What is permeability and what is effective permeability? Permeability has a well defined definition in the context of a laboratory coreflood, where it is the transport coefficient in Darcy's Law. What happens in the reservoir, where flow patterns are three dimensional, "cross-sectional areas" are arbitrary, and "pressure drops per unit length" can be defined in many different ways? In this case the definition of effective permeability is, to a large extent, arbitrary, as it depends significantly upon the conceptual experiment which underlies its measurement. This is as true for a direct field determination of permeability as it is when attempting to derive an effective value by an upscaling calculation. This arbitrariness has long been recognised as an issue in the upscaling literature, where the effective permeability may be defined and calculated in a multiplicity of inequivalent ways.
In this paper, effective permeability will be treated as an engineering concept. Instead of attempting to derive a "first principle's" definition of permeability, we must first decide upon the use of the effective parameter, and then choose a method of calculation which is designed for that requirement. In particular, three uses are commonly made of permeability:
to model the continuity of sands,
to model the continuity of barriers, and
to model fluid flow around barriers.
Each of these flow patterns provide different elements of a reservoir characterisation, and require quite distinct calculations.
The superiority of this approach to a conventional single definition of permeability is demonstrated with detailed geologic models of the two turbidite reservoirs of the Magnus field. The challenge of one reservoir is that of locating the oilwithin a mature field, with high risk of rapid water-cut rise for infill targets. The other is low net-to-gross, with the issues of connectivity and large scale pressure support.
In this report we continue the themes introduced in a previous publication [1], in which we examined the application of novel upscaling approaches to the Lower Kimmeridge Clay Formation (LKCF) of the Magnus reservoir. The LKCF is a geologically complex generally low net-to-gross reservoir, consisting mainly of sequences of sandstones and mudstones inter-bedded on the centimeter to meter scale. The field average net-to-gross is less than 25%, but it varies from over 65% in the crest of the field, to essentially zero on the margins. With such a wide variation of net-to-gross, internal communication is clearly expected to be an issue.
The LKCF is in communication with the upper reservoir, the Magnus main sand member (MSM). The MSM is a large Upper Jurassic, sand dominated, turbidite reservoir deposited above a field wide shale [2]. It has been on production since August of 1983. Decline has been managed since January of 1995 with an active infill drilling program [3]. RFT measurements indicate that the LKCF has been partially depleted. There is no indication of compartmentalization, but neither is there any indication of vertical equilibration. On a large scale the effective permeability is non-zero but only to the extent of generating a sequence of tortuous baffles to flow.
