A Facies-Based Approach to Distribution of Petrophysical Properties in an AeolianSandstone Reservoir, Unayzah Formation, Saudi Arabia

摘要

Defining and distributing petrophysical reservoir propertywithin a simulation ready geocellular model is an industrywidedilemma; the result of which has a large business impacton reservoir management decisions. The objective of thisstudy is to develop porosity-to-permeability transforms,multiphase relative permeability and capillary pressure curvesfor each distinct depositional facies. The key to this process isthe ability to recognize each of the four eolian facies anddemonstrate that even at deep formation; the originaldepositional facies controls the petrophysical properties.Through core and image-log studies, the geologist divided thereservoir into the four distinct depositional facies thatcomprise a ”wet” eolian depositional system: dune, sand sheet,paleosol, and playa. Each facies was distributedproportionately by zone (vertically) and by region (area) in a3D geocellular model using an object based modelingtechnique and a wet eolian depositional model. The dunefacies is the best reservoir unit followed by sand sheet;together the dune and sand sheet comprise a high proportionof the reservoir gross thickness. Conventional and special coreanalysis measurements were conducted on hundreds of coresamples from three gas wells. The core data were thoroughlyevaluated for pertophysical properties, and then grouped bydepositional facies. Porosity-to-permeability transforms,multiphase dynamic relative permeability and capillarypressure were derived for each facies. The reservoir propertieswere then geostatistically distributed within each facies objectusing the facies and zone specific transforms.Results showed a quantifiable porosity-to-permeabilitytransform, and corresponding capillary pressure andcondensate-gas relative permeability for each depositionalfacies in the reservoir. This paper shows that by grouping thereservoir into recognizable depositional faceis hassignificantly enhanced the distribution of reservoir and fluidproperties, which were incorporated into a compositionalmodel. This will result in a dependable fluid-flow model thatwill support investment, reservoir development and reservoirmanagement decisions.