--> Integrated Thermal Reservoir Characterization: The Grosmont Formation at Saleski, Alberta, Canada

AAPG Annual Convention and Exhibition

Datapages, Inc.Print this page

Integrated Thermal Reservoir Characterization: The Grosmont Formation at Saleski, Alberta, Canada

Abstract

The Upper Devonian (Frasnian) Grosmont Formation in north-central Alberta contains 406 billion barrels of bitumen in-place (ERCB, 2013) hosted in a paleo-karsted dolostone. A thermal pilot has been in operation since 2010 to prove the commercial viability of extracting this resource. The targets are two bitumen-saturated dolostone reservoirs approximately 350 m below the surface: the Grosmont “C” unit, a laterally continuous intensely-fractured vuggy dolostone that resulted from solution enhancement of moldic porosity, and the overlying Grosmont “D” unit, a stratiform breccia that resulted from the dissolution of interbedded evaporites. The pilot employs horizontal wells for steam injection and bitumen recovery and seven vertical and fully instrumented observation wells provide continuous pressure and temperature monitoring. Repeated high-resolution 4D seismic surveys tied to core and petrophysical log data help provide a detailed characterization of the carbonate reservoir. The nominal acquisition geometry of the base 3D seismic survey acquired in 2011 is orthogonal shot and receiver lines laid out with an interval of 80 m and orientated east-west and north-south with 20-m source and receiver intervals, resulting in a natural bin size of 10 × 10 m. A 5-m station indent ensured that sources and receivers were not co-located, and allowed for sub-binning to 5 × 5 m. Single source charges of ¼ kg dynamite were loaded to a depth of 6 m. Subsequent 4D monitors acquired each year repeated this geometry and were processed with identical processing streams, including amplitude normalization, statics and velocities, and prestack time migration to 5 × 5 bins. Monitors were match-filtered in phase and amplitude to the base survey and simultaneously inverted to impedance volumes. The differences between the monitors and the baseline survey likely reflect where injected steam is present at the time of acquisition, and enclose a rock volume that is conformable to the injected steam volumes taking into account the produced liquids. Interestingly, there is very little ‘time sag’ of the underlying horizons including the base of the reservoir in contrast to the sag often observed with thermally heated clastic reservoirs; we interpret this to be the result of the relatively rigid framework of the carbonate reservoir. Importantly, the data reveals uniform thermal conformance along the entire length of the horizontal well bores in the paleo-karsted reservoir.