Refined Processing and Quantitative Interpretation of DAS-Monitored Steady-State Injection Profiles

For a long time, improving extraction efficiency while minimizing environmental hazards has been a key concern for petroleum reservoirs worldwide. Distributed acoustic sensing (DAS), being environmentally friendly, offers a new solution for long-term monitoring of production and injection wells. However, in actual logging applications, this technology still faces challenges including low-amplitude responses and significant background disturbance. This study proposes an integrated processing and interpretation workflow for steady-state injection profiles based on DAS monitoring data, which sequentially addresses propagation-related amplitude attenuation, complex noise interference, and the mixed vibration responses of wellbore and reservoir fluids. Firstly, amplitude compensation is applied to the raw signals through wavefront diffusion energy compensation. Then, systematic denoising of the DAS data is performed using a denoising algorithm that integrates whale optimization algorithm (WOA)-optimized variational mode decomposition (VMD), wavelet-threshold filtering and the OTSU global threshold. Finally, the processed results are subjected to Hough transform line detection to divide the frequency bands of vibration signals, and quantitative analysis is conducted on the reservoir-fluid-related frequency components using the integral-area method. The workflow is evaluated using one water-injection well with three active water-intake intervals, with isotope logging used as an independent reference. The DAS-derived relative injection contributions are 67.22%, 8.30%, and 24.48%, compared with 68.20%, 8.44%, and 23.36% from isotope logging, respectively, yielding an overall MAE of 0.75 percentage points, RMSE of 0.86 percentage points, and MAPE of 2.63%. These results show good agreement with the isotope-logging reference for the investigated field case and support the feasibility of the integrated workflow for semi-quantitative estimation of relative zonal injection contributions under the investigated conditions.