Prediction of the behaviour of the Silt Density Index (SDI) in the Caribbean Seawater and its impact on RO desalination plants

The environmental characteristics of the seawater and its impacts on planning and operating RO desalination plants were always the concern of several scientists around the world. In this article, a study has been performed to investigate the seasonal characteristic of seawater in the Caribbean. The Radial Basis Function Networks (RBFN) was the methodology used in this paper where the redistribution of centres and the input training data possess significant effects in the model. The proposed method is based on clustering of input space vectors and computing weights of Euclidian distances and histogram equalization within each cluster for determining the centre and width of each receptive field. The model allows predicting the behaviour of the SDI using field data (turbidity and salinity) taking into account the tides (the Gulf of Paria) and seasonal changes (hurricane or dry seasons). The results indicated that during the hurricane season, the parameters can change up to 32% between seasons along year due to the change of direction of the inshore currents. Also values higher than 6 in the SDI indicated a very high potential for fouling. This article offers to the project engineers a new door in the application of precautionary principles in the design of desalination plants.