Collecting And Analyzing RAM Data as Part of Maintenance Management in FPSO Operations

Abstract

Different decision support tools are introduced to help the decision makers and managers in different industries to make reliable decisions. Teekay Petrojarl Production as the largest FPSO operator in North Sea is moving toward operational excellence and implementation of such tools in decision making is known as an accepted approach. Such decision support tools need input data to be able to produce the results. For this, a RAM tool was developed by SINTEF/NTNU and is implemented in the company for supporting decisions.

Hence, in this master thesis the methods and techniques for providing such data are discussed. This includes studying the required parameters for different decision situations in addition to explore the required structure of a database for recording thedata. Being familiar with required parameters and database structure, in the next step, different analysis approaches are introduced and a case study using current data ismade. Specifically, use of Bayesian technique in practice is presented.

Generic data and plant specific data are the main data sources in hands of each reliability analyst and the estimations are produced out of these data.

Hence, this study, is supposed to provide a theoretical backbone regarding the estimation techniques and challenges in collecting and analysis of data in addition to implementing these theory to real data from TKPJ.

This work is a complementary to the specialization project which was done in Autumn 2010 which was supposed to provide necessary theoretical background for RAM data collection and Analysis in TKPJ by approaching the topic from a structural perspective.