Predictive Maintenance for Pump Systems and Thermal Power Plants: State-of-the-Art Review, Trends and Challenges

Area of Science:

• Engineering
• Computer Science
• Data Science

Background:

• Thermal power plants are crucial for energy infrastructure, but component faults, especially in pumping systems, cause significant losses.
• Effective maintenance of pumps is vital for cost reduction and high availability in various industries.
• Prognostics and Health Management (PHM) and Predictive Maintenance (PdM) are gaining traction for data-driven system condition estimation.

Purpose of the Study:

• To conceptualize the field of Predictive Maintenance (PdM).
• To provide a structured literature overview of PdM applications, focusing on thermal power plants and pump systems.
• To identify current challenges and future trends in PdM.

Main Methods:

• Literature review and conceptualization of the PdM field.
• Structured overview of existing research and applications in PdM.
• Analysis of domain-specific challenges and trends in thermal power plants and pump systems.

Main Results:

• Numerous experimental data-driven models for PdM have been successfully deployed.
• A significant need exists for more industrial case studies in the PdM field.
• Scalability of PdM models and automatic model maintenance are key areas for future investigation.

Conclusions:

• The PdM field has seen successful implementation of data-driven models, particularly in experimental settings.
• Further research should focus on industrial case studies and the scalability of PdM solutions for large-scale deployment.
• Automatic maintenance of PdM models presents an interesting avenue for future research and development.