Machine Automation in Seismic Interpretation and Modelling: Applications to CO2 Storage. A Case Study from the Gippsland Basin, Victoria, Australia

Abstract

Carbon capture and storage (CCS) projects are of utmost importance in mitigating greenhouse gas emissions and addressing climate change (IPCC, 2005). Accurate reservoir and storage models play a significant role in the success of CCS initiatives (Bachu, 2008). Understanding and characterizing seals, reservoirs, and faults are crucial aspects for developing reliable reservoir and storage models (Shapiro et al., 2004). The mapping and characterization of these geological features enable a better assessment of storage capacity, identification of potential leakage pathways, and optimization of injection and storage processes (Chadwick et al., 2018). This study aims to interpret a modern 3D seismic volume using advanced machine-augmented interpretation technology (Pauget et al., 2009). This globally consistent seismic interpretation and modelling technique provides detailed structural and stratigraphic understanding by providing rapid and high-resolution interpretation throughout the seismic volume. The outputs from this process are then directly fed into unique QI and modeling tools, contributing to the improvement of reservoir characterization for effective CCS implementation. The Gippsland Basin in Australia has a rich petroleum exploration history, with significant discoveries dating back to the 1960s (Evans et al., 2003; Blevin et al., 1998). The basin has been a prolific producer of oil and gas, with major fields such as the Esso-BHP joint venture's Longford and Barracouta fields contributing to Australia's energy resources (Bradshaw & Morse, 2002). A significant amount of historical open-file data is available in the region which provides valuable insight into the geology and potential of the basin for CCS. The abundance of data and proximity to onshore infrastructure made the basin an obvious candidate for potential CCS projects. The Pelican Carbon Capture, Utilization, and Storage (CCUS) project aims to capture and store carbon dioxide (CO2) emissions from an industrial facility located in the Gippsland Basin (Hoffman & Carman, 2015). The project involves the capture of CO2 from a gas-fired power station and its transport via pipeline for permanent storage in a deep saline formation beneath the seabed. In order to better characterize the storage site, a modern high-resolution 3D seismic volume was acquired in 2018 (Pelican 3D) (Hoffman, 2021). The Pelican 3D dataset covers some 160 km2 of the offshore Gippsland basin down to 2 seconds TWT, with a bin spacing of 12.5m × 12.5m and a vertical sample rate of 2 ms. The final SEGY volume is 5.6 GB. Here, we apply a unique automated interpretation methodology to interpret the Pelican 3D volume, with the results used to better understand the structural and stratigraphic relationships in the dataset and provide inputs to a static model.