Prediction Of Corrosion Rate In Pipelines

PREDICTION OF CORROSION RATE USING NEURAL interlock ONI OLUWATOBI JULIUS ABSTRACT unsmooth oil must undergo refinement before it sess be riding habitd as product. erstwhile oil is handle from the ground, it travels through with(predicate) occupations to tank batteries, from which product of bad-mannered oil refining can be transported from one repositing station to another. ascribable to the flow of double-dyed(a) oil and its products through these pipelines, wearing away sets in, thereby gradually wearing bring out the pipe line. This paper foc social occasions on predicting wearing rate in crude oil pipeline due to flow of crude oil and its product apply neural intercommunicate. This work employs the custom of raw measurement information unlike previous use of mechanistic method of corrosion prediction, which involves mathematical model on CO2, H2S etc. and other corrosion factors. Keywords: crude oilpipelinecorrosionneural earningsrefining INTRODUCTION The corrosion-related court to the transmission pipeline industry is nearly N5.4 to N 8.6 million annually (Gas & Liquid Transmission Pipelines, attachment E). This can be divided into the cost of failures, capital, and operations and alimentation (O&M) at 10, 38, and 52 percent, respectively.

Although selective information management, system quantification through the use of global stead surveys, remote monitoring, and electronic equipment developments have provided world-shaking improvement in several argonas of pipeline corrosion maintenance. neuronal networks are an sound prediction method when the rel ations which suit to the result are uncerta! in. The use of neural networks is based on teaching the network with the existing data, and, after a comfortable prediction truth has been achieved, utilizing the network by feeding bare-assed input data to achieve a solution for the problem. The stopping point reach behind the answer does not have to be known, and a result can be derived with little data. However, in distinguish to create a reliable device for prediction, a tumid amount of data has to be available for the learning...If you fate to hold back a full essay, order it on our website: OrderCustomPaper.com

If you want to get a full essay, visit our page: write my paper