Data-Driven Approaches for Operation and Maintenance Decision on Power Plants

The paper proposes a decision-making system based on Proactive Operation & Maintenance (POM) data-driven approaches, supported by Condition-Based Maintenance (CBM) and Statistical Data Analysis (SDA). The CBM approach is comprised of both online and offline data acquisition methods. To develop the system, the online CBM data acquisition involves the analysis of real-time sensor data at the asset of power plants that are categorized based on their numerical variables and then mapped on the asset register. On the other hand, offline CBM data acquisition is conducted by performing in-situ measurements at the power plants site. Statistical data analyses of failure data derived from the Computerized Maintenance Management System (CMMS) are also utilised to support the decision system. The POM approach supported by data-driven method and set point values of the controllable parameter is developed. In this decision-making system, the POM approach through the combination of CBM and SDA is enhanced to be structured by proactive online recommendations and comprehensive analysis space followed by expert judgments. This study presents a novel strategy for the creation and implementation of a proactive decision-making system in the context of operation and maintenance of power plants. It presents a decision-making paradigm that encompasses CBM and SDA, which can raise the amount of confidence in expert judgment, as well as the types of recommendations that can be made. The impacts show that the performance of the power plants increased.

Keywords: Condition-based maintenance, Data-driven, Decision-making system, Operation & maintenance, Power plant, Reliability.

Download PDF