Interpretation Method of Donwhole Temperature and Pressure Data for Detecting Water Entry in Horizontal/Highly Inclined Gas Wells

Zhu, Ding (Texas A&M University) | Achinivu, Ochi Ikoku (Texas A&M University) | Furui, Kenji (ConocoPhillips Co)

OnePetro 

Accurate and reliable downhole data acquisition has been made possible by advanced permanent monitoring systems such as downhole pressure and temperature gauges and fiber optic sensors. These downhole measurement instruments are increasingly incorporated as part of the intelligent completion in complex (highly slanted, horizontal, and multilateral) wells where they provide bottomhole temperature, pressure and sometimes volumetric flow rate along the wellbore. To fully realize the value of these intelligent completions, there is a need for a systematic data analysis process to improve our understanding of reservoir and production conditions using the acquired data and to make decisions for well performance optimization.

We have successfully developed a model to predict well flowing pressure and temperature (i.e. the forward model), and applied an inversion method to detect water and gas entry into wellbore using synthetic data generated by the forward model (i.e. the inversion model). It is concluded that temperature profiles could provide sufficient information to identify fluid entries, especially in gas wells. However, both the mathematical complexity and advanced well structure lead to challenges in model validation and application. In this paper, hypothetical examples based on previously published field example were used to explain how the interpretation model works. Some parameters used in the model are discussed. Practical guidelines on how to initialize the inversion process and achieve a quick conversion are presented. Judgments should be used based on the understanding of temperature and pressure behavior when initializing the forward model and this can increase efficiency of model application. The study results and guidelines developed in this study will help us to design permanent monitoring systems and set realistic expectation for predictive capability of intelligent well systems.