Abstract

In this paper we identify and characterize known and new environmental consequences associated with CO2 capture from power plants, transport by pipeline and storage in geological formations (CCS). The DPSIR framework, describing environmental Drivers, Pressures, States, Impacts and Responses, is used to systematically review environmental impact assessment procedures and scientific literature on CCS. Also, it is investigated whether crucial knowledge on environmental impacts is lacking that may postpone the implementation of CCS projects. The findings of this study are that the capture of CO2 from power plants results in a change in the environmental profile of the power plant. This change encompasses trade-offs and synergies in the reduction of key atmospheric emissions, being: NOx, SO2, NH3, particulate matter, Hg, HF and HCl. The largest trade-offs are found for the emission of NOx and NH3 when equipping power plants with post-combustion capture. Synergy is expected for SO2 emissions, which are low for all power plants with CO2 capture. An increase in water consumption ranging between 32% and 93% and an increase in waste and by-product creation with tens of kilotonnes annually is expected for a large-scale power plant (1 GWe), but exact flows and composition are uncertain. The cross-media effects of CO2 capture are found to be uncertain and not quantified. For the assessment of the safety of CO2 transport by pipeline at high pressure an important knowledge gap is the absence of validated release and dispersion models for CO2 releases due to pipeline failures. There is also uncertainty in estimating the failure rates for CO2 pipelines. Furthermore, uniform CO2 exposure thresholds, detailed dose-response models and specific CO2 pipeline regulation are absent. Most gaps in environmental information regarding the CCS chain are identified and characterized for the risk assessment of the underground, non-engineered, part of the storage activity. This uncertainty is considered to be larger for aquifers than for hydrocarbon reservoirs. Failure rates are found to be heavily based on expert opinions and the doseresponse models for ecosystems or target species are not yet developed. Integration and validation of various sub-models describing fate and transport of CO2 in various compartments of the geosphere is at an infant stage. Concluding, it is not possible to execute a quantitative risk assessment for the non-engineered part of the storage activity with high confidence. Finally, several recommendations have been formulated to deal with the knowledge gaps identified in this study.