The tendency to fouling is a common problem that affects an oil refinery's processing units. Residues naturally present on the oil or chemical products added during its transport deposit and foul tubes and internal surfaces of equipments. A fouled heat exchanger loses its capacity to adequately heat the oil, needing to be shut down periodically for cleaning. Previous knowledge of the best period to shut down the exchanger may improve the energetic and production efficiency of the plant. In this work we develop a system to predict the fouling on a heat exchanger from a real refinery, based on data collected in a partnership with Petrobras. Recurrent Neural Networks are used to predict the heat exchanger's flow in future time. This variable is the main indicator of fouling, because its value decreases gradually as the deposits in the tubes reduce their diameter. The prediction could be used to help plan the maintenances, providing previous information about the fouling evolution.