(112b) A Novel Way of Monitoring Heat Exchanger Performance:  the Dynamic TH-? Plot

In many industrial situations, from milk processing to refining, heat exchanger performance changes over time due to deposition of material on exchange surfaces (fouling) leading to losses in duty (reduced thermal efficiency) and increased pressure drop (reduced hydraulic efficiency). Operating performance is usually measured by monitoring just thermal efficiency, typically using thermal efficiency measures such as fouling resistance, R_f, that are easy to calculate. The reasons of thermal inefficiency, however, can be quite varied, ranging from different deposition rates, deposition of materials with different thermal conductivity, coking of existing deposits, to removal of deposits by shear, often having opposite effects. On its own, simply monitoring fouling resistance can give quite misleading indications of performance, and of the underlying conditions of the exchanger.

The dynamic TH-λ plot is an easy to understand, rich visualisation that combines information about both the thermal and hydraulic evolution of the performance of a heat exchanger. In its simplest form, it evidences the changes in duty and pressure drop relative to clean performance and permits tracking and assessing the approach to thermal (T) and hydraulic (H) limits. The addition of characteristic reference lines for deposits of constant thermal conductivity, the so called λ-lines, allows monitoring and in many cases helps identify the degree of deposit coking, abnormal events such as acute fouling and inorganics breakthrough. When process conditions (flowrates and temperatures of the streams) are variable, as is common in practice, the reference performance (clean and l-lines) are obtained using a sophisticated dynamic model of the exchanger under fouling.

Using a variety of industrial examples, the presentation will illustrate the ability of such a representation and underlying models to resolve the ambiguities of simpler thermal resistance metrics, provide an improved understanding of the conditions of an exchanger, and doing so much faster that would otherwise become apparent. This is achieved by extracting a lot more information from the measured plant data. The role of such new tool for operations monitoring and diagnosis of abnormal events will be discussed.