Successful storage of CO₂ in underground aquifers requires robust monitoring schemes for detecting potential leakage. To aid in this challenge we propose to use statistical approaches to gauge the value of seismic monitoring schemes in decision support systems. The new framework is based on geostatistical uncertainty modeling, reservoir simulations of the CO₂ plume in the aquifer, and the associated synthetic seismic response for both leak and seal scenarios. From a large set of simulations we assess the leak and seal conditional probabilities given seismic data over time, and build on this to compute the value of information of the seismic monitoring schemes. The Smeaheia aquifer west of Norway is used to exemplify the approach for early leakage detection and decision support regarding CO₂ storage projects. For this case study, we find that the optimal monitoring time is about 10 years after injection starts.

成功地将CO ₂储存在地下含水层中需要鲁棒的监测方案以检测潜在的泄漏。为了应对这一挑战，我们建议使用统计方法来评估决策支持系统中地震监测方案的价值。新框架基于地统计不确定性建模，CO ₂储层模拟含水层中的羽状流，以及针对泄漏和密封情况的相关综合地震响应。通过大量模拟，我们评估了随时间变化的给定地震数据的泄漏和密封条件概率，并以此为基础来计算地震监测方案的信息价值。挪威西部的Smeaheia含水层用于举例说明有关CO ₂封存项目的早期泄漏检测和决策支持的方法。对于本案例研究，我们发现最佳的监视时间是注射开始后约10年。