Developing a statistical model of oil and gas potential prediction by gas showings in the Verkhnekamskoye deposit strata of potassium and magnesium salts

• Main Authors: Vladislav I. Galkin, Oleg A. Melkishev, Stanislav V. Varushkin, Segey S. Andreiko, Tamara A. Lialina
• Format: Article
• Language: English
• Published: Perm National Research Polytechnic University 2020-03-01
• Series: Perm Journal of Petroleum and Mining Engineering
• Subjects: oil and gas potential prediction; gas content of salts; gas-dynamic effects; probabilistic and statistical models; probability; verkhnekamskoye deposit of potassium and magnesium salts; perm krai
• Online Access: http://vestnik.pstu.ru/get/_res/fs/file2.pdf/9141/Developing+a+statistical+model+of+oil+and+gas+potential+prediction+by+gas+showings+in+the+Verkhnekamskoye+deposit+strata+of+potassium+and+magnesium+saltsfile2.pdf

The Verkhnekamskoye deposit of potassium and magnesium salts is located in Solikamsk depression within Pre-Urals foreland basin. There is a salt deposit in the near-surface section, with a number of oil and gas deposits below it. The gas-oil ratio is known to be central to the process of gas showings during drilling of geological exploration wells and initiation of gas-dynamic phenomena at underground operations. For this purpose, all evidence on gas showings recorded when drilling salt exploration wells at the Verkhnekamskoye potash deposit was collected, as reflected in archival data and reports on prospect evaluation surveys. The findings were summarized and used to build a probabilistic and statistical model of oil and gas potential prediction. This study considers 18 characteristics of 374 wells, related to the thickness of salt productive formations and their quantity. The characteristics were compared using the Student's t-test and Pearson's chi-squared test (χ2). This dependence was used to calculate probability values Pp (Zm) for all 856 studied salt exploration wells. At the first stage, individual one-dimensional probabilistic models of gas potential prediction were built. The obtained individual probabilities served the basis for obtaining a discriminant function (Zm) to predict the gas potential in the salt massive. The obtained values of Zm discriminant function were used to build a regression model of oil and gas potential prediction Pp (Zm). Following this dependency, Pp (Zm) values were calculated for all 856 studied wells drilled for exploration and prospecting operations for the purpose of salt extraction. The average value (± standard deviation) of the probability for the class within the outline of the oil and gas zone was 0.510 ± 0.068 unit fractions. For the class outside the oil and gas zone, the average value was 0.490 ± 0.070 unit fractions. The obtained models enable constructing gas-show prediction schemes and an oil and gas potential prediction scheme for the Verkhnekamskoye potash deposit.