2014 GSA Annual Meeting in Vancouver, British Columbia (19–22 October 2014)

Paper No. 20-5
Presentation Time: 9:05 AM

SUPERCRITICAL CO2 EXTRACTION OF HYDROCARBONS FROM THE MARCELLUS SHALE


JARBOE, Palma J.1, CANDELA, Philip A.2, ZHU, Wenlu3 and KAUFMAN, Alan J.3, (1)Department of Interior, U.S. Geological Survey, 12201 Sunrise Valley Drive, MS 956, Reston, VA 20192, (2)Department of Geology, University of Maryland, College Park, MD 20742, (3)University of Maryland, Department of Geology, Geology Building (#237), College Park, MD 20742-4211

Hydraulic fracturing enhances permeability of unconventional shale reservoirs, allowing economically viable quantities of oil and natural gas to be extracted. However, due to variability in both organic and inorganic matrix components found among different shale formations and within a single reservoir, there is a need to develop different fracturing fluid systems that can optimize well productivity while minimizing damage to the reservoir. Carbon dioxide (CO2) may present an extraction alternative to conventional water-based fluids in both water-sensitive and oil-wet formations.

In this study, supercritical CO2 was used to extract n-aliphatic hydrocarbons from samples of Marcellus shale and to evaluate recovery as a function of sample matrix particle size (sieve size). All samples are from vertically drilled wells (subsurface depths ranging from 6,300 - 8,200 feet) in central and western Pennsylvania, USA. Results show that supercritical CO2 has the potential to liberate diesel-range n-aliphatic hydrocarbons from high-maturity shale at estimated in situ pressure and temperature conditions. Total quantity of resolvable n-aliphatic hydrocarbons ranges from approximately 0.3 - 12 milligrams of hydrocarbon extracted per gram of total organic carbon. No significant differences in yield are observed between crushed samples of different sieve sizes (1,000 - 500 µm, 250 - 125 µm, and 63 - 25 µm). However, Brunauer-Emmett-Teller surface area results also reveal that there is no significant variation in exposed sample surface area among sieve sizes as was anticipated. Additionally, scanning electron microscopy images confirm that sieve size is not an accurate indication of crushed rock matrix grain size distribution for the size ranges analyzed. Although no significant variation in either hydrocarbon recovery or specific surface area is found between sieve size fractions for a given sample, overall trends in hydrocarbon recovery are observed among all samples, a finding that warrants further investigation. These observations include: 1) recovery reaches a maximum threshold with the intermediate 250 - 125 µm size fraction; and 2) recovery increases as a function of S1 (free hydrocarbon content) and specific surface area, regardless of sieve size.