A.V. Pinchuk, E.A. Pylev, E.E. Polyakov, M.A. Tvorogov, I.V. Churikovа
Optimisation of cluster drilling based on integrated seismic attributes and well log data analysis using neural network algorithms: Chayandinsky oil and gas condensate field
Key words: Chayandinsky oil and gas condensate field; geological model; Khamakinsky and Talakhsky horizons; well; prediction of lithotype occurrence; lithology; reservoir; neural network; classification; attribute analysis; seismic attribute; lithology cube.
For citation: Pinchuk A.V., Pylev E.A., Polyakov E.E., Tvorogov M.A., Churikova I.V. Optimisation of cluster drilling based on integrated seismic attributes and well log data analysis using neural network algorithms: Chayandinsky oil and gas condensate field. Geologiya nefti i gaza. 2022;(2):17–30. DOI: 10.31087/0016-7894-2022-2-17-30. In Russ.
Chayandinsky oil, gas and condensate field is one of largest in Russia. The main gas accumulations are found in the Vendian Botuobinsky, Khamakinsky, and Talakhsky pay horizons. The field is confined to the large non-structural trap in the north-eastern part of the Nepsky Arch; the field has a rather complicated geological structure that causes numerous challenges in its development. With the purpose to optimise cluster drilling and improve the efficiency of the Chayandinsky oil, gas and condensate field development, prediction of reservoir occurrence was carried out, and their lithological membership within the yet undrilled development well clusters was updated with adjustments based on the wells drilled. The authors discuss the methodology for integration of lithotypes identified from well log data with seismic data, which is based on application of an innovative neural network algorithm. They present the new method of building the predicted local geological models, which is created by them, including the following: re-interpretation of well log data from development wells; integrated interpretation of seismic and drilling data using the method of trainable neural networks; creating adaptive geological cluster models of pay horizons in the Chayandinsky oil, gas and condensate field. The results were lithology cubes accounting for wells data and probability cubes for identified lithological varieties. Comparison of the obtained lithology cubes with geological modelling results being a part of reserves assessment is presented. The authors note a more differentiated distribution of lithological varieties across the section of pay horizons and, as a consequence, more differentiated maps of net thicknesses. The use of the proposed tool will make it possible to update the distribution of zones with better reservoir properties with the purpose of optimizing the placement of production well clusters and increasing the development efficiency of Vendian terrigenous deposits in the Chayandinsky oil, gas and condensate field.