Grantee Research Project Results
2011 Progress Report: Expert-Based Development of a Site-Specifc Standard in CO2 Sequestration Monitoring Technology
EPA Grant Number: R834384Title: Expert-Based Development of a Site-Specifc Standard in CO2 Sequestration Monitoring Technology
Investigators: Nicot, Jean-Philippe , Hovorka, Susan D.
Current Investigators: Nicot, Jean-Philippe , Hovorka, Susan D. , Remington, Randy L , Sun, Alex , Yang, Changbing , Sava, Diana , Zeidouni, Mehdi , Mickler, Pat
Institution: The University of Texas at Austin
EPA Project Officer: Aja, Hayley
Project Period: December 1, 2009 through November 30, 2012 (Extended to November 30, 2013)
Project Period Covered by this Report: December 1, 2010 through November 30,2011
Project Amount: $899,958
RFA: Integrated Design, Modeling, and Monitoring of Geologic Sequestration of Anthropogenic Carbon Dioxide to Safeguard Sources of Drinking Water (2009) RFA Text | Recipients Lists
Research Category: Drinking Water , Water
Objective:
(1) Quantitatively evaluate potential monitoring strategies to select an array of strategies and guidelines for application to specific sites;
(2) Test the results of evaluation against the growing array of field measurements, gathered from past and current test sites in the United States and the world;
(3) Develop widespread consensus that these strategies are adequate when properly applied; and
(4) Compile a test/teaching set of cases for testing strategies and then train practitioners in applying the strategies to an array of sites.
Progress Summary:
During the second year, major accomplishments include (1) extensive effort seeking input and updates from experts in the field, (2) study of progress made by EPA in developing guidance and assessment of updates from parallel international work on monitoring, (3) developing models and analytical tools to assess effectiveness of various monitoring approaches, and (4) presentations for expert review assessing how this information can be effectively conveyed to site developers and regulators.
This second year report was updated in May 2012, to assess progress toward completion. Several elements have taken longer than anticipated because of complexity and data availability. Data are now being released from a number of field studies that must be input into the analysis underway. In addition, a breakthrough in locating software to simply assess the impact of site-specific variables on time-lapse (4-D) seismic has been achieved; however, the analysis is incomplete. Combination and standardization of the analytical approaches developed has just been started; this is a key step in preparing a workbook for review and testing.
2.1 Project Status
The purpose of this study is to provide technical background information to support regulators and project developers in matching selected monitoring tools and protocols with sites and uncertainties at those sites. Because monitoring for GS projects is rapidly evolving, during Year 2 we have continued the process of Year 1, drawing information from diverse sources and discussing our ideas with many researchers and policy developers. Published overviews reviewed this year include Shell Canada Limited (2010); Canadian CCS standards (2011), and Det Norske Veritas, (2009 and 2010). Technical expertise is provided by tool developers and data reduction specialists, and monitoring experts with field experience provide information on performance in terms of durability and sensitivity of tools under actual field conditions. Significant information from regulators was gained through analysis of draft guidance (Hovorka, written communication, 2012). Information was gained from industries with potential to supply CO2, and from policy developers. In addition, during this year the public has provided input on the areas of concern that should be considered in developing a monitoring program. Calculations using site-specific parameters and measured uncertainties and noise for field cases provide the inputs to this study. Conversion of these data from test cases to the generalized format needed by this project is partly completed.
Results from Year 2
3.1 Development of Nomograms
The development of easily accessible and standardized plots of tool sensitivity has proved more difficult than anticipated during conceptualization. In Year 1, we prepared an inventory of tools and parameters that we recommend be considered in monitoring tool selection. First, principles for assessment of sensitivity of tools, e.g., under what conditions is it expected that a tool would be sensitive, were prepared. An update of this inventory was submitted to EPA.
3.1.1 Difficulty in developing standardized plots.
Three problems have been encountered in developing standardized plots of sensitivity of monitoring tools to site-specific parameters. First, tool sensitivity is a complex function, with a large number of mutually interacting variables. For example, one deployment of a tool may be optimum for one site and a different deployment optimum for a different site. How then do we quantify the sensitivity of the tool? This complex integration of deployment with sensitivity is particularly difficult in geophysics. The solution we are using is to standardize the tool-specific parameters to values suitable for a typical tool at an average deployment, biasing decisions toward the sites where we are aware storage is occurring. Notes associated with the analysis make clear that site-specific and tool-specific optimization are possible. However, significant work remains to complete this optimization to make results broadly applicable, easy to use, and technically rigorous.
The second barrier to simplification results from our ambition to provide quantitative comparisons among tools. Different classes of tools measure the perturbation that would be created by leakage differently. For example, confined-zone pressure response is sensitive to the rate of change and therefore is expressed by volume leaked over time. A soil gas flux measurement, however, has an explicit area specification required because it is detected over a plane; this is not needed by a pressure signal. Geochemical detection of leakage has a third type of expression, where the leakage rate can be much less relevant than it is in a pressure or flux measurement because the geochemical signal (absent transport considerations) is a function of concentration. However, concentration responds to not only the leakage rate and the time over which it occurred but volume of water into which the leakage occurred. Water volume, in turn is linked to assumptions about leakage mechanism and pathways. The methods considered to manage these intrinsic non-parallel measurement units of different methods include non-parametric models and standardization of conditions. If done correctly, the non-parallel measurement units can be made less intrusive. In all cases, the models presented will be prototypes leading site developers and regulators to further calculate unique solutions at their site.
The third area of uncertainty is the noise in the measurement. Noise is a strongly site-specific parameter, and is a key threshold in determining which measurements will be successful in leakage detection. Our approach to noise is to collect data and provide site-specific representations of noise. Noise comes from different sources for different tools. For example, in soil gas flux, noise is related to environmental variables. In time-lapse geophysics, noise is related to the complexity of the geologic environments through which the signal must past. Based on field studies, we are preparing guidance on how to assess noise in the environment during characterization and intersect this with the signal calculation to determine if a tool is adequately sensitive.
3.1.2 Progress in developing standardized plots (nomograms).
3.1.2.1 Pressure in above-zone monitoring intervals (AZMI)
Monitoring pressure in an above-zone monitoring interval (AZMI) is proposed as a mechanism for detecting any fluid (CO2 or brine) that leaves the injection zone, assuming that the flow path intersects the AZMI. To assess the sensitivity of pressure changes in the AZMI to fluid flow, a new single-phase analytical model for fluid flow through a leaky fault was developed. The analytical model can be used to evaluate the leakage rate and pressure perturbations related to fault leakage both in the injection zone and in an overlying formation separated by an impermeable confining layer. The solution is extended to evaluate the vertical leakage attenuation considering multiple overlying formations with alternating aquitards. Such calculations can be done quickly without the need for spatial and time discretization, which can ease uncertainty quantification and Monte-Carlo type analysis. The model also can be used to characterize a fault based on above-zone pressure monitoring.
3.1.2.2 Geochemical modeling
EPA traditionally relies on geochemical methods to track changes in water quality, and underground sources of drinking water are the key protected resources in the Underground Injection Control (UIC) program. During Year 2, we completed a sensitivity study of groundwater geochemistry to potential CO2 leakage. The test case used was a geochemical model comparison of sensitivity of the chemistry of Cranfield and SACROC CO2 injection areas shallow freshwater aquifers to changes resulting from introduction of CO2 from a leak. In this phase of the study, transport within the aquifer was not assessed.
It is of interest to study the sensitivity of groundwater geochemistry, especially pH and carbonate parameters (dissolved inorganic carbon [DIC], HCO3-, alkalinity and stable carbon isotopes of DIC) because the Cranfield and SACROC shallow aquifers represent two major aquifer types: with and without carbonates in the aquifer sediments. Groundwater chemistry data from the two shallow aquifers indeed show significant differences. Groundwater pH, and HCO3- in the Cranfield shallow aquifer are lower than in the SACROC shallow aquifer. Ratios of HCO3/SiO2 for most water samples collected from the SACROC shallow aquifer are higher than 10, suggesting groundwater chemistry is dominated by carbonate mineral weathering. Ratios of HCO3/SiO2 for most water samples collected from the Cranfield shallow aquifer are smaller than 5, suggesting groundwater chemistry is dominated by silicate mineral weathering. δ13C of DIC is heavier in the SACROC shallow aquifer than in the Cranfield shallow aquifer because of presence of carbonate minerals in the SACROC shallow aquifer sediments.
Sensitivity of pH and carbonate parameters to potential CO2 leakage is evaluated using a modeling approach. Results suggest that DIC shows a strong sensitivity to CO2 leakage in both aquifers. HCO3- and alkalinity show a strong sensitivity in the SACROC shallow aquifer but a weak sensitivity in the Cranfield shallow aquifer. Groundwater pH shows a strong sensitivity in the Cranfield shallow aquifer but a weak sensitivity in the SACROC shallow aquifer. δ13C of DIC shows a strong sensitivity in the Cranfield shallow aquifer because of simple end members of carbon sources and significant shift between δ13C of DIC in the background and CO2 injected.
3.1.2.3 Seismic sensitivity
Numerical modeling is classically used to determine if changes in seismic response to emplacement of CO2 in a storage site is expected to be successful. We recently received a donation of software (FSM_SIMULATOR) developed by the Carbon Capture Project (CCP) that will let us plot the seismic sensitivity to fluid changes against input of the estimated elastic moduli for the saturated fluid, rock frames, and the porous rocks. Modeling using this software has not yet started, and therefore no appendix of supporting information has been generated.
3.1.2.4 Thermal monitoring
Thermal change is a classic well-leakage detection technique. Warm fluids migrating upward within or adjacent to a leaking well are used commercially for leak detection and to design well remediation. In addition, cooling during flash of CO2 from dense to gas phase can be detected with thermal monitoring. In this study, we develop code that lets us assess the sensitivity of this method with respect to rate of leakage under various scenarios. Calibration data are used from well-based measurements at field studies at Cranfield to show the uncertainties that arise from the simulation. In addition, we undertook an assessment of how far from the leak point the thermal change could be measured. This is relevant to fault leakage, where the zone of maximum flow may be focused at one point in the fault plane and the well at which thermal profiles are measured is some distance away. This model can be calibrated with data from in-zone injection at Cranfield. Additional work is needed to make these models robust.
3.1.2.5 Monitoring changes in shallow soil/groundwater isotopes
The major criterion in determining if the leakage signal is sensitive using carbon isotopes is the contrast between the δ13C of ambient near-surface carbon sources and the injected CO2 δ13C. However, downward transport by rainwater has a significant impact on sensitivity through dilution. The CO2 mixing model developed for this study assesses the interactions of rainfall and temperature, and experiments with simplifying assumptions and types of output. The influence of injectate CO2 on the vadose gas CO2 δ13C values can be recognized when the natural variability in soil-derived CO2 δ13C values is exceeded.
3.1.2.6 Uncertainty
Uncertainty quantification (UQ) is a necessary step in risk assessment of CO2 sequestration. Risks are associated with, for example, overpressure in confining formation and leakage to AZMI through leaky faults and abandoned wells. The purpose of UQ is two-fold. First, it helps identify dominant system and environmental variables that contribute to variability in the output, which is important for data collection and design of operational and monitoring strategies. Second, UQ yields bounding scenarios (or worst-case scenarios) for a particular set of variables and thus provides direct input to robust risk assessment.
As one case, we have developed a pressure-anomaly-based inversion algorithm for locating leaky wells and leakage rates. This methodology provides a framework into which other types of assessments can be completed.
3.2 Model Case Selection
Models for case studies have so far been derived from GCCCs field projects, which has provided substantive leverage suitable for a wider spectrum of sites because of the probabilistic approaches used. Using a range of values is useful in producing failure scenarios, where leakage of fluids out of the injection zone have been modeled to give a range of responses, such as lateral migration in excess of predicted, pressure build-up in the reservoir, fluid leakage to shallower zones, or slow to fast leakage of CO2 into underground source drinking water (USDW).
An important contribution to project completion has been the project team's participation in monitoring field projects. The SECARB project at Cranfield, Mississippi, has completed 3 years of monitoring under separate funding but provides input into the recommendations in development for this study. These monitoring data provide a robust starting point for pragmatic assessment of site-specific applicability of monitoring approaches. Care is taken not to over-generalize the results of tool deployment at this setting; however, the intersection of monitoring approaches with a real case study provides deep experience on viable approaches, both those that are sensitive at this setting but might be obscured by noise at another setting, and more commonly, the reverse, tools and approaches that are poorly sensitive at this setting but might be successful in a different setting. Above-zone pressure is an example of a tool that shows promise of working well at Cranfield where a well-defined zone at suitable depth is available, but less well where too many, too few, too discontinuous, or too thick above zone units are found. Examples of tools for which the needed signal is degraded by noise at Cranfield are in-zone pressure monitoring and time-lapse seismic.
Additional value is obtained by planning site-specific monitoring at two field injection sites in development. These are Hastings Field in Brazoria County, Texas, where anthropogenic CO2 will be injected in 2014, and West Ranch Field, Jackson County, Texas, where CO2 captured from NRG's J.W. Parrish power plant will be injected. Planning monitoring for these sites has accelerated the development of site-specific monitoring approaches and provided substantive practical as well as conceptual leverage for the work done this year to develop tools for assessment of sensitivity.
Another major method of data collection has been seminars, conference presentations, by-invitation reviews, and peer review of manuscripts and draft regulations. Through this process, we inventoried U.S. and global progress on performance of monitoring tools as well as available information on best approaches. Because of rapid evolution, in-person exchange of ideas has been essential, as we note that publications lag public release of results by more than a year. In-person discussion also provided feedback on the approaches in development for this study. In addition, our team has been working with a number of project developers, which has sharpened our understanding of needs and issues exposed by developing monitoring plans at field sites.
Future Activities:
A main Year 3 activity is to assess nomograms of tool sensitivity against model test cases. As additional field data become available, we will keep incorporating data into the data sets.
The second main Year 3 activity is to organize the test cases into a teaching set in the form of a manual. This manual will be used by prospective site developers and regulators for training in designing an appropriate site and specific monitoring plan. A phased period of reviews is needed to assure that this manual attains widespread acceptance.
References:
Canadian CCS standards, 2011, Geological storage of carbon dioxide, CSA Z741-11, public review draft (final not yet released).
Det Norske Vertitas, 2009, Guideline for Selection and Qualification of Sites and Projects for Geological Storage of CO2 DNV Report No: 2009-1425.
Det Norske Vertitas, 2010 CO2QUALSTORE Guidance for users of the guideline, DNV Report No.: 2010-0036.
Environmental Protection Agency, 2010a (part RR and UU). Mandatory Reporting of Greenhouse Gases: Injection and Geologic Sequestration of Carbon Dioxide, Federal Register CFR 40 parts 72, 78, and 98; http://edocket.access.gpo.gov/2010/pdf/2010-29934.pdf.
Environmental Protection Agency, 2010b (UIC program). Federal requirements under the underground injection control (UIC) program for carbon dioxide (CO2 ) geologic sequestration, (GS) wells, http://www.gpo.gov/fdsys/pkg/FR-2010-12-10/pdf/2010-29954.pdf.
Environmental Protection Agency, 2010. General Technical Support Document for Injection and geologic sequestration of carbon dioxide: subparts RR and UU (Clean Air Act) Greenhouse Gas reporting program. Download from https://www.epa.gov/climatechange/emissions/downloads10/Subpart-RR-UU_TSD.pdf last accessed May, 2012.
Environmental Protection Agency, 2012. Draft Underground Injection Control (UIC) Program Class VI Well Testing and Monitoring Guidance, http://water.epa.gov/type/groundwater/uic/class6/upload/GS_TestingMonitoring_Guidance_DRAFT_01262012_508.pdf, accessed March 2012.
Hovorka SD, 2012. Comments on Geologic Sequestration of Carbon Dioxide: Draft Underground Injection Control (UIC) Program Class VI Well Testing and Monitoring Guidance, March 12, 2012.
Li R, Urosevic M, 2006. FSM-Simulator help file, CO2 CRC Report No. RP06-0171.
Li R, Dodds K, Siggins AF, Urosevic M, 2006. A rock physics simulatorand its application for CO2 sequestration processes. Exploration Geophysics 37:87-72.
Shell Canada Limited, 2010. Quest Carbon Capture and Storage Project Volume 1: Project Description Appendix A: Measurement, Monitoring and Verification plan; written communication.
Journal Articles on this Report : 2 Displayed | Download in RIS Format
Other project views: | All 17 publications | 15 publications in selected types | All 15 journal articles |
---|
Type | Citation | ||
---|---|---|---|
|
Sun AY, Nicot J-P. Inversion of pressure anomaly data for detecting leakage at geologic carbon sequestration sites. Advances in Water Resources 2012;44:20-29. |
R834384 (2011) R834384 (2012) R834384 (Final) |
Exit Exit Exit |
|
Sun AY, Zeidouni M, Nicot J-P, Lu Z, Zhang D. Assessing leakage detectability at geologic CO2 sequestration sites using the probabilistic collocation method. Advances in Water Resources 2013;56:49-60. |
R834384 (2011) R834384 (Final) |
Exit Exit Exit |
Progress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.