CO2 sequestration monitoring using a high-precision time-lapse reverse time migration method

Carbon dioxide (CO₂) capture and sequestration (CCS) is a key strategy in mitigating greenhouse gas emissions. CCS demands that the captured gas remains securely within the geological formation to prevent unintended migration or leakage. This necessitates continuous monitoring during and after injection to mitigate economic losses and potential damage to local ecosystems due to leaks. Among various monitoring techniques, seismic monitoring is recognized as one of the most effective methods for CCS projects globally. Moreover, the emerging carbon trading systems require precise quantification of CO₂ migration. Early CCS projects have utilized seismic time-lapse analysis, examining temporal changes in seismic attributes. Techniques such as normal moveout velocity analysis or AVO (Amplitude Versus Offset) analysis were employed to qualitatively assess changes in subsurface parameters caused by CO₂ injection through variations in amplitude and time shifts in specific horizons. However, these techniques suffer from the loss of pre-stack information, like angle and amplitude, which can compromise the accuracy of model-building outcomes. To improve the effectiveness of CCS monitoring, it's crucial to develop or refine methods that retain and leverage the full spectrum of seismic data, ensuring both safety and economic viability of these environmental initiatives.

Theoretically, the Reverse Time Migration (RTM) method can utilize all reflected and scattered waves for imaging, serving as an advanced technique for CO₂ migration imaging. RTM employs Claerbout's imaging principle to locate subsurface reflectors by correlating the source-side forward-propagating wavefield with the receiver-side backward-propagating wavefield. From a rigorous theoretical perspective, imaging algorithms should apply the inverse operator of the forward operator to seismic data. However, the aforementioned imaging algorithm applies the adjoint operator (conjugate transpose) of the forward operator to seismic data, resulting in low resolution and unbalanced amplitude in the final imaging result. This abstract introduces a high-precision time-lapse seismic imaging method based on accurate two-way illumination compensation of the seismic source- and receiver-side wavefields. The proposed method extends the two-way illumination compensation RTM approach, developed for seismic exploration using a scattering integration algorithm (SI-RTM), into time-lapse mode for CO₂ sequestration monitoring. The SI-RTM method accurately calculates the diagonal elements of the Hessian operator, allowing for better preconditioning of the final subsurface image, thereby improving amplitude preservation in the imaging outcomes. The proposed time-lapse SI-RTM method capitalizes on the advantages of SI-RTM by reducing the impact of acquisition geometry and alleviating the need for data consistency in the time-lapse imaging algorithm. Numerical experiments using the Kimberlina CO₂ sequestration model demonstrate that the proposed time-lapse SI-RTM method enhances the image quality of deep CO₂ reservoirs. It minimizes the impact of inconsistent acquisition geometry for different surveys and better reconstructs the subsurface changes caused by CO₂ injection/migration, facilitating high-resolution monitoring of CO₂ migration.