Optimisation of Vessel Maintenance Traditionally, vessel maintenance has always been carried out empirically, mainly based on the recommendations of the manufacturers and the requirements of the classification societies, but does this method optimise maintenance costs? In general, most drilling equipment include systems and parts that are subject to scheduled maintenance according to the prescriptions of the manufacturers. Such prescriptions, besides being usually on the conservative side do not take into account the systems mission profile and the actual environment in which they operate. Therefore, there is no evidence that they are economically optimal. To move from a prescriptive to an optimised and customised approach, it is necessary to review the scheduled maintenance intervals established by manufacturers on the basis of the actual reliability performance, safety and economic constraints. Such process falls within the framework of the more general approach known as Reliability Centred Maintenance (RCM). Computerised Maintenance Management System (CMMS) can thus be used to demonstrate the application of preventative maintenance concepts in order to achieve better quality, punctuality, speed and a higher service level for its fleet. Objectives are reached by planning maintenance repairs at the most convenient time, without interfering with the equipments availability, and thus reducing unexpected maintenance actions (i.e. breakdowns) while maintaining (or even improving) the original safety levels. Although not specifically aimed at reliability calculations, the presence of CMMS suggested the application of RCM techniques in order to add the economical optimisation of maintenance to the above objectives whilst improving the safety level. The ultimate goal of the Reliability Centred Maintenance method, based on reliability, availability and maintainability (i.e. RAM analysis), is the operating of the maintenance process to take into account both the risk of accidents, which can occur if primary systems fail, and the costs associated with their preventative and corrective maintenance. The optimisation process is the result of balancing risk and cost on the basis of well-defined RAM objectives of:

Safety Economic Commercial/quality

The RCM approach can thus be implemented by utilising risk analysis methods applied to safety criteria defined by regulations and the cost minimisation criteria as defined by the vessels owner. The basic methodology used in the implementation of the analysis included the following nine stages:

1. Preliminary risk analysis to identify the critical items selected. 2. Identification of the items for which maintenance is more expensive and the quality

and quantity of data 3. Familiarisation with the type of maintenance strategy currently applied to the selected

items (see table) 4. Collection of historical data of the scheduled and unscheduled maintenance actions

on the items under analysis 5. Confirmation of the data received from the drilling vessels owner to evaluate their

consistency with those utilised in the preliminary analysis 6. Statistical analysis of the items for which a significant number of failures has been

collected 7. Definition of the costs associated with failure and maintenance of each selected item,

in terms of down times, off-hire, spare parts and resources 8. Upgrade of maintenance strategy and, if possible, definition of optimal maintenance

intervals 9. Update of statistical analysis and of the maintenance intervals on a regular basis

Following the data analyses of the main systems of drilling vessels, it was concluded that the savings deriving from the customisation of the time intervals for the replacement of the selected items were outstanding. E.g. The frequency of replacement was abated by a factor of more than 2 (even 6 in one case). Five broad classes of maintenance strategies can be distinguished Maintenance Model Corrective

(unscheduled) maintenance

Preventive (scheduled) maintenance

Corrective maintenance only

replace at failure none

Age replacement policy (ARP) replace at failure replace after a time of operation

Block replacement policy (BRP) replace at failure replace after a fixed elapsed time

Minimal repair policy (MRP) minimal repair by failure replace after a fixed elapsed time

Periodic testing (for standby components)

none replace after a fixed elapsed time

Notes: For all strategies except MRP, the item is assumed to be restored as good as new after repair. The term minimal repair means that the item is repaired to the state it had immediately before the failure occurred. The parts of the systems selected in the study were subject to the ARP policy only. For these calculations, the definition of failure was crucial. Here, failure meant the actual breakdown of an item, and for every reliable component, this can determine the significance of the calculation. In such cases, the failure mechanisms and the actual purpose of the prescribed maintenance operations have to be thoroughly understood before applying the revised time intervals. It should also be stressed that the adoption of the optimal replacement time of an item does not necessarily minimise the number of failures. In the extreme case of equal preventive and corrective maintenance costs, it would be convenient to run the item to failure, thus tolerating the failures rather than carrying out the preventive maintenance actions. For this reason, this methodology is best applied to items where failure is expensive compared to preventive operations, but not immediately critical for the drilling vessels safety, so the risk of a failure can, in principle, be afforded. It is theoretically possible to apply this methodology to very critical equipment, on condition that the costs associated with its failure are correctly accounted for, which could be ve ry difficult when safety issues are involved. Another significant factor is the actual possibility of getting a statistic within a reasonable time frame; obviously, this can be achieved when the samples are sufficiently wide, thus a significant amount of data can be collected in a relatively short time. When the samples are small, the trend tests can be ignored and the analysis can proceed as illustrated. Care should be taken, however, when the economic risk associated with the items analysed is high. The uncertainties associated with optimal time intervals can be significant, and the decision to adopt the mean, upper or lower value must be taken by making use of the experience of the personnel in charge of maintenance activities.

A final consideration can be made for those items for which a statistical trend if detected and, consequently the quantification of the optimal time intervals. Even in this case, the statistical analysis can provide added value, in that it highlights the need for further investigation to understand the causes of the trend (in particular, if it represents decreasing reliability). The following checks should be made:

verify if the statistical results can be justified physically or are due to insufficient or poorly elaborated data;

if they are justified, understand the reason for the trend; e.g. parts replaced with inadequate spares, lack of skill of the operators performing the replacement, etc.

This process can highlight different maintenance performances among ships in the same fleet. A ship owner who has installed a CMMS onboard his ships obtained a real advantage by employing the maintenance data for statistical analysis. The cost of this is surely negligible with respect to the possible savings and the insights provided, on condition that qualified resources are involved in the process; expertise is particularly required in the phase of interpretation of the field data, which actually affects the final results.