Technical condition indexes for auxiliary marine diesel engines

Abstract

Knowledge about the technical condition of a ship is an important issue. It will help in improving the planning of the maintenance actions, and giving earlier alerts to the organization if there is any development that might affects the ship’s integrity. MARINTEK and NTNU have developed a methodology able to quantify and calculate the technical condition of any technical system through a Technical Condition Index (TCI). The TCI enable the user at early stages to reveal the development of undesired potential failure to the system in question and consequences for the safety and the environment.

The main objective of this PhD work is to develop a methodology for performance and health monitoring for ship’s auxiliary engine (AE), such that it can be implemented in the maintenance management systems in a life cycle perspective. TOCC1 has proposed a performance monitoring model for this type of engines. This model is utilizing the TCI concept and has been originated from a similar model; however for the ship’s main engine (ME). The main difference between the AE and ME is the turbocharging concept; constant pressure turbocharging concept in ME, while pulse turbocharging concept in AE. Not all the sensors mounted on the ME’s turbocharger are mounted too on the AE’s one. This is a challenge, and so a new model has been proposed by TOCC for the AE’s turbocharger. This proposed model has been examined in this thesis.

To utilize the AE TCI model, most of its input parameters need to be corrected first to ambient conditions. Moreover, knowledge about the performance of the AE under faulty conditions is another step to test the TOCC’s proposed model. Therefore, a simulation model has been built, using a software package called GT-Power. This model has been calibrated, and then validated by executing sensitivity and certainty analysis. Several simulations have been done using the validated model to understand the performance of the AE under changing ambient conditions and under selected faults. From the simulation results, new ambient correction factors have been developed for most of the parameters input to the AE TCI model. Moreover, the currently used ambient correction factors have been examined and a conclusion about using them has been stated.

Two groups of faults have been simulated in this work. The first group contains faults that influence the amount of air delivered to the engine, while the second group contains faults that influence the amount and timing of fuel injected to the engine. The results have been explained and analyzed, and then it has been concluded that the response of the proposed AE TCI is quite good. Moreover, it has been concluded that the TCI concept, beside being used for performance monitoring, it can be used also for fault detection. A fault matrix has been developed such that it can be a basis for a complete fault detection and diagnosis system.