The Ketch Field – From Core Description to Development Drilling in 1 Year

摘要

The Ketch gas Field was discovered by Shell/Esso in 1984 with first production in 1999. The field is reservoired in the Carboniferous age Ketch and Cleaver Formations of the Westphalian C/D groups and sealed by the Silverpit shales that overlie the Base Permian Unconformity. In March 2005* the field was acquired by Tullow Oil UK. This presentation outlines results from the field's re-development from core description through geological modelling to successful drilling. The sub-surface work combined a revised geophysical interpretation based on a Pre Stack Depth Migrated seismic dataset with new geological data which were incorporated within Petrel to form a tightly integrated geological model. The geological steps involved core re-description from 7 wells and the breaking out of a new facies classification scheme (C1 (<8' thick), C2 (8-15' thick), C3 (15-50' thick) channels, S1, S2 overbanks, S2F deltas and mud) which were then identified and described in the non-cored wells. Porosity – Permeability data were collated from all core and side-wall core data