In the process of constructing forecast models, based on which the prediction of the future dynamics of oil production is built, it is highly crucial to quantify the mutual influence of wells within the target area. The complexity of such estimates is associated with among others the need for consideration time-lagging response of injection wells and the production wells of one group (cluster). The main purpose of this research is to develop a methodology for estimation of mutual influence of injection and production wells in a quantitative manner that will take into account the information regarding time-lagging response within a target area. The empirical basis for constructing models was a field that is presented by 115 wells (24 groups of wells), where 84 of them are producing ones. The data was considered in daily dynamics for a specific period of time - 9 years and 6 months and it was presented by various parameters such as fluid flow rate for production wells, reservoir and bottom-hole pressures, volume of injected fluid for injection wells. A vector autoregression model was considered as a quantitative representation the parameters of which were found using an approach of Bayesian estimation. In all constructed models for 24 groups of wells, there was a statistical significance (p <0.05) dependences of the fluid flow rate of production wells on the following parameters: 1) fluid volume of the adjacent production wells in the group of wells with a time lag of up to 4 days, 2) difference between the bottom-hole and reservoir pressures at the current time; 3) pressure difference in adjacent wells along with the cluster/group (for some wells); 4) cumulative injection rate for all injection wells in the cluster/group. The average forecast error found on the test dataset of 30-day dynamics for 76 out of 84 producing wells and was 3.84%, while for another 4 wells this parameter was 14-18%, and for the other 4 wells it was 36-51%. All model estimates were considered to be as robust and asymptotically consistent. Application of the developed methodology for constructing a vector autoregression model with parameter estimation based on the Bayesian approach made it possible to consider the mutual influence of the production rate of wells on each other, taking into account the delayed effect. Moreover, the obtained estimates of this impact were reliable since it was indirectly proved by the small value of the forecast errors calculated using the corresponding models.
1. Grif A.M., Persova M.G., Soloveichik Yu.G., Determination of the effect of injection wells on production wells in their work dynamics by using hydrodynamic modeling science, Bulletin of the Novosibirsk State Technical University, 2019, no. 4(77), pp. 31–44, DOI: 10.17212/1814-1196-2019-4-31-44.
2. Weber D., Edgar T.F., Lake L.W. et al., Improvements in capacitance-resistive modeling and optimization of large scale reservoirs, SPE-121299-MS, 2009, DOI:10.2118/121299-ms
3. Kaviani D., Interwell connectivity evaluation from wellrate fluctuations: a waterflooding management tool: Doctoral dissertation, Amirkabir University of Technology (Tehran Polytechnic), 2009.
4. Khatmullin I.F., Tsanda A.P., Andrianova A.M. et al., Semi-analytical models for calculating well interference: limitations and applications (In Russ.), Neftyanoe khozyaystvo = Oil Industry, 2018, no. 12, pp. 38–41, DOI: 10.24887/0028-2448-2018-12-38-41.
5. Khasanov M.M., Bakhitov R.R., Lakman I.A., Review of research on modeling the geological structure and processes of field development (In Russ.), Neftyanoe khozyaystvo = Oil Industry, 2021, no. 10, pp. 46–51, DOI: 10.24887/0028-2448-2021-10-46-51
6. Sagitov D.K., Studying of changes in the interaction of well during the water injection process (In Russ.), Izvestiya vuzov. Neft' i gaz, 2019, no. 2, pp. 81–85, DOI: 10.31660/0445-0108-2019-2-81-85
7. Ponomareva I.N., Martyushev D.A., Chernyy K.A., Research of interaction between expressive and producing wells based on construction of multilevel models (In Russ.), Izvestiya Tomskogo politekhnicheskogo universiteta. Inzhiniring georesursov, 2021, V. 332, no. 2, pp. 116–126, DOI: 10.18799/24131830/2021/2/3048
8. Suhartono D., Prastyo H., Kuswanto M., Hisyam L., Comparison between VAR, GSTAR, FFNN-VAR and FFNN-GSTAR models for forecasting oil production, MATEMATIKA, 2018, V. 34, no. 1, pp. 103–111, DOI:10.11113/matematika.v34.n1.1040.