Full text loading...
-
MULTIPLE REGRESSIONS AND ANN TECHNIQUES TO PREDICT PERMEABILITY FROM PORE STRUCTURE FOR TERRIGENOUS RESERVOIRS, WEST-SHEBELYNSKA AREA
- Publisher: European Association of Geoscientists & Engineers
- Source: Conference Proceedings, Monitoring 2019, Nov 2019, Volume 2019, p.1 - 5
Abstract
In the processes of exploration, allocation of producing intervals, and development of hydrocarbon deposits, the important part is the accurate determination of the poro-perm properties. According to the complex of laboratory studies of the reservoir rocks from the West-Shebelynska area (depth interval 4929–5380 m), the authors predicted the permeability coefficient using the quantitative distribution of different types of voids in (the reservoirs of different types (intergranular, secondary, and fractured). With the help of Multiple Linear Regression (MLR), it was established that the filtration of fluid in the investigated complex reservoir rock samples occurs both in intergranular and secondary voids as well as in fractures. It is shown that Artificial Neural Network (ANN) and Multiple Nonlinear Regression (MNLR) algorithms can provide a stable model with a high degree of confidence that can be used to predict the permeability coefficient at the intervals studied.