Permeability Prediction Using Hybrid Neural Network Modelling

Maslennikova, Yulia

doi:10.2118/167640-stu

Cited by 9 publications

(2 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…relative permeability and capillary pressure curves). The flow properties of the matrix are the primary controller parameters on the fluid exchange between the matrix and fractures (Babadagli and Al-Salmi 2004;Maslennikova 2013). The conventional permeability evaluation using Routine Core Analyses (RCA) and well logs were generally employed to populate the matrix block with the permeability.…”

Section: Methodology and Simulation Workflowmentioning

confidence: 99%

Modelling the transient effect in naturally fractured reservoirs

Aljuboori

Lee

Elraies

et al. 2022

J Petrol Explor Prod Technol

View full text Add to dashboard Cite

Modelling of Naturally Fractured Reservoirs (NFR’s) is a very challenging task. NFR’s are often simulated using the dual-porosity (DP) model, which requires significantly lower computational time and less simulation cost compared with the fine-grid modelling. However, the DP model is unable to capture the transient effect, the saturation front, and its tendency to overestimate the predicted oil recovery, as the invaded fluid is immediately reaching the gridblock centre once its saturation exceeds the critical saturation. The matrix block discretisation and modifying the transfer function are among the widely investigated areas to represent the transient effect and to improve the simulation accuracy. Adjusting the transfer function often results in complicated and unstable solutions, which make this approach limited in use. However, the matrix discretisation technique, such as Multiple Interacting Continua (MINC), is one of the utilised approaches to improve the simulation of NFR’s. The model’s layers provide the required mean to capture the transient effect and to include the matrix heterogeneities. In this work, we present an improvement to the original MINC model to enhance its accuracy and stability using a particular case of two subdomains model. We suggested using an equal volume of the matrix sub-blocks, besides performing an adjustment to the calculation of fluid saturation. The new adjustments have provided a stable solution and improved simulation results compared with the original MINC model. Besides the advantage of the matrix layers to handle various heterogeneities with excellent simulation accuracy and marginal errors, and hence a reliable performance and prediction.

show abstract

Section: Methodology and Simulation Workflowmentioning

confidence: 99%

Modelling the transient effect in naturally fractured reservoirs

Aljuboori

Lee

Elraies

et al. 2022

J Petrol Explor Prod Technol

View full text Add to dashboard Cite

show abstract

“…Using as inputs gamma ray, bulk density, and induction resistivity, the authors predict k with sufficiency against selected semi-empirical equations. Other researchers have worked in the same direction (Mohaghegh et al 1997;Helle et al, 2001;Habibian and Nabi-bidhendi, 2005;Singh, 2005;Maslennikova, 2013;Kohli and Arora, 2014;Al Khalifah et al, 2020) using practically the same inputs than Mohaghegh et al (1997) with slightly larger training sets. However, due to the nature of the training sets used in these works, the resulting models cannot be applied directly to heterogeneous conditions and their inputs/output ranges are still small-scale.…”

Section: Introductionmentioning

confidence: 99%

Neural model to estimate permeability from well logs and core data

García

Adame-Arana

2023

Bol. Geol.

View full text Add to dashboard Cite

A case study testing the effectiveness of neural networks for permeability determination in heterogeneous media using basic rock properties is presented. The dataset used consists of 213 core samples from the Morrow and Viola formations in Kansas, United States. The characterizing parameters of the cores are porosity (ϕ), water and oil saturations (Sw and So), and grain density (GD), and the additional variables from well logs are induction resistivity (ILD), gamma ray (GR) and neutron-porosity (NPHI). The neural predictions are compared with permeability values obtained from three semi-empirical models (Timur, Coates, and Pape) widely used in reservoir characterization. It is concluded that the neural network provides the best overall prediction quantified by the highest correlation coefficients (R and R2) far above those achieved with conventional methods in formations with rock heterogeneity and complex diagenetic nature. Applying Timur’s method R was 0.58 and R2 was 0.343, for Coates’ model R was 0.60 and R2 0.365 and for Pape’s model R was 0.60 and R2 was 0.372, while for the neural model, 0.97 and 0.94 were obtained for R and R2, respectively.

show abstract