(220c) Novel Approach in Heat Exchanger Fouling Monitoring

<align="center">Novel Approach in Heat Exchanger Fouling Monitoring

H. Erdener Akinç¹, M. Becer²,  G. Sahin², S. Orman¹, S. Bas Hekimoglu¹, M. Bakir¹

¹TÜPRAS, Turkish Petroleum Refineries Corporation Head Office, Kocaeli/TURKEY

²TÜPRAS, Turkish Petroleum Refineries Corporation Izmir Refinery, Kocaeli/TURKEY

(Corresponding Author: metin.becer@tupras.com.tr)

Energy management is an important aspect for every industry especially for refineries. Due to everchanging refinery margins; the prevention of energy losses and maximum energy recovery are hot topics. Fouling in preheat exchangers is considered as the main sources of energy loss, reaching up to 2% of the refinerys’ total energy consumption. It is of great concern due to lowering of heat exchanger capacities and thus causing an increase in furnace heat loads, meaning extra fuel energy and greenhouse gas emissions; and as a consequence leads an increase in unit operation cost.

This paper aims developing a novel approach for heat exchanger fouling monitoring; which is based on heat exchanger simulation, real time process data monitoring and fouling prediction. In this respect a software was being developed and applied in both crude oil distillation unit preheat train exchangers and diesel hydro-processing unit feed/effluent heat exchangers in TÜPRAª refinery. In this software, the clean heat transfer conditions of the heat exchangers are simulated by considering the effects of physical and chemical properties such as temperature, pressure, composition, etc. of both tube and shell side flows of each exchanger considering the heat exchanger design specifications. Then the fouling factor, (or fouling resistance, RF) is being monitored from the difference between this clean heat transfer coefficient and the actual heat transfer coefficient obtained from real time process data. The simulation results were found to be consisted with real time operational data taken from the refinery. Going one step further, according to the fouling models -which is developed by relating the unit charge type (from storage tank or cracked, etc.) with the fouling factor models-the maintenance period of the units can be optimized. In this respect, the developed genetic and simulation algorithm run simultaneously in terms of economic aspects. The objective function of the optimization states a comparison between marginal process cost and maintenance cost due to unit shut down and extra fuel cost arising due to fouling; and the optimization constraints are defined according to the operational variables. The optimization algorithm was applied to the diesel hydroprocessing unit and the optimum cleaning time period of the heat exchangers were found.