and bearing wear – often weeks or months before an alarm is triggered.
Condition monitoring is equally strong in oil analysis. Particle counting, spectrometric metal analysis and water content measurement give reliable conclusions about internal component condition without disassembly. For cylinders of two-stroke engines, analysis of scavenging drain oil can provide insights into liner and piston ring condition.
Exhaust gas temperatures per cylinder are also a classic CM parameter. Systematic deviations between cylinders – corrected for load influence and ambient conditions – point to problems with injector nozzles, exhaust valves or piston rings. The key lies in normalisation: absolute values say little, but relative comparisons between cylinders and trend progressions over weeks are extremely reliable.
There are many areas where fixed intervals remain sensible or mandatory. Safety-critical systems such as life-saving appliances, fire protection systems and emergency generators are subject to clear regulatory requirements under SOLAS and class rules. Here, time-based maintenance is not optional but compulsory – regardless of measured condition.
For components with sudden failure patterns – such as seals, diaphragm valves or safety valves – condition monitoring often provides no reliable advance warning. These parts function until they fail, without gradual deterioration. Here, fixed replacement intervals based on experience and manufacturer specifications remain the better strategy.
Furthermore, many classification societies are not yet prepared to accept purely condition-based maintenance plans without a time-based fallback. DNV GL, Lloyd’s Register and Bureau Veritas offer programmes for condition-based class extensions (such as DNV’s “Machinery Planned Maintenance”), but require a documented risk assessment and defined minimum intervals as a safeguard.
Best practice is a mixed logic: time-based maintenance for mandatory and baseline tasks, condition monitoring for critical assets. In practice, this means a PMS structure with two tiers. The first tier contains all intervals prescribed by regulation and manufacturer – these are not altered. The second tier adds condition-based triggers that can initiate brought-forward or deferred maintenance.
An example: the cylinder inspection on the main engine is planned every 8,000 hours. Simultaneously, the condition monitoring system tracks exhaust gas temperatures and oil analysis. If the data shows deterioration at 6,000 hours, the inspection is brought forward. If it shows perfect condition at 8,000 hours, a documented extension of 1,000–2,000 hours can be considered – provided the manufacturer and class permit it.
IACS Rec. 74 offers a practical framework here. The recommendation describes how operators can structure a condition-based maintenance programme that is accepted by classification societies. Key elements are a systematic criticality assessment, defined measurement procedures, documented thresholds and a traceable escalation process.
The choice of measurement method depends on the asset type and available infrastructure. For main bearings and turbochargers, broadband vibration measurement (velocity, 10–1,000 Hz) has proven effective as an entry point. More advanced analyses use envelope analysis for detecting rolling bearing damage and FFT spectral analysis for identifying specific frequency patterns.
For lube oil analyses, a distinction is made between online systems that continuously measure water content, viscosity and particle count, and laboratory analyses that additionally provide spectrometric metal analysis, acid number and base number. For maritime use, a combination is recommended: online monitoring for rapid trend capture and quarterly laboratory samples for detailed condition assessment.
Thermography is another effective tool that is still used too rarely on vessels. Infrared inspections of switchboards, transformers and engine casings can detect localised overheating before it leads to failure. The cost of a portable thermal imaging camera has dropped considerably in recent years, and operation requires only basic training for the ship’s engineer.
The critical point with all methods: they must be applied under reproducible conditions. A vibration reading taken once at 50% load and once at 85% load is not comparable. Measurement SOPs (Standard Operating Procedures) with defined load conditions, measurement points and instrument settings are a fundamental prerequisite for any robust trend analysis.
Introducing a mixed maintenance logic changes procedures both on board and ashore. The Chief Engineer needs clear instructions on which measurements to take when and in which format to report results. The superintendent must have the competence to evaluate trends and make decisions – or bring in a technical adviser.
For spare parts management, the shift means less blanket stockholding, but more targeted procurement based on condition data. If oil analysis shows elevated iron concentration in cylinder oil, the corresponding spare parts (liners, piston rings) are ordered proactively – rather than discovering during the next scheduled inspection that they are needed.
The organisational effort is manageable if introduction is gradual. From experience, a superintendent requires approximately 2–4 additional hours per week for trend evaluation of 3–5 vessels. This effort pays for itself with the first prevented unplanned failure.
A tanker operator with six vessels ran separator maintenance on a purely time-based schedule: a bowl overhaul every 4,000 running hours. The cost per overhaul was approximately EUR 12,000 including spare parts and labour. On recommendation, vibration-based condition monitoring was introduced – a portable sensor measured monthly at the separator bearing.
Result after 18 months: on three vessels, intervals could be extended to 5,500–6,000 hours because condition data showed faultless operation. On one vessel, a brought-forward overhaul at 3,200 hours was necessary because vibration values early indicated a bearing problem. Net saving across the fleet: approximately EUR 40,000 per year, with investment costs below EUR 3,000 for the measuring equipment.
Condition monitoring is worthwhile when the asset has a high failure impact, offers measurable wear parameters, and the crew is capable of carrying out the measurements reliably. The decision should be based on a simple criticality matrix: failure probability multiplied by failure consequences yields the priority.
Not every asset justifies CM. For non-critical consumers such as workshop compressors or sanitary pumps, time-based maintenance is perfectly adequate. The effort for CM would here exceed the possible saving. Focus on the top 10–15% of assets by criticality delivers the best ROI.
Free initial consultation – we analyze your situation and find the best path forward.
Request Consulting