DBPapers
DOI: 10.5593/SGEM2014/B11/S6.090

EASE THE TRANSPORT OF VISCOUS CRUDE OIL BY PIPELINE THROUGH PERFORMING RIGHT BLENDING RECIPES

T.Chis, A. E. Sterpu,C. I. Koncsag, A. I. Dumitru
Wednesday 1 October 2014 by Libadmin2014

References: 14th International Multidisciplinary Scientific GeoConference SGEM 2014, www.sgem.org, SGEM2014 Conference Proceedings, ISBN 978-619-7105-07-0 / ISSN 1314-2704, June 19-25, 2014, Book 1, Vol. 1, 695-702 pp

ABSTRACT
In Romanian Oil Field, heavy oil viscosity and pour point affect transportation by
pipeline. In many cases, the oil is transported by car and railway with some drawbacks (high costs, special condition at discharging, etc.). This is why, for the purpose of lowering costs, it is necessary to blend viscous crudes in recipes leading to physical properties prone to transportation by pipeline, The aim of this paper is to study the evolution of additive properties (density, sulfur content) and non-additive properties (viscosity, pour point, distillation curve) of oil mixtures with their mixing ratio. Mixtures of two crude oils or three crude oils were considered. A new algorithm based on discrete-time neural network was worked out for finding the oil blends properties evolution. The theoretical analysis of stability and convergence of the neural networks is also presented. An application example is provided to illustrate the neural network-modeling approach. We believe that modelling of crude oil blending via neural networks is a very effective method.

Keywords:oil heavy, blend, neural network