CARBON STORAGE ESTIMATES IN ORGANIC-RICH SHALES OF THE MARCELLUS AND UTICA–POINT PLEASANT INTERVALS IN THE EASTERN UNITED STATES

Geologic carbon storage is generally investigated relative to saline reservoirs and enhanced oil and gas recovery with ultimate storage in conventional reservoirs, but for regional assessments, unconventional shale plays should also be investigated. The Midwest Regional Carbon Sequestration Partnership is investigating the potential for CO₂ storage and enhanced gas recovery in the Middle Devonian Marcellus Shale and Upper Ordovician Utica Shale and Point Pleasant Formation in the Appalachian Basin. Organic-rich shale assessments require that regional total organic carbon estimates be derived from downhole log data based on the best comparisons to lab data.

Carbon storage in the Marcellus Shale is expected to occur as mostly adsorbed gas associated with organic matter. Storage estimates were made for a play area in the basin with depths of more than 2,500 ft, and net thickness of shale with at least 2 percent TOC and a gamma-ray log value of 180 API units or greater. A comparison of lab-derived TOC values with four different models showed that Schmoker’s density model was the best fit for the lab data—a similar result to previous work in the Ohio Shale. A median (50th percentile) TOC curve was gridded and processed to produce an estimated carbon storage volume ranging from 887 million short tons at 3 percent efficiency to 2,955 million short tons at 10 percent efficiency.

Carbon storage in the Upper Ordovician Utica and Point Pleasant shales is expected to occur as a free gas in the matrix porosity of the units and as adsorbed gas associated with the organic matter. Free gas in the matrix (greater than 98 percent) is more important in these units than adsorbed gas storage because of the shales’ lower overall organic content. Storage estimates were limited to a play area in the basin more than 2,500 ft deep and a cutoff of net thickness of shale with at least 1.5 percent TOC. TOC derived from existing models did not adequately estimate lab-derived TOC in this interval, so previous models were modified and new non-parametric linear regression and multi-variate linear regression models were tested. Data were then gridded and processed to produce an estimated carbon storage volume of 2,072 million to 6,907 million short tons. Nearly 60 percent of the total storage is estimated to occur in the Point Pleasant Formation.