PREDICTION OF MINIMUM TRANSPORT CONDITION (MTC) OF SAND PARTICLES

ABSTRACT

The production of sand with hydrocarbons from reservoirs is unavoidable as the formation of the reservoirs is as a result of sedimentation in the earth. The deposition of sand or producing the sand at very high velocity poses serious detriment to the condition of the petroleum pipeline as it can lead to pipeline degradation and other hitches in the assurance of flow. This makes it important to investigate the minimum hydrocarbon velocity that prevents sand deposition in the pipelines and ensure it is maintained at different conditions. In this project, we employ the perceptron optimization approach to choose the best coefficient values for the model by reducing the mean square error between the actual and predicted velocity using a freshly built model based on considerable experimental data from a laboratory. The prediction was done with an artificial algorithm (PERCEPTRON) using the PYTHON software. Also, the Support Vector Machine (SVM) learning algorithm was used to model the same data used for the optimization process using python language to build the code. The input parameters were sand concentration, particle diameter, viscosity, density and superficial velocity of the hydrocarbons.