Fouling in refinery heat transfer units is a major problem that affects plant’s economics, operability, safety and environmental impact. Traditional heat exchanger design methodologies based on fixed values for the fouling resistance (e.g. TEMA fouling factors) have drawn several critiques in the past 40 years and were found responsible for exacerbating fouling rather than mitigating it. The fouling factors approach is, in fact, highly empirical and neglects fouling dynamics and its dependency on process conditions. The ability of capturing such dependency is therefore pivotal to overcome traditional design limitations. A novel dynamic, distributed model for a multi–pass shell–and–tube heat exchanger undergoing crude oil fouling was recently proposed by Coletti and Macchietto. The model takes into account the exchanger geometry and configuration, the variation of fluid temperature, velocity, physical properties and fouling rate along the length of each unit and captures the interactions between the fouling layer growth and the fluid–dynamics by solving a moving boundary problem. In this paper, the model is validated over a wide range of operating conditions (i.e. temperatures and flowrates) with data from four different industrial units (2 single and 2 double shells). Geometries and process conditions used are those of two refineries belonging to major oil companies (ExxonMobil and Shell). Some model parameters are estimated for each exchanger using measurements during the first 60 days after a mechanical cleaning. The model is then used in a fully predictive mode for subsequent times. Results indicate that for all units the outlet temperatures (in °C) are predicted over extended periods (i.e. 4–16 months) with an excellent accuracy of ±1% for the tube-side and ±2% for the shell-side. It is concluded that the model can be used with confidence on a wide range of operating conditions to calculate reliable temperatures and fouling resistances.

This content is only available via PDF.
You do not currently have access to this content.