Datapages, Inc.Print this page

Deterministically Guided Geocellular Modeling of Deep-Water Deposits in East Coast of India : A Case Study

Sanjay K. Shrivastava, Amit K. Sinha, Ramanathan Nachiappan, and Jyoti P. Rout
Subsurface Development, Reliance Industries Ltd., Navi Mumbai, India

A giant deep-water gas field discovery is under fast track development in the east coast of India. The reservoir consists of high frequency multiple episodes of channel formation and channel fill. Massive unconsolidated sand and laminated sand mudstone intervals of varying bed thickness are the primary reservoir facies. Drilling and production from these unconsolidated, Upper Pliocene reservoirs have been planned based on reservoir analyses using advance methods and processes. High end seismic inversion has added considerable value in discriminating the reservoir facies, albeit with inherent uncertainty. An attempt has been made to characterize the reservoir and build a 3D geocellular model using conventional 3D seismic, inverted seismic volumes, core calibrated petrophysics aided by detailed seismic mapping. Nine major channel episodes within the reservoir system were mapped. The mapped channel base and top surfaces facilitated stratigraphic framework building for deterministically guiding the stochastic property population. Regions were created to differentiate areas within and outside the channel and used for optimizing grid design and pattern, layering scheme, grid size and property population. The main advantage of modeling properties in such a grid is to ensure that the geostatistical processes are tightly linked with sedimentary processes of channel filling due to adequate parameterization of depositional space. Facies quality distribution was constrained by relative position of depositional element. An improved correlation co-efficient with secondary variable was achieved by this method as against conventional models constrained by broad sequence stratigraphic surfaces. Multiple realizations were created and evaluated to capture uncertainty for unknowns.