Session: 02-11-01 Reliability Based Maintenance

Submission Number: 182227

Comparison of Preventive Maintenance Intervals for FSPO Items Based on Zero-Failure Reliability Models and Uwild Program 

In Floating Production Storage and Offloading (FPSO) units, several items such as sea chests are cleaned periodically as part of its preventive maintenance plan to prevent marine growth and clogging. This periodicity is typically defined by the Underwater Inspection In Lieu of Drydocking (UWILD) program, which establishes fixed calendar intervals, often once or twice every five years. However, as these components rarely exhibit failures throughout their operational life, the use of time-based intervals may lead to non-optimal maintenance scheduling and unnecessary interventions. In this context, this study compares the preventive intervals defined by the UWILD program with those obtained from different probabilistic reliability models developed for zero-failure and censored data. These models include approaches such as based on Bayesian posterior estimation using non-informative priors, modified maximum likelihood estimation (MMLE) for Weibull-distributed lifetimes, bootstrap-based confidence bounds, and matching distribution curve methods. These models allow the estimation of the time-to-failure distribution even in the absence of observed failures, by incorporating censored information from suspended data and prior knowledge on the expected service life of the items. By applying these models to typical FPSO items with high design reliability and long operating lifetimes, this work aims to evaluate how the estimated mean time to failure (MTTF) and reliability functions influence maintenance planning. The comparison highlights the potential deviation between time-based preventive intervals and those predicted by probabilistic zero-failure models. In addition, the results may support optimization using the Cost per Unit Time (CPUT) criterion to define economically justified maintenance intervention intervals. It is expected that the results will demonstrate more optimized intervals between preventive interventions, aligning maintenance periodicity with the statistical reliability of components rather than fixed schedules. Such findings can contribute to improving resource allocation and maintenance planning within the context of maintenance of FPSO underwater systems.

Presenting Author: Edilson Gabriel Veruz Escola Politécnica - Universidade de São Paulo

Presenting Author Biography: Edilson received the degree of Mechanical Engineer from Universidade Tecnológica Federal do Paraná. He received the Master of Engineering Degree from Escola Politécnica - University of São Paula. Currently he is a Doctor of Engineering student at Escola Politécnica - University of São Paulo. His research interests are: risk-based analysis of equipment and structures and risk-based maintenance planning.

Authors:

Edilson Gabriel Veruz Escola Politécnica - Universidade de São Paulo
Alecio Julio Silva Escola Politécnica - University of São Paulo
Renan Favarão Da Silva Escola Politécnica - University of São Paulo
Miguel Angelo De Carvalho Michalski Escola Politécnica - University of São Paulo
Gilberto Francisco Martha De Souza Escola Politécnica - University of São Paulo
Fernando Andrade Rodrigues Leopoldo Américo Miguez de Mello Research and Development Center - Petrobras