A large amount of energy in oil refineries is used to heat the crude oil in furnaces, following a pre-heat train (PHT), a network of heat exchangers which maximise energy recovery from the hot products. The crude oil flowing in the exchangers deposits unwanted layers of material (fouling) which impede the energy recovery and can obstruct the flow of oil in the pipes. This results in severely decreased thermal and hydraulic efficiencies, large production losses (due to throughput reduction and need for periodic cleaning shut-downs). Energy recovery losses in the PHT must be compensated by substantial increase in fuel burned in the furnace, with large CO2 emission. Avoiding or minimising fouling is therefore important.

Fouling deposition is not well understood, and current exchanger designs (based on empirical fouling factors) and mitigation solutions (ranging from use of chemical additives to mechanical cleaning devices) do not prevent efficiency losses. A US$ 5.4M project (CROF, for CRude Oil Fouling), involving a consortium of universities supported EPSRC and industry, is presently underway, aiming at producing a step change in understanding, data, predictive models and methods for the improved design and operation of PHT.

In this paper, a detailed dynamic, distributed mathematical model for a single tubular heat exchanger with fouling is proposed. By including local effects of process variables (temperature and velocity) and deposition mechanisms, the model accounts for interactions between fouling layer and fluid flow and reproduces well overall experimental behaviour.

Various retrofit options for individual heat exchangers are assessed and their energy and CO2 release impact established. For the future, it is envisaged to use the approach to simulate and optimise the design and operation of the overall energy recovery network, its cleaning schedules and energy efficiency strategies. Results so far indicate the return on such work can be substantial.