A complete procedure for leak detection and diagnosis in a complex heat exchanger using data-driven fuzzy models

Area of Science:

• Engineering
• Control Systems
• Artificial Intelligence

Background:

• Heat exchangers are critical components in many industrial processes.
• Effective leak detection is essential for operational efficiency and safety.
• Traditional methods may lack the precision for early-stage leak identification.

Purpose of the Study:

• To design and validate an efficient fuzzy model-based leak detection algorithm for a pilot heat exchanger.
• To develop a dynamic fuzzy model for real-time process monitoring.
• To assess the algorithm's performance in detecting leaks of varying magnitudes.

Main Methods:

• Derivation of a dynamic fuzzy model from input-output measurements using fuzzy clustering.
• Parallel execution of the fuzzy model with the physical process for symptom generation.
• Experimental validation on a real co-current heat exchanger.

Main Results:

• The fuzzy model-based algorithm demonstrated high efficiency in leak detection.
• The system successfully identified leaks of different magnitudes in the water circulation pipe.
• The approach proved robust for real-world heat exchanger applications.

Conclusions:

• Fuzzy model-based approaches offer an efficient solution for heat exchanger leak detection.
• The developed algorithm provides a reliable method for monitoring the integrity of water circulation pipes.
• This technique enhances operational safety and reduces potential downtime caused by leaks.