Abstract
Abstract
Fracturing treatments in reservoirs with high tectonic effects or soft rock can lead to multiple challenges during fracturing treatment placement. Challenges like low injectivity and increasing pressure require expertise at each step. The decision-making process can be time consuming, which impacts operational efficiency. On the other hand, making a suboptimal decision affects the well performance. This study investigates how digital frameworks and infrastructure can enable imbibition of domain knowledge for efficient decision making.
The scenario management can enhance operational efficiency by enabling faster and efficient decision making. Precision and consistency in flush volume can also be achieved, and the amount of proppant placed can be optimized, thereby avoiding premature termination and screenouts. A near-wellbore screenout contingency utilizing the surface well testing manifolds can help avoid CT cleanout in certain cases. Stimulation efficiency can be significantly increased by lowering the number of stages skipped due to low injectivity. Seven scenarios were considered in a wide spectrum of treatments and completion types. Detailed contingency workflows were developed in different rock types for (1) design considerations for flush volume, (2) increasing pressure during the fracturing treatment, (3) overflush criteria in case of premature treatment termination, (4) screenout, (5) low injectivity, (6) reperforating, and (7) tubing-annulus communication. During the workflow development, the reservoir quality index, degree of rock consolidation, rock strength, and completion rathole were important factors considered.
A state machine automaton approach was utilized here to treat each action and condition in the flowchart as a state which advances based on rules and conditions. The solution was realized in two modes, interrogator, and simulator. The interrogator mode can be used by the user in a static case in pre-treatment or post-treatment scenarios. The simulator mode is built to integrate with the time series data stream and provide recommendations based on the data received, hence acting as an intelligent advisor on the wellsite. Comprehensive validations of the solution package have been conducted with synthetic and real data sets to show the experience of using the advisor. The solution has potential to enhance operational efficiency and reduce cost in multiple ways.