Sensitivity to initial data errors in interpreting temperature logging of an isolated injection well segment

K. A. Potashev   ; D. R. Salimyanova   ; A. B. Mazo   ; A. A. Davletshin   ; A. V. Kosterin

Učënye zapiski Kazanskogo universiteta. Seriâ Fiziko-matematičeskie nauki, Uchenye Zapiski Kazanskogo Universiteta. Seriya Fiziko-Matematicheskie Nauki, Tome 166 (2024) no. 2, pp. 238-249

Résumé

This study considers the inverse problems inherent in interpreting temperature logging data from an isolated segment of the injection well in order to ascertain its operating period and the thermophysical properties of the oil reservoir. The forward problem of thermal conductivity was reduced to a one-dimensional axisymmetric formulation within the oil reservoir layer, disregarding the vertical thermal exchange with neighboring layers. The inverse problem of determining the well operating period was solved by reformulating the forward problem with regard to the temperature field derivative, which enabled the use of first-order optimization methods. Thus, Nesterov’s method was applied. An algorithm to automatically scale one of the method’s parameters (step length) was developed, and the optimal value of the second parameter (inertial step) was calculated. This increased the efficiency of the method by 10 – 15 % in solving the problem under consideration. The algorithm’s stability against perturbations in the initial data on temperature and thermophysical properties was demonstrated. The sensitivity analysis revealed that a 1 % error in the temperature measurements results in a standard deviation of the solution, which is about 2 % from the true value of the well operating period. A similar level of error was seen when the thermal diffusivity was over- or underestimated by approximately 15 %. The solution was little sensitive to variations in the heat transfer coefficient between the oil reservoir and the well \vskip0.2cm at characteristic magnitudes; even with a twofold distortion, the error in the determination of the well operating period did not exceed 1.5 %. To mitigate the error in thermometry interpretation to 1 %, temperature measurements must have an error margin of no more than 0.25 %, alongside precisely specified thermophysical properties of the oil reservoir, or, alternatively, when temperature is measured accurately, the rock thermal diffusivity must be set within an error margin of less than 3 %, but it is nearly impossible under real conditions. Increasing the number of temperature measurements diminishes the sensitivity to measurement errors, with the optimal efficacy achieved at 10 measurements, rendering further increments impractical. Therefore, the algorithm’s stability and the solution’s sensitivity of the inverse problem of determining the reservoir thermal diffusivity for a given operating period of the well relative to temperature measurement errors were found. The results show that a 1 % error in temperature measurements leads to a standard deviation of about 6 % from the true value.

Détail

Mots-clés : oil reservoir
Keywords: temperature logging, measurement error, geological uncertainty, inverse problem, numerical modeling.

Bibliographie

[1] Proselkov Yu.M., Heat Transfer in Wells, Nedra, M., 1975, 223 pp. (In Russian)

[2] Pudovkin M.A., Salamatin A.N., Chugunov V.A., Thermal Processes in Active Wells, Izd. Kazan. Univ., Kazan, 1977, 167 pp. (In Russian)

[3] Witterholt E.J., Tixier M.R., “Temperature logging in injection wells”, Proc. Fall Meet. of the Society of Petroleum Engineers of AIME (San Antonio, TX, Oct. 8–11, 1972), 1972, SPE-4022-MS | DOI

[4] Filippov A.I., Akhmetova O.V., Temperature Field in the Oil Reservoir and Well, Akad. Nauk Resp. Bashk., Gilem, Ufa, 2011, 336 pp. (In Russian)

[5] Tikhonov A.N., Arsenin V.Ya., Solutions of Ill-Posed Problems, Nauka, M., 1974, 288 pp. (In Russian)

[6] Özisik M.N., Orlande H.R.B., Inverse Heat Transfer: Fundamentals and Applications, Heat Transfer, 2nd ed., CRC Press, Boca Raton, FL, 2021, 297 pp. | DOI

[7] Baikov V.A., Bakirov N.K., Yakovlev A.A., Mathematical Geology, v. I, Introduction to geostatistics, IKI, Izhevsk, 2012, 228 pp. (In Russian)

[8] Mazo A.B., Fundamentals of Heat Transfer Theory and Calculation Methods, A Study Guide, Izd. Kazan. Univ., Kazan, 2013, 149 pp. (In Russian)

[9] Barenblatt G.I., Scaling, Self-Similarity, and Intermediate Asymptotics. Theory and Applications to Geophysical Fluid Dynamics, Gidrometeoizdat, L., 1982, 256 pp. (In Russian)

[10] Salimyanova D.R., Potashev K.A., Mazo A.B., Davletshin A.A., “Estimating the operating period of injection wells based on thermometry results and taking into account initial data errors”, Proc. XVI Sci.-Pract. Int. Conf. Mathematical Modeling and Computer Technologies in Field Development, Neft. Khoz., M., 2024, 38–39 (In Russian)

[11] Nesterov Yu.E., “A method of solving a convex programming problem with convergence rate O(1/$\text{k}^2$)”, Dokl. Akad. Nauk SSSR, 269:3 (1983), 543–547 (In Russian)

[12] Nemirovsky A.S., Problem Complexity and Method Efficiency in Optimization, Nauka, M., 1979, 384 pp. (In Russian)

[13] Kostyuk F.V., “Heavy ball method in unconstrained optimization problems for continuously differentiable functions with bounded Lebesgue sets”, Model., Dekompoz. Optim. Slozhnykh Din. Protsessov, 34:1(34) (2019), 151–159 (In Russian) | DOI

[14] Gill P., Murray W., Wright M., Practical Optimization, Mir, M., 1985, 510 pp. (In Russian)

[15] Yakovlev B.A., Solving the Problems of Petroleum Geology Using Thermometry Methods, Nedra, M., 1979, 138 pp. (In Russian)

[16] Hill A.D., Production Logging: Theoretical and Interpretive Elements, Monograph Ser., 14, 2nd ed., Soc. Pet. Eng., Richardson, TX, 2021, 256 pp. | DOI

[17] Lilliefors H.W., “On the Kolmogorov–Smirnov test for normality with mean and variance unknown”, J. Am. Stat. Assoc., 62:318 (1967), 399–402 | DOI

[18] Seredkin I.N., “Methods for measuring the thermophysical properties of rocks”, Gorn. Inf.- Anal. Byull., 2015, 249–255 (In Russian)