Paper Information

Title: 

ESTIMATION OF POROSITY IN GACHSARAN OIL FIELD BY USING ARTIFICIAL NEURAL NETWORKS AND MULTIPLE LINEAR REGRESSION

Type: PAPER
Author(s): JALALI LICHAEI M.,NABI BID HENDI M.*,MIRZAEI S.
 
 *INSTITUTE OF GEOPHYSIC, TEHRAN UNIVERSITY
 
Name of Seminar: PROCEEDING OF IRANIAN MINING ENGINEERING CONFERENCE
Type of Seminar:  CONFERENCE
Sponsor:  IRANIAN SOCIETY OF MINING ENGINEERING
Date:  2005Volume 1
 
 
Abstract: 

Porosity is an important parameter for formation evaluation and is a key parameter in Petroleum Engineering. The core measurements and well logs are usually used for porosity determination. The first method is very expensive and the second one does not give reliable results. In addition, statistical methods will have some problems due to the input data obtained from well logs. Artificial neural network is a new method, recently has been used in oil industry for prediction of petrophysical properties. In this paper, artificial neural network and Multiple Linear Regression (MLR) have been used for porosity prediction. For this reason, the data of four wells (well Nos. 25, 31, 32, 287) from Gachsaran oil field have been used. The data consist of petrophysical logs and their core measurements. For porosity prediction, the data of wells 25, 32 and 287 have been divided to three sets for Training, Testing and Validation Processes. The results of these processes have been compared with the porosity obtained from core tests. Finally the data of well 31 is used for Generalization and a correlation coefficient of 0.94 is obtained.

 
Keyword(s): ARTIFICIAL NEURAL NETWORK, POROSITY, BACK PROPAGATION NETWORK, MULTIPLE LINEAR REGRESSION
 
Yearly Visit 3   tarjomyar
 
Latest on Blog
Enter SID Blog