Optimizing Drilling Efficiency: A Fast Fourier Transform Approach to Energy Loss Detection from Surface Measurements

Abstract This study presents a novel method for detecting inefficiencies and dysfunctions in drilling operations using Fast Fourier Transform (FFT) analysis of surface torque and pump pressure variations across various well construction activities. By focusing on energy as the key indicator of inefficiency, this method enables the identification of specific parameters adjustments to improve efficiency and prevent undesired events. FFT is applied to surface measurements recorded at standard sampling rates, such as every 5 seconds, to decompose signals into frequency components, computing a single scalar energy value, which can be correlated to surface parameters, in a constant time window, identifying energy changes that correlate with potential dysfunctions or downhole issues. The method offers a computationally efficient and accessible solution for real-time monitoring and post-operation analysis, particularly in scenarios where high-frequency data, downhole measurements, and/or sophisticated diagnostic tools are unavailable. However, when these tools are available, this method can also be used to correlate specific dysfunctions, providing insights that can be applied to future projects lacking such advanced resources. The incorporation of game theory further enhances the method by identifying key parameters contributing to inefficiencies, offering guidance for parameter adjustments to optimize efficiency. The results demonstrate a strong correlation between total energy derived from surface measurements and instances of dysfunctions in the drill string or bottom hole assembly (BHA). During problem-free intervals, energy values remained consistently low, reflecting efficient operations. Conversely, significant deviations in energy readings corresponded to dysfunctions such as erratic pipe motion, inadequate hole cleaning, and other operational problems. This method’s flexibility—particularly its adaptability to lower sampling rates, low computational cost, and ability to operate independently of downhole telemetry and high-frequency data— makes it an ideal candidate for standard use in optimizing well construction processes in a wide range of projects.