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

Task 4:  Large-Scale Hydrological Impacts of CO2 Geological Storage

Task 2 Leads:  Jens Birkholzer and Quanlin Zhou

Overview:

      LBNL will continue research activities on evaluating large-scale hydrological and environmental impacts of GCS. We propose for FY12 a realignment of the three-year research plan that was developed a few years ago for the time period FY10 through FY12. This research plan included three subtasks as follows, Subtask 4.1: Improve Understanding and Prediction Reliability of Basin-Scale Brine Pressurization and Migration; Subtask 4.2: Evaluate the Impact of Brine Pressurization on Groundwater Resources and Provide Technical Basis for Area-of-Review Determination; and Subtask 4.3: Evaluate Pressure Management Schemes via Brine Extraction to Improve Storage Capacity and Remediate CO2 Leakage.
     Subtask 4.2, with its several research activities, will be finalized by the end of FY11, as scheduled in the three-year research plan, and will not be continued in FY12. Subtask 4.3 started in the second quarter of FY11 and will continue through FY12, as scheduled in the three-year research plan (more details below); however, this subtask will be renamed as Subtask 4.2 in FY12. With respect to Subtask 4.1, we have in FY10 and FY11 finalized the first two of three planned activities, namely (1) a detailed sensitivity analysis to identify the most sensitive parameters and detailed literature and data review, and (2) high-resolution modeling studies of small- to intermediate-scale heterogeneity in permeability to better understand pressure propagation. The remaining activity, to be conducted in FY12 according to the three-year research plan, is validation modeling of a selected industrial-scale application site with plenty of near- and far-field pressure data (e.g., from a large RCSP Phase III demonstration project). We plan to defer this activity to FY13 or FY14, when more data from RCSP Phase III demonstration projects will be available. Instead, as the research goals are closely aligned with the objectives of Task 4, we propose here to start development of a high-performance regional-scale simulation model to conduct predictive simulations of CO2 flow, pressure propagation, and brine migration in response to multi-site CO2 injection in the Basal Aquifer in the Northern Plains – Prairie Region of North America.
     The northern Great Plains basal aquifer system extends over nearly 575,000 square miles of the north-central United States and south-central Canada, forming a deep saline system that is considered a very important target for CO2 storage in North America. A bi-national Canada-USA multi-organizational consortium, led by Alberta Innovates – Technology Futures (AITF) in Edmonton for the Canadian side and the Energy and Environmental Research Center (EERC) at the University of North Dakota for the US side, started in FY11 a three-year project to: 1) assess the volumetric and dynamic CO2 storage capacity of the Northern Plains – Prairie Basal Aquifer, 2) assess the effect of injecting such large volumes of CO2 on resident brine and on shallow groundwater resources in areas of aquifer outcrop in Manitoba, South Dakota and Montana; 3) assess the effect of potential leakage of CO2 and/or brine through wells that penetrate this aquifer. As a partner in this consortium, LBNL is responsible for development and application of the high-performance regional-scale simulation model that is necessary to assess dynamic storage capacity and environmental impact. LBNL’s modeling work is planned to start in early FY12, when ongoing work on CO2 source evaluation and hydrogeological aquifer characterization performed by the other consortium partners has been finalized. We propose to incorporate the model development and application work for the Northern Plains – Prairie Basal Aquifer as Task 4.1 of LBNL’s Consolidated Sequestration Research Project. We expect this work to be conducted within the next two years, in FY12 and FY13.

The Large-Scale Hydrological Impacts project has two primary subtasks:

Subtask 4.1 Basin-Scale and Local-Scale Simulation Models for Evaluation of Dynamic Storage Capacity, Brine Displacement, and Groundwater Impact: 
A high-performance regional-scale simulation model will be developed to conduct predictive simulations of CO2 flow, pressure propagation, and brine migration in response to multi-site CO2 injection in the Northern Plains – Prairie Basal Aquifer. The purpose of the basin-scale model is to:

  1. determine the distribution, migration, and long term fate of multiple CO2 plumes corresponding to large CO2 sources in the region,
  2. evaluate the pressure perturbation and brine migration effect at the basin scale, and
  3. evaluate the dynamic storage capacity of the aquifer based on the predicted pressure build-up and brine migration results.

