Cambrian–Ordovician strata of the midwestern United States are considered a promising reservoir for geologic storage of carbon dioxide. To assess the potential of the Ordovician St. Peter Sandstone, storage-resource estimates were generated using a hierarchical approach to estimating prospective storage resources. The method employs a series of increasingly sophisticated analyses to better facilitate an understanding of the uncertainty in the estimates. Results demonstrate how uncertainty of storage-resource estimates varies as a function of data availability and quality as well as the underlying assumptions used in the application of specific storage efficiency factors.
In the simplest analysis, storage-resource estimates were calculated from updated regional-scale mapping of the gross thickness of the formation and by applying a single best estimate of the mean porosity for the entire formation. This analysis follows the technique prescribed by the US Department of Energy and yields storage-resource estimates ranging from 3.3 to 35.1 billion t CO2 in the Michigan Basin and 1.0 to 11.0 billion t CO2 in the Illinois Basin at the 10% and 90% probability levels. The second analysis incorporated generalized models of the diagenetic history of the formation throughout the two basins by implementing depth-dependent functions of porosity that lead to more realistic portrayals of spatially variable results. Similar resource estimates were calculated for the Michigan Basin, but reduced estimates (43%) were found for the Illinois Basin. The third analysis explicitly accounted for the local-scale spatial variability in reservoir quality using net-porosity calculations, resulting in a significant increase in the low-range resource estimate for the Michigan Basin and dramatic increases for Illinois Basin resource estimates (factor of 3 to 11 increases). A fourth analysis was conducted for the Michigan Basin that used advanced reservoir characterization to define reservoir properties for multiple reservoir facies and yielded resource estimates significantly larger than the third analysis and a larger range of uncertainty. This study highlights how different factors impact the expected uncertainty in storage-resource estimates, and analysis suggests that estimates from the first two approaches provide excessively conservative results, whereas the second two approaches tend to overestimate the resource.