Estimating the Porosity and Initial Water Saturation in South Structure of X Field Using Artificial Neural Network
Andrian Sutiadi (a*), Muhammad Taufiq Fathaddin (b*), Suryo Prakoso (b), Dwi Atty Mardiana (b)

a) PT. Prima Energi, Sahid Sudirman Centre, 53rd Floor, Jl. Jend. Sudirman Kav. 86, Jakarta Pusat 10220, Indonesia
*andrian.sutiadi[at]primaenergy.id
b) Department of Petroleum Engineering, Universitas Trisakti, Jakarta Barat, Indonesia
*muh.taufiq[at]trisakti.ac.id

Abstract

Determining the location of development wells requires rock and fluid data by carrying out petrophysical correlation between existing wells. That is done to estimate the potential of future wells or newly drilled wells. In this study, the porosity and saturation distribution of formations nearby CS-01 well with coordinates X = 722861.58 and Y = 9300235.29 at depths between 5377.5 ft to 6399.5 ft was estimated by applying an artificial neural network model (ANN). The ANN model was developed using data from three wells in X field. The data used includes measured depth, gamma ray, resistivity log, neutron log, density log as input parameters. Based on the results obtained correlation coefficients for training, validation, and testing processes for sequential porosity prediction are 0.9278, 0.9147, and 0.9303.
Meanwhile, correlations for training processes, validations, and tests for initial water saturation prediction in sequence are 0.8787, 0.9162, and 0.8220. The implementation of ANN model shows prediction of porosity and initial water saturation in average of 0.24 and 0.49 respectively.

Keywords: artificial neural network, flow efficiency, flowrate, skin factor, oil reservoir

Topic: Reservoir engineering