A copula-based statistical method of estimating production profile from temperature measurement for perforated horizontal wells

Estimating production profile from temperature data is of great significance for production increase and cost reduction in well production process. This study aims to provide a cost-effective direct method of interpreting flow profiles from temperature measurements. The concept of representative inspired by randomized distribution regarding heat conduction representative points for each no-influx section is proposed. Three assumptions regarding existence, distribution and independence of representatives are introduced to establish a copula-based statistical method of converting fitted distributions of representative, to distributions of relaxation distances for each no-influx section, and then to estimate the production profile. The method is applied on a horizontal gas well on the east slope of western Sichuan depression, and the simulated results are compared with the production logging data. The error of estimators for each no-influx section is between 6.7% and 11.8%, which is satisfactory. Sensitivity analyses are conducted regarding temperature measurement error, and the results verify the robustness of this method. This paper provides a reliable solution to interpret production profile from relatively inaccurate temperature data when the perforation zones are adjacently close. The findings of this study can help for better understanding of low-cost production allocation and uncertain flow state modeling problems.

Automated summary

This study proposes a cost-effective direct method of interpreting flow profiles from temperature measurements, by establishing a copula-based statistical method. The method is applied on a horizontal gas well and compared with production logging data, showing satisfactory error rate between 6.7% and 11.8%. Sensitivity analyses verify the robustness of this method, providing a reliable solution to interpret production profiles from relatively inaccurate temperature data.