Vijaysai Prasad

Predictive Heat Exchanger Efficiency Monitoring

The performance of heat exchangers degrades with time due to fouling or deposition of material on the heat transfer surface. The fouling of critical exchangers in manufacturing plants results in a significant cost impact in terms of production losses, energy efficiency, and maintenance costs. While most plants monitor their exchangers to some degree, the ability to effect real and sustainable improvements requires four components: (1) real time monitoring; (2) an advance warning mechanism; (3) the ability to diagnose the cause of fouling; and (4) the ability to treat the cause in order to slow or reverse the degradation. CHeX is a comprehensive tool which monitors, predicts, and diagnoses heat exchanger performance. The unique features of this advanced technology include: numerous data cleaning steps to improve data quality and isolate a net fouling trend, an adaptive model which learns from the past to predict performance three years in advance, and knowledge-based diagnostics which identify the probable cause(s) of fouling and recommend corrective actions. The final control action is performed by a field engineer in adjusting the fouling treatment. The scope of the current paper includes only the detection and prediction features. To date, CHeX has been validated at three chemical processing plants, for fourteen exchangers. Selected case studies shall be presented to demonstrate the power of its algorithms over traditional calculations.© 2005 ASME