Neural network boosted with differential evolution for lithology identification based on well logs information
- PDF / 1,429,643 Bytes
- 8 Pages / 595.224 x 790.955 pts Page_size
- 72 Downloads / 180 Views
RESEARCH ARTICLE
Neural network boosted with differential evolution for lithology identification based on well logs information Camila Martins Saporetti1 · Leonardo Goliatt2
· Egberto Pereira3
Received: 26 March 2020 / Accepted: 25 September 2020 © Springer-Verlag GmbH Germany, part of Springer Nature 2020
Abstract Lithology identification of geological beds in the subsurface is fundamental in reservoir characterization. Recently, automated log analysis has an increasing demand in reservoir research and the oil industry. In this context, Machine Learning (ML) techniques arise as a surrogate model to provide lithology identification in a fast way. However, to achieve suitable performance, ML techniques require the adjustment of some parameters, and that can become a hard task, depending on the difficulty of the problem to be solved. This paper presents an Artificial Neural Network (ANN), assisted by an adaptive Differential Evolution (DE) algorithm to classify petrophysical data in the Southern Provence Basin. The main contribution is searching for a competent ANN configuration, including architecture, activation functions, regularization, and training algorithms. The proposed approach outperformed four classifiers and two results previously published. The computational methodology proposed here is able to assist in the classification of petrophysical data, helping to improve the procedure of reservoir characterization and the idealization of the development of production. Keywords Lithology identification · Neural networks · Evolutionary algorithms · Automated machine learning
Introduction Lithology identification of geological beds in the subsurface is fundamental in reservoir characterization, as one cannot predict the fluid content of any geological bed without knowing the lithology that the fluid is associated with. To make petrophysical properties estimations, such as porosity, clay volume, water saturation, permeability, the various lithologies of the reservoir interval must be identified,
Communicated by: H. Babaie Leonardo Goliatt
[email protected] Camila Martins Saporetti [email protected] Egberto Pereira [email protected] 1
UEMG, Divinopolis, Brazil
2
UFJF, Juiz de Fora, Brazil
3
UERJ, Rio de Janeiro, Brazil
and their properties understood. Accurate determination and understanding of lithology are fundamental to other petrophysical analyses, critical for practical exploration, and production of hydrocarbon. The economic potential of an oil reservoir depends on the quality and description of lithology (Abbey et al. 2018). The determination of lithology can be carried out by direct or indirect methods. Direct examination of underground cores sample from the intervals of interest is costly, partially reliable and biased different geologists can provide dissimilar analysis. Indirect methods use well log data to measure the physical aspects of geological formations, providing most of the data accessible to a geologist. Considering their influence on decisions, they are also fund
Data Loading...
