Search

• Western Nebraska has potential to host a CCS project. • Dedicated storage options are subject to significant uncertainty. • Public outreach would be a vital element for CCS in this region. The Energy & Environmental Research Center (EERC) has conducted a pre-feasibility study for a commercial-scale carbon dioxide (CO 2) geologic storage complex in western Nebraska, integrated with potential CO 2 capture at the coal-fired Gerald Gentleman Station (GGS). This pre-feasibility project has been executed as part of the U.S. Department of Energy (DOE) CarbonSAFE (Carbon Storage Assurance Facility Enterprise) Program, where projects are required to demonstrate the potential to capture and store at least 50 million tonnes (Mt) of CO 2 over a minimum 25-year operational period. Using publicly available data, this study has shown that western Nebraska has potential to host a commercial-scale carbon capture and storage (CCS) project. However, three key challenges would need to be overcome. Firstly, the business case for deploying CCS projects is uncertain; recently announced federal tax credits and sales for enhanced oil recovery might not cover full project costs. Secondly, the potential dedicated storage site defined in this study is at a low level of readiness to support a CCS project. Thirdly, public outreach would be a vital element in western Nebraska, where sensitivities around such environmental issues as water resource protection and pipeline construction would need to be carefully addressed. [ABSTRACT FROM AUTHOR]

• Multiscale workflow to analyze heterogeneity and estimate rock properties. • Data from three scales included thin section and micro-CT images and well logs. • Fractals, Lorenz coefficient, porosity, and permeability estimations across scales. • Combining Lorenz coefficient and fractals to discern differences in pore networks. • Machine learning, fractals, and Lorenz coefficient can connect multiscale data sets. Geologic heterogeneity is a key feature that must be considered when translations of scaled data are performed. This paper presents the assessment of geologic heterogeneity using a multiscale workflow that includes image analysis-based methods coupled with well log analysis to provide data in which fractals and machine learning methods estimate the carbon dioxide (CO 2) storage resource potential of a reservoir. The heterogeneity of rock properties of the complex Bell Creek reservoir in Montana, USA, was explored at the pore scale (∼nm to mm), core scale (∼mm to m), and well scale (∼cm to m). The data used in this study included advanced image analysis of micro-CT (computed tomography) images (pore scale), thin sections (pore scale), plugs and core images (core scale) and well logs (well scale). The micro-CT images were segmented using a U-net segmentation approach into objects of pores and grains. The segmented images were reconstructed into subvolumes of different sizes. Physical properties (porosity and permeability) and fractal dimensions were calculated for the various subvolumes, and Lorenz coefficient (Lc) values, a single parameter to describe the degree of heterogeneity within a pay zone section, were calculated from thin-section images and well logs. Porosity and fractal dimension values were used to estimate the 188-µm threshold of representative elementary volume (REV) in this study. Both the Lc and fractal dimension values were found to be negatively correlated. When these two parameters are combined, it is possible to discern differences in the complex porous networks of the samples analyzed in this study. [ABSTRACT FROM AUTHOR]

Pulsed-neutron well logs (PNLs) were acquired to monitor CO 2 storage associated with enhanced oil recovery. This work quantifies the precision of repeat PNLs using data from four wells and 15 repeat PNLs. Root-mean-square (RMS) precision for the repeat PNLs was less than 3%, indicating good agreement between the baseline and repeat PNLs. Evaluations of scaled relative difference (Scaled-D) showed variation in precision among individual wells and formations. Analysis of false-positive rates (FPRs) across the entire data set showed that a Scaled-D threshold in sandstone formations of approximately ±8% resulted in a 1% FPR. These Scaled-D precision thresholds were used to estimate the value of CO 2 saturation able to be confidently distinguished from baseline. The detection limit for CO 2 is lowest for high-porosity formations filled with saline water and is highest for low-porosity formations filled with fresh water. Thus, detection of vertical out-of-zone CO 2 migration using repeat PNLs is a function of instrument precision, petrophysical properties, and hydrology, all of which must be taken into account as part of the monitoring program. The results of this work provide insight into how PNLs may be included within monitoring plans to detect vertical out-of-zone CO 2 migration along a wellbore or instances of wellbore failure and provide a quantitative basis for establishing detection limits of repeat PNLs to distinguish change from baseline conditions. [ABSTRACT FROM AUTHOR]

