Converting data into knowledge with RCA methodology improved for inverters fault analysis

Area of Science:

• Renewable Energy Engineering
• Knowledge Management Systems
• Reliability Engineering

Background:

• Knowledge management in energy evaluation, particularly for maintenance teams and failure detection, has faced integration challenges.
• Existing expert systems for pattern recognition in energy equipment inefficiencies show limitations in incorporating maintenance team knowledge.
• Degradation of energy equipment in solar plants necessitates improved analysis methods for reliability.

Purpose of the Study:

• To develop an enhanced root cause analysis (RCA) methodology for photovoltaic (PV) inverters.
• To integrate knowledge management principles with failure mode analysis for improved energy equipment diagnostics.
• To address limitations in detecting sudden failures and early degradation in PV solar plant components.

Main Methods:

• Applied knowledge management methodology to energy evaluation and maintenance team training.
• Utilized Erdös-Rényi graphs from graph theory for quantitative and qualitative analysis.
• Conducted failure mode analysis on 164 inverters across 21 failure modes using extensive signal data.

Main Results:

• The study analyzed 120,561 signals and 3,014,025 patterns from 2018 to 2021.
• A novel RCA approach was developed, combining knowledge management with failure mode analysis.
• The method demonstrated effectiveness in complementing solutions for sudden failures and early degradation.

Conclusions:

• The proposed RCA method, enhanced by knowledge management, offers a significant improvement for PV solar plant maintenance.
• Integrating graph theory and knowledge management provides a robust framework for analyzing inverter failures.
• This approach aids in early detection and mitigation of energy equipment issues, boosting plant reliability.