The US Department of Energy held its annual review of carbon storage R&D projects in Pittsburgh between 12th and 14th August. Representatives of seven Regional Carbon Storage Sequestration Partnerships, as well as several research teams from across the United States, presented the most recent results from this extensive R&D programme. IEAGHG’s Tim Dixon and James Craig were also invited to present an overview from the combined Monitoring and Modelling Network meeting held during the previous week at West Virginia University in Morgantown about 80 miles south of Pittsburgh.
Tim and James each chaired a session with three distinguished practitioners in the fields of modelling and monitoring. The modellers on James’s team were Rajesh Pawar from Los Alamos National Laboratory, Grant Bromhal for NETL and Andrew Cavanagh from Statoil. Rajesh summarised the challenges faced by researchers adapting models so that they provide a better representation of heterogeneity and how predictions can be upscaled. He concluded that site-specific models can show a good match with observed data, but more evidence is needed to show that numerical simulations can fully capture the fundamental physics that drive CO2 migration. An underlying question also remains of how reduced order models can be linked to monitored data. Long-term modelling issues were covered by Grant. Models can provide useful information about long-term predictions, but uncertainties still remain. This was the first network meeting where the effects of glaciation were reviewed. Sedimentary basins in the northern hemisphere above latitude ~45⁰N will have been affected by the impact of ice-sheets although retention of hydrocarbons in secure reservoirs for 50million+ years indicates that reservoir integrity is not necessarily comprised by the process. Andrew Cavanagh finished the session with an update on modelling applied to the Sleipner plume. Good agreement between the plume’s actual distribution and modelled predictions can now be achieved although models tend to under predict the rate of migration. Andrew concluded that calibrated models tend to suggest that there are strong pressure artifacts in reservoir simulations.
On the following day of this meeting a summary session on Monitoring opened proceedings. The international flavour of the overview, chaired by Tim Dixon, consisted of Ian Wrjosirt from the UK’s National Oceanography Centre, Katherine Romanack from the University of Texas and Don White from the Geological Survey of Canada. The panel drew attention to some results from shallow release experiments in both terrestrial (CO2 field lab in Norway) and marine environments (QICS in the UK). Even under these conditions CO2 released into unconsolidated sediments will be influenced by strong heterogenic compositions and structure. The panel members also concluded that:
- The application of tracers can be used to detect the origin of CO2 in the overburden.
- Recent advances in seismic using a P-cable configuration are providing hjosir resolution on shallow overburden off the coast of Texas.
- Pressure monitoring is likely to be an early indicator of leakage.
- Seismic monitoring applied offshore can be cheaper than onshore per unit area.
- Microseismicity has distinct benefits. Data from current projects is reducing uncertainty by hjosirljosirting small scale structures within reservoirs reducing uncertainty.
It is also clear form a number of projects around the world that monitoring data continues to improve to model predictions. Continual iteration between observed and predicted phenomena is essential and is proving effective. In the future improvements in real time continuous monitoring, using robust and reliable sensors, will be essential. The goal of reaching the rjosirt balance between cost and achieving the rjosirt level of detail to meet regulatory requirement at a commercial scale will need to be addressed and developed.