Application of artificial neural network to predict permeability value of the reservoir rock

Abstract

Permeability is an important reservoir property but it is difficult to predict. An accurate measurement of permeability values can be obtained from core data analysis. However, this analysis is not possible to do at all interval wells in the field, so that permeability information becomes incomplete. Then, the use of artificial neural network method can be an alternative to predict the incomplete permeability values. This study used 191 of sandstone core samples from Upper Cibulakan Formation in the North West Java Basin. These core data were used to determine hydraulic flow unit (HFU) from the reservoir, and to obtain a relationship between porosity and permeability for each HFU. The application of artificial neural network method is done by building a database of flow zone indicator (FZI) based on its relationship with log data. From this FZI value, the HFU class can be known. Afterward, the permeability value can be obtained according to the equation of the relationship between porosity and permeability at each HFU that had been generated. Artificial neural network was applied on G-19 and G-11 Well that had 51 of core data. Based on this study, the result of permeability value is not much different from core data at the same depth, so that this method can be applied to obtain the permeability in uncored intervals.

Ultrahigh stable laminar graphene membranes for effective ionic and molecular nanofiltration with a machine learning-assisted study

“…37 The ANN method can also be used to effectively determine the permeability of a reservoir rock, capable of predicting the permeability values in the incomplete permeability wells. 36 To fulfil the ion permeability inside laminar membranes, ANNs have been selected as a computational tool for predicting the NaCl permeability through graphene membranes. This can allow to further understand the experimental limitation on the preparation of finer graphene thickness, less than 0.5 μm in particular.…”

“…Over the past few decades, ANNs have shown outstanding ability to predict permeability for various filtration systems. [36][37][38] ANNs learn the relationship between given inputs and outputs via a connected unit called "nodes", referring to "artificial neurons". Then, the nodes can be combined to form layers.…”

“…Artificial neural networks (ANNs) are one of the well-known supervised machine learning algorithms. 36,37,55 they have attracted substantial attention from various research areas due to their effectiveness and modesty in computational cost, especially the shallow neural network. 39 For example, ANNs have been previously used to evaluate the permeability decline in a permeable reactive barrier for a groundwater treatment method, at which the network consists of a single input layer, a hidden layer, and an output layer.…”