OPTIMAL PREVENTIVE POLICIES FOR PARALLEL SYSTEMS USING MARKOV DECISION PROCESS: APPLICATION TO AN OFFSHORE POWER PLANT

This work proposes a Markov Decision Process (MDP) model for identifying windows of opportunities to perform preventive maintenance for multi-unit parallel systems subject to a varying demand. The main contribution lies in proposing: (i) a reward function that does not depend on maintenance costs, which are typically difficult to assess and classify; and (ii) a new metric for prevention. By optimizing the capacity utilization rate and the decision flexibility, which is denoted in terms of standby units, for a set of typical operational scenarios, the optimal opportunities for preventive interventions are identified within the respective prevention ranges, in relation to an offshore power plant (case study). The sequential decision problem is solved using the Value Iteration algorithm to obtain the optimal long-term policies. As a result, a backlog management decision-support solution is developed, using a low-cost computational model, which provides scenario-dependent preventive policies and promotes the integration of operations with maintenance, being easy to implement, maintain and communicate with stakeholders.

RELIABILITY AND RISK ANALYSIS OF AN ACTIVE HEAVE COMPENSATION DRAWWORKS

This paper presents a reliability analysis performed from a qualitative risk identification (HAZID), followed by analysis through failure modes, effects, and criticality analysis (FMECA) and fault tree analysis (FTA). The idea was to obtain a comparison between the risk levels of two solutions for rig motion compensation systems available on the market. In order to establish a comparison between the two technologies, a group of experts was consulted and, based on data and information collected inside and outside the company, the study was carried out encompassing the following steps: (i) Survey of scenarios where the heave compensation function is critical in offshore well construction and maintenance operations; (ii) failure modes, effects and criticality analysis (FMECA) in the raised scenarios; (iii) Identification of the winch system model with active compensation to serve as a reference; (iv) The design, simplification and resolution of fault tree analysis (FTA) for the scenarios/hypotheses of interest; (v) Data collection with the estimation of failure rates/probabilities and; (vi) Interpretation and discussion of the results obtained. The results of the work are presented and discussed, as well as the challenges and opportunities of the HAZID-FMECA-FTA approach used.

BIG DATA ANALYTICS FOR PREDICTIVE MAINTENANCE MODELING: CHALLENGES AND OPPORTUNITIES

This paper presents a comparison of different approaches for RUSBoost and Random Forest (RF) classifiers, in constructing such a prognostic system for a specific class of turbogenerator failures from a specific Floating Production Storage an Ofloading (FPSO) of Petrobras. Besides the comparison of different classifiers, a contribution of this work lies on the use of, not only telemetry data, which comes from the machine sensors, but also of non-structured data, such as official reports (e.g.: operator´s machine event annotations) regarding the most critical failures. Those reported annotations were correlated to telemetry data to identify real critical failures, and simultaneously avoid false positives.