High-resolution computed tomography scan dataset of lower Mount Simon Sandstone samples from the Illinois Basin

This dataset encompasses high-resolution computed tomography scans of small samples of the lower Mount Simon Sandstone from the subsurface of the Illinois Basin. Samples were collected as part of various geological carbon storage characterization efforts and publications focusing on the Mount Simon as a storage reservoir, with scanning performed at the National Energy Technology Laboratory. Thirty-seven three-dimensional (3D) volumes at various resolutions are described and presented as a resource that illustrates the pore and grain size distributions, as well as other petrographic characteristics. This high-quality, fine resolution, 3D image dataset of an important carbon storage target rock formation can be utilized by researchers as a training dataset for machine learning algorithms and for further reservoir characterizations.


Value of the Data
• Dataset represents a large assemblage of high-resolution digital rock data on the Illinois Basin Mount Simon Sandstone, a high-priority target of multiple ongoing geologic carbon sequestration effort s in the United St ates of America.• High resolution 3D Computed Tomography images are a source of information on porosity and permeability in a format suitable for a variety of research or modeling needs.• Dataset can be a source of other advanced petrographic and diagenetic characteristic classifications, such as mineral composition, sediment maturity, and degree of cementation.• This is a valuable resource to researchers interested in characterizing CO 2 storage reservoirs, testing petrophysical relationships, and improving the information used to generate reservoir models for geological carbon storage.• The range of image resolutions provides a training ground for machine learning based analysis to help improve such algorithms.

Background
The data gathering effort [ 1 ] was spurred by a desire to characterize and understand reservoir properties and heterogeneity of the Mount Simon Sandstone at a pore-to-core scale, allowing for the upscaling of acquired data and their application to regional field trends and relationships.The samples were sourced from a variety of stratigraphic characterization wells within the Illinois Basin that penetrated the Mount Simon Sandstone, including from the basal portion of the formation.They were then scanned with varying acquisition parameters at the National Energy Technology Laboratory, in Morgantown, West Virginia.The computed tomography scanning and core flow facility was built to non-destructively evaluate rock properties from millimeter to micron resolutions [ 2 ].The equipment in this facility has been used to investigate a variety of fluid interactions in the Mount Simon Sandstone, including the trapping of CO 2 droplets in pore spaces [ 3 ] and geomechanical alterations due to the flow of carbonated brines [ 4 ].

Data Description
The basal portion of the Mount Simon Sandstone represents some of the oldest (Cambrian) strata in the Illinois Basin and serves as a potential injection zone for multiple geological carbon storage projects [ 5 , 6 ].The Mount Simon is a fine to coarse-grained, poorly sorted, quartz-rich sandstone with varying arkosic regions.Subdivided into lower, middle, and upper sections, the Mount Simon shows transgressive terrestrial to shallow marine sequences [ 7 ].The Mount Simon is regionally extensive, greater than 457 m thick in portions, is overlain by low permeability sealing formations, and is over 1,524 m deep, making it a viable geologic carbon storage reservoir in the Midwest of the United States [ 8 ].
The Mount Simon samples scanned and described ( Fig. 1 ) here are primarily from Verification Well #1 (39 °52 47.23 N, 88 °53 36.21W) and Verification Well #2 (39 °53 32.07 N, 88 °53 32.82 W) ( Fig. 2 ) drilled as part of the Illinois Basin Decatur Project led by the Illinois State Geological Survey with numerous partners and partially funded by the U.S. Department of Energy.These verification wells were added to this early geological carbon storage site to improve the understanding of plume migration and post-injection site behavior, as described in the NETL Carbon Sequestration Atlas ( https://netl.doe.gov/coal/carbon-storage/atlas ).Portions of these cores were made available to researchers investigating CO 2 rock interactions at this location through the Energy Frontier Research Consortium GSCO2, which was in operation from 2014 to 2018 ( https://science.osti.gov/bes/efrc/Old-Centers/GSCO2).
The dataset consists of 37 reconstructed computed tomography scans ( Fig. 3 ) and associated metadata and is housed on NETL's Energy Data eXchange (EDX) [ 1 ].All data are organized by   brine present in pore space.The experimental condition for each of the tests can be found in Table 1 .
For samples scanned in a dry or a dry epoxy-stabilized state, the sample was placed in the Zeiss Versa XRM-400 CT scanner for scanning at ambient pressure and temperature conditions.
Samples scanned with fluids present were placed in a rubber confining sleeve with spacers at each end.Then the samples were secured inside a custom designed, computed tomographytransparent beryllium core holder.Three Teledyne ISCO pumps were used in the core flooding system maintaining pore and confining pressure [ 3 ].Samples were then injected with brine or CO 2 -saturated brine (denoted respectively as Brine and CO 2 Sat Brine in Table 1 ).For samples with supercritical CO 2 pressure and temperature conditions were elevated to achieve supercritical conditions (14.5 MPa confining pressure, and 47 °C).Trace brine is visible in one sample scanned at ambient conditions (2-Mt.Simon_6986ft_10x).
All Micro-CT scans were conducted using one of three optic lenses: the 10x, the 4x, or the M70, depending on desired resolution.The 10x optic allows for the highest resolution scans and was used to obtain scans in the 0.7-1.1 μm resolution range.The 4x optic was used for scans in the 1.0-2.8μm resolution range, and the M70 for scans with resolutions of 4.0-6.6 μm.For detailed information on which optic and what resolution was achieved for each scan, please refer to Table 1 .
All micro-CT scans were reconstructed using proprietary Xradia software and exported as 16-bit .tiffimage stacks.All image processing, consisting of cropping and scale application, was performed in ImageJ/Fiji [ 11 ].

Limitations
This data curation effort was undertaken with respect to samples analyzed over a period of several years.The prolonged duration nature of the data acquisition meant some parameters could no longer be verified.Specifically, the exact number of small samples could not be established, as many physical samples have already been returned to their research entities of origin.While multiple scans of the same well and depth exist, it was not always possible to establish whether these encompass various parts of the same physical sample, or whether they relate to multiple physical subcores obtained from the same larger core and correspond to the same depth.All available sample information was included in the naming scheme and data table, including the well and depth from which samples were sourced, the scan type (IndCT, M70, 4x, 10x) and the area scanned when one physical sample was confirmed to have been scanned repeatedly (e.g., AreaA, AreaB).
Additionally, due to the proprietary nature of some experiments, three scan groups do not have a physical well location attached to them.In those cases, known parameters include only the generalized source within the Illinois Basin, and the depth from which samples were obtained.

Fig. 1 .
Fig. 1.Segmentation: example of pore space and matrix grain isolation in a 2.1 mm diameter digital representation of scan 2-Mt.Simon_6986ft_10x.Pore space is shown in blue, and matrix grains in yellow, with original computed tomography data seen in the background in grayscale.

Table 1
Compilation of all scans and folders (Scan Group) along with file and scan parameter information.