The work scope in FY12 and FY13 can be summarized as follows:

  1. Basin-Scale Model - Development Phase (FY12): Model development will be conducted in close collaboration with other consortium partners, in particular those engaged in source and aquifer characterization. LBNL will review all available geologic and hydrogeologic data to develop model input parameters. For model development, LBNL will perform the following research activities:
    • Determine the appropriate model domain of the basin-scale simulation model, including the target formation and overlying formations, based on literature and data review, scoping simulations, and discussion with other consortium partners;
    • Develop a basin-scale model grid with adequate far-field boundaries and appropriate local refinement around projected plume areas and known faults, on the basis of the static geologic model to be provided by EERC and AITF;
    • Parameterize the model based on existing well data and other geologic and hydrologic information, and constrain the large-scale model parameters by using and analyzing existing pumping/injection tests, if necessary;
    • Calibrate model parameters (e.g., large-scale permeability) using existing observations on hydraulic head and salinity in terms of regional-scale recharge, discharge, and flow rates, which will be provided by EERC and AITF;
    • Explore and test high-performance computational options, such as super-computers available at the LBNL’s super-computer center (NERSC), and a large Linux cluster owned by LBNL geologic sequestration program;
    • Perform preliminary simulations using the parallel TOUGH2/ECO2N simulator to explore the fate of CO2 in the storage formation and to predict pressure response and brine migration.
    • Perform preliminary sensitivity simulations to assess model uncertainty and guide additional aquifer characterization needs.
  2. Basin-Scale Model - Model Application Phase (FY13): LBNL will work with other consortium partners to develop a set of potential future injection and storage scenarios. The storage scenarios may include low, medium, and high injection rates, may consider different injection rates in subregions of the Basal Aquifer, and may account for staged implementation of CO2 storage in the region with early projects and late projects. For each storage scenario, LBNL will perform the following activities:
    • Develop a scenario-based numerical model, with local refinement around each storage project and known faults;
    • Use the calibrated basin-scale model to predict pressure response within the injection formation as well as relevant overlying formations;
    • Evaluate dynamic storage capacity of the aquifer based on pressure buildup results;
    • Evaluate the fate of brine displaced as a result of injection in terms of direction, rate and extent, and of the potential of affecting the shallow areas of the aquifer where it serves as groundwater supply;
    • Compare predictions from detailed regional-scale model with more simplified semi-analytical solutions developed by other consortium partners, e.g., the Princeton University Group;
    • Conduct sensitivity analysis with respect to reservoir formation and primary seal properties, fault properties, and sand/shale interbedding;
    • Use the regional-scale model to design a large-scale monitoring program for pressure impacts and assess the feasibility of novel monitoring options (e.g., InSAR, tiltmeter data) to measure land surface deformations.

Subtask 4.2 Evaluate Pressure Management Schemes via Brine Extraction to Improve Storage Capacity and Remediate CO2 Leakage:
     Creative pressure management schemes can be used to lower pressure-related constraints on storage capacity and remediate CO2 leakage when occurring. These schemes, which become particularly important for industrial-scale applications, may involve extraction of resident brine from storage formations and re-injection into overlying/underlying saline aquifers, or pressure-driven brine transfer in passive pressure relief wells from the storage formations to overlying/underlying saline formations. A pumping/re-injection scheme has been successfully employed since 1953 in the Herscher natural gas storage field in Illinois, where gas leakage out of the reservoir was detected in shallow wells, and reduced by extracting brine at the periphery of the gas bubble and re-injecting it into the immediately overlying formation (Buschbach and Bond, 1974). Subtask 4.2 will evaluate the technical feasibility of pressure management schemes for geologic sequestration, and will investigate the fundamental issues and potential merits of the suggested pressure manipulation schemes by means of analytical and numerical simulations for hypothetical injection scenarios. The task activities include (1) development of a suite of possible pressure management options, (2) design and optimization of pressure management strategies (i.e., well patterns, withdrawal rates and duration, cost estimates) via analytical solutions and numerical modeling for hypothetical injection scenarios and different management options.
     In FY11, we have developed a portfolio of possible pressure management options, including active extraction, passive pressure relief wells, combined active and passive wells, active extraction coupled with re-injection into suitable formations. We have also developed the concept of “impact-driven pressure management (IDPM)”, which means targeted (or local) fluid extractions that are optimized to meet local performance criteria (i.e., that limit pressure increases where environmental impact is a concern). We applied a new single-phase analytical solution for pressure management studies to evaluate the effectiveness of passive and/or active wells in mitigating (managing) pressure buildup in the storage formation. In these studies, we designed well patterns and brine extraction rates in a trial-and-error mode to achieve a given target performance criterion (e.g., a maximum allowable pressure increase in a fault zone). In FY12, we will develop an automated inverse modeling method (such as iTOUGH2) to design and optimize “impact-driven” pressure management options for CO2 sequestration projects. We will initially evaluate simplified sequestration scenarios using the above-mentioned analytical solution, which produces calculation results much faster than numerical models and is thus particularly useful in automated inversion procedures. In the second step, we will move to more complicated scenarios with more complexity with regards to hydrogeology, heterogeneity, and processes. In these cases, the inverse modeling method will be coupled to a multi-phase numerical simulation model to account for the complexities that analytical solutions are unable to handle.