Earth Sciences Division (ESD) Department of Energy (DOE) Lawrence Berkeley National Laboratory (LBNL)

Task 3:  Sim-SEQ:  A Model Comparison Study Using Results from CO2 Field Tests

Task 3 Leads:  Sumit Mukhopadhyay and Jens Birkholzer


     The Sim-SEQ project intends to objectively evaluate the modeling efforts of different modeling groups as they are applied to the Phase III geologic carbon sequestration (GCS) field tests to be undertaken in the Regional Carbon Sequestration Partnerships (RCSP) in the United States over the next several years. Modeling efforts will first be conducted for one specific RCSP field test site near Cranfield, Mississippi, hereafter referred to as the Sim-SEQ Study Site (the S-3 Site), where CO2 injection started in November 2009. The participating modeling groups will each develop individual models for the S-3 Site, based on a single set of site characterization data as provided below, but using their respective model approaches and simulators. Being able to evaluate different model results for one site guarantees a direct comparison of models and approaches, and allows understanding and explaining their differences, stemming, for example, from process simplifications and parameter choices.

     It is expected that the Sim-SEQ project will demonstrate in an objective manner that the observed system behavior at GCS sites can be predicted with confidence, and that the remaining differences between models and measurements at the S-3 Site, as well as between different models for the same site, are well understood. The project will also ensure that model uncertainties are evaluated and their impact is assessed, and that lessons learned and improvements made by one specific modeling group are documented and available to all other research teams for use to improve future modeling efforts. The project will help to convince stakeholders that the subsurface processes expected in GCS sites are well understood and can be predicted with a reasonable degree of confidence.

     To address the overall objectives outlined above, multiple modeling teams will develop numerical models of one or two selected GCS test sites based on a common set of characterization data, iteratively revising their models to include monitoring data as they become available. Predictions of the system state during natural and test conditions are then used for model-to-data and model-to-model comparisons. Comparison of an individual model’s predictions to data, and model-to-model cross-comparisons of selected, pre-defined outputs (performance measures) will be performed and evaluated against pre-defined acceptance limits. The scope of work described here pertains to the simulation studies performed by each modeling team associated with the project, including activities necessary to participate in discussions and meetings related to the model comparison study. The LBNL Sim-SEQ team coordinates and facilitates all Sim-SEQ activities. The proposed work to be performed consists of:

  •  review of initial characterization data and geological framework model provided by the owners of the S-3 site, as well as review of monitoring data as they become available,
  • conversion of the geological model into one or multiple conceptual models of the hydrogeology at the site,
  • development of parameter sets (and their uncertainties) describing hydrogeological properties,
  • definition of scenarios and cases to be simulated,
  • numerical model development (mesh generation, property assignment, initial and boundary conditions),
  • deterministic and/or probabilistic simulation runs,
  • compilation of results through calculation of pre-defined performance measures (e.g., plume migration, pressure buildup, etc.), and
  • participation in the Sim-SEQ Technical Team and support of its development, simulation, and evaluation activities.

Please visit the official Sim-SEQ website, which provides additional details of the project, historical information, and a discussion page.

The Sim-SEQ Project has several subtasks:

Subtask 3.1  Review of Site Characterization Data

    Familiarization with the S-3 Site and data evaluation: Get familiar with the S-3 site through review data review of site characterization data available on the web portal; read the literature about the S-3 site geology/hydrogeology provided on the web portal; download first set of characterization data and geologic framework model provided by the owners of the sites; develop set of performance measures and related acceptance criteria with other project participants and the Sim-SEQ Technical Team.

    Participate in Data Review Telecon: Report to Sim-SEQ project management team about adequacy of the site characterization data; report any inconsistencies in uploaded data and/or any missing data; submit request for additional data requirement; project management to address issues related to site characterization data.

Subtask 3.2  Initial Model Development

    Conceptual Model: Develop conceptual model of selected site based on analysis of available characterization data and the geologic framework model; document model development process, specifically regarding chosen model parsimony (in terms of simulated process, dimensionality, granularity, etc.) and assumptions in the light of the intended use of the model.

    Numerical Model: Develop a numerical model based on the conceptual model above; document model development, specifically regarding chosen discretization, parameterization/zonation, initial and boundary conditions, property values and their uncertainties.

Subtask 3.3  Initial Predictive Simulations

    Selection of Scenarios and Cases: Define modeling scenarios, as well as simulation and parameter variation cases; describe selected cases and rationale for examining them.

    Simulation Runs: Perform simulations of CO2 field test for all selected scenarios and cases.

SubTask 3.4  Performance Measures

    Evaluate Performance Measures: Sim-SEQ project management, in consultation with the Sim-SEQ Technical Team, will develop performance metrics for comparison of results; modeling teams to compile simulation results calculating pre-defined performance metrics; document measures and discuss uncertainties. Performance metrics may be defined for: migration of the CO2 plume as a function of time, trapping mechanisms such as solubility trapping, residual trapping, evolution of injection pressure and injectivity assessments, large-scale evolution of pressure within reservoir and above it, brine displacement and changes in salinity, stress changes and induced seismicity, land surface uplift, geochemical changes (gas and brine chemistry, precipitation, dissolution), etc.

    Transfer, Review and Discussion of Model Results: Report on and discuss findings, and transfer results of simulations (i.e., performance measures that will be evaluated using acceptance criteria) to Sim-SEQ Technical Team for subsequent model comparison study.

Subtask 3.5  Comparison Study

    Support Comparison Study: Contribute to Sim-SEQ model comparison study (clarification of results, visualizations, evaluation of model-to-model measures and model-to-data and acceptance criteria, conclusions, lessons learned, etc.)

Subtask 3.6  Model Refinement

    Perform Formal Model Calibration and Validation: Calibrate model against pre- and post-testing injection and monitoring data, and perform formal model validation exercise; evaluate prediction uncertainties and model reliability; refine model based on insights gained during calibration and validation activities.

    Transfer, Review and Discussion of Model Results: Report on and discuss calibration and validation results and enhanced modeling studies, exchange gained knowledge, expertise, lessons-learned, and capabilities; plan follow-on phase of Sim-SEQ initiative.

Subtask 3.7  Technical Team

    Participation in Sim-SEQ Technical Team Activities: Members of the modeling teams will join the Sim-SEQ Technical Team, which is the main venue for the discussion and comparative evaluation of model results. The Sim-SEQ Technical Team also moderates the exchange of relevant information among the modeling teams. The Technical Team—which convenes regularly via teleconferences, videoconferences, and occasional workshops—comprises members of all participating RCSP modeling teams, as well as the Lawrence Berkeley National Laboratory (LBNL) Sim-SEQ team.