TODAY’S STUDY: A FRACKING STUDY SAYS NO WATER POLLUTION
An Evaluation of Fracture Growth and Gas/Fluid Migration as Horizontal Marcellus Shale Gas Wells are Hydraulically Fractured in Greene County, Pennsylvania
15 September 2014 (National Energy Technology Laboratory)
This field study monitored the induced fracturing of six horizontal Marcellus Shale gas wells in Greene County, Pennsylvania. The study had two research objectives: 1) to determine the maximum height of fractures created by hydraulic fracturing at this location; and 2) to determine if natural gas or fluids from the hydraulically fractured Marcellus Shale had migrated 3,800 ft upward to an overlying Upper Devonian/Lower Mississippian gas field during or after hydraulic fracturing.
The Tully Limestone occurs about 280 ft above the Marcellus Shale at this location and is considered to be a barrier to upward fracture growth when intact. Microseismic monitoring using vertical geophone arrays located 10,288 microseismic events during hydraulic fracturing; about 40% of the events were above the Tully Limestone, but all events were at least 2,000 ft below producing zones in the overlying Upper Devonian/Lower Mississippian gas field, and more than 5,000 ft below drinking water aquifers.
Monitoring for evidence of fluid and gas migration was performed during and after the hydraulic fracturing of six horizontal Marcellus Shale gas wells. This monitoring program included: 1) gas pressure and production histories of three Upper Devonian/Lower Mississippian wells; 2) chemical and isotopic analysis of the gas produced from seven Upper Devonian/Lower Mississippian wells; 3) chemical and isotopic analysis of water produced from five Upper Devonian/Lower Mississippian wells; and 4) monitoring for perfluorocarbon tracers in gas produced from two Upper Devonian/Lower Mississippian wells.
Gas production and pressure histories from three Upper Devonian/Lower Mississippian gas wells that directly overlie stimulated, horizontal Marcellus Shale gas wells recorded no production or pressure increase in the 12-month period after hydraulic fracturing. An increase would imply communication with the over-pressured Marcellus Formation below.
Sampling to detect possible migration of fluid and gas from the underlying hydraulically fractured Marcellus Shale gas wells commenced 2 months prior to hydraulic fracturing to establish background conditions. Analyses have been completed for gas samples collected up to 8 months after hydraulic fracturing and for produced water samples collected up to 5 months after hydraulic fracturing. Samples of gas and produced water continue to be collected monthly (produced water) and bimonthly (gas) from seven Upper Devonian/Lower Mississippian gas wells.
Current findings are: 1) no evidence of gas migration from the Marcellus Shale; and 2) no evidence of brine migration from the Marcellus Shale.
Four perfluorocarbon tracers were injected with hydraulic fracturing fluids into 10 stages of a 14-stage, horizontal Marcellus Shale gas well during stimulation. Gas samples collected from two Upper Devonian/Lower Mississippian wells that directly overlie the tracer injection well were analyzed for presence of the tracer. No tracer was found in 17 gas samples taken from each of the two wells during the 2-month period after completion of the hydraulic fracturing.
Conclusions of this study are: 1) the impact of hydraulic fracturing on the rock mass did not extend to the Upper Devonian/Lower Mississippian gas field; and 2) there has been no detectable migration of gas or aqueous fluids to the Upper Devonian/Lower Mississippian gas field during the monitored period after hydraulic fracturing…
Microseismic monitoring results indicate that stress imposed on rock formations by hydraulic fracturing did not extend to the Upper Devonian/Lower Mississippian gas field. However, numerous microseismic events were observed above the Tully Limestone, which is thought to be an upper barrier to fracture growth from hydraulic fracturing in the Marcellus Shale when intact.
The geometry and orientation of microseismic event clusters located above the Tully Limestone suggests that energy from hydraulic fracturing was focused along pre-existing joints, low-offset faults, and bedding planes.
Seven wells in an Upper Devonian/Lower Mississippian gas field were monitored for evidence that hydraulic fracturing in the underlying Marcellus Shale had breached 3,800 ft of intervening strata and allowed gas and fluid migration between the two reservoirs. Monitored parameters included: 1) production and pressure history, 2) isotopic composition of gas, 3) Sr isotope ratio in produced water, and 4) presence of PFC tracer. Background conditions were established by sampling gas and produced water 1 month prior to hydraulic fracturing. Monitoring frequency, monitoring duration, and current status of sample analysis are summarized in Table 3 below.
A comparison of production and pressure records from Upper Devonian/Lower Mississippian wells collected weekly for 1 year after hydraulic fracturing with records collected for a 3-year period before hydraulic fracturing did not reveal pressure or production increases that might indicate migration of gas from the over-pressured Marcellus Shale below. Isotopic analysis of gas and produced water from the Upper Devonian/Lower Mississippian wells did not detect the presence of gas or fluids from the Marcellus Shale. These lines of evidence indicate that there has been no detectable migration of gas and fluids from the Marcellus Shale to the overlying Upper Devonian/Lower Mississippian gas field such as could be provided by open fractures or unplugged wells.
PFC tracers were employed to augment pressure/production and isotope monitoring by detecting the low-volume migration of gas between the hydraulically-fractured formation and the monitored zone in the overlying Upper Devonian/Lower Mississippian gas field. Because PFC tracers are man-made and do not occur naturally, PFC detection in gas produced from Upper Devonian/Lower Mississippian wells would be strong evidence that inter-formational gas migration has occurred. For 2 months after hydraulic fracturing and tracer injection, the gas produced from two wells (UD-2 and UD-5) in the monitored zone was sampled frequently (every 2–6 days) and analyzed for PFC tracers. No PFC tracers were detected (detection limit = 0.001 ppb) in gas samples collected from UD-2 and UD-5 during the 2-month period after hydraulic fracturing.