Carbon dioxide (CO 2 ) enhanced oil recovery (EOR) will likely be the primary means of geologic CO 2 storage during the early stages of commercial-scale carbon capture and storage (CCS) because of the inherent economic incentives as well as the abundant experience and demonstrated success in the United States, where CO 2 EOR has been employed since 1974. The work presented here estimates CO 2 storage efficiency factors in CO 2 EOR operations using a unique industry database of CO 2 EOR sites and 12 different reservoir simulation models. The simulation models encompass fluvial clastic and shallow shelf carbonate depositional environments for reservoir depths of 1219 and 2438 m (4000 and 8000 feet) and 7.6-, 20-, and 64-m (25-, 66-, and 209-foot)-thick pay zones. A novel statistical modeling technique incorporating the Michaelis–Menten function is used to generate empirical percentile estimates of CO 2 storage efficiency factors. West Texas San Andres dolomite water alternating gas (WAG) CO 2 flood performance data were used to derive P10, P50, and P90 CO 2 storage efficiency factors of 0.76, 1.28, and 1.74 Mscf/STB (stock tank barrel) of original oil in place. Median CO 2 storage efficiency factors from continuous CO 2 injection following conventional waterflood varied from 15% to 61% and 8% to 40% for fluvial clastic and shallow shelf carbonate simulation models, respectively, while those from WAG injection varied from 14% to 42% and 8% to 31%, respectively. Variation in the CO 2 storage efficiency factors was largely attributable to reservoir depth (a surrogate for reservoir pressure and temperature) and lithology (clastic versus carbonate). The results of this work provide practical information that can be used to quantify CO 2 storage resource estimates in oil reservoirs during CO 2 EOR operations (as opposed to storage following depletion) and the uncertainty associated with those estimates. [ABSTRACT FROM AUTHOR]

• Developed a data assimilation workflow to predict CO2 plume migration in GCS. • Deep learning-based surrogate models are used to accelerate the workflow. • An ES-MDA framework is adopted for inversion and uncertainty quantification. • Complete history matching and forecasting in less than one hour on a workstation. • Intelligent treatments are applied to make the workflow relevant and practical. Fast assimilation of monitoring data to update forecasts of pressure buildup and carbon dioxide (CO 2) plume migration under geologic uncertainties is a challenging problem in geologic carbon storage. The high computational cost of data assimilation with a high-dimensional parameter space impedes fast decision-making for commercial-scale reservoir management. We propose to leverage physical understandings of porous medium flow behavior with deep learning techniques to develop a fast data assimilation-reservoir response forecasting workflow. Applying an Ensemble Smoother Multiple Data Assimilation (ES-MDA) framework, the workflow updates geologic properties and predicts reservoir performance with quantified uncertainty from pressure history and CO 2 plumes interpreted through seismic inversion. As the most computationally expensive component in such a workflow is reservoir simulation, we developed surrogate models to predict dynamic pressure and CO 2 plume extents under multi-well injection. The surrogate models employ deep convolutional neural networks, specifically, a wide residual network and a residual U-Net. The workflow is validated against a flat three-dimensional reservoir model representative of a clastic shelf depositional environment. Intelligent treatments are applied to bridge between quantities in a true-3D reservoir model and those in a single-layer reservoir model. The workflow can complete history matching and reservoir forecasting with uncertainty quantification in less than one hour on a mainstream personal workstation. [ABSTRACT FROM AUTHOR]

• Methodology for carbon dioxide storage resource estimation in residual oil zones. • National and regional scale prospective CO 2 resource estimation. • CO 2 storage in residual oil zones will predominately take place in the free phase. Residual oil zones (ROZs) are increasingly gaining interest as potential reservoirs for carbon dioxide (CO 2) storage. Here, we present a national- and regional-scale methodology for estimating prospective CO 2 storage resources in residual oil zones. This methodology uses a volumetric equation that accounts for CO 2 storage as a free phase in pore space and as a dissolved phase in oil and does not assume any oil production associated with CO 2 storage. Reservoir modeling and the CO 2 -SCREEN tool are used to demonstrate that CO 2 storage in residual oil zones will predominantly take place in the free phase (approximately 92–97%) with some storage as dissolution in oil (approximately 3–8 %). Based on this preliminary demonstration, the CO 2 storage efficiency for ROZs using this national- and regional-scale method ranges from 0.61 to 7.1 %. This range indicates ROZs have a similar efficiency potential for storing CO 2 as deep saline formations (0.51–5.4 %). [ABSTRACT FROM AUTHOR]