Reservoir
Characterization and Geocellular Modeling of Deep-Water Sandstones in the
Ranger Zone of the Long Beach Unit, East Wilmington Field, California
Tye, Robert S.1, Creties
Jenkins1, Don Michel2, Scott Grasse3, Genliang
Guo3, Scott Prior4 (1) DeGolyer and MacNaughton, Dallas,
TX (2) DeGolyer and MacNaughton, (3) Oxy Long Beach Inc, Long Beach, CA (4)
THUMS Long Beach Company, Long Beach, CA
Waterflooding of deep-water sandstones in
the Long Beach Unit (LBU) began in 1965. Over the years, production from these
unconsolidated, Mio-Pliocene reservoirs has been influenced by numerous
technological advances and reservoir analyses. In 2005-06, a reservoir
characterization study of the eastern portion of the most prolific reservoir
zone (the Ranger Zone) was undertaken using a 3D seismic survey, core data, and
logs from more than 900 wells.
The seismic data facilitated structural
horizon mapping and fault delineation, and also holds promise for guiding the
distribution of reservoir properties. Core data allowed depositional
interpretations and the calibration of net sandstone, porosity, and water
saturation calculations from the logs. Petrophysical analysis of the
unconsolidated sediments was complicated by conflicting core data, variable
lithofacies, and heavy minerals.
An existing layering scheme based on
field-wide shale picks (Unit markers) was refined by correlating additional
stratigraphic markers representing unconformities and flooding surfaces. These
were corroborated with injection, production, and pressure data to define
fluid-flow barriers/baffles and delineate flow units.
The seismic, core, and log data were used
to build a geocellular model layered with a combination of constant-thickness
and proportional-thickness cells. Lithofacies, porosity, and permeability data
were distributed using Sequential Indicator Simulation and Sequential Gaussian
Simulation. A height vs. saturation function was used to distribute fluid
saturations. Original oil-in-place volumes were calculated and sensitivities to
oil-water contacts, pore volume, and other parameter
variations were tested. The resulting model will be used in numerical
simulation to understand reservoir mechanisms, define under-injected areas, and
identify bypassed pay.
AAPG Search and Discover Article #90063©2007 AAPG Annual Convention, Long Beach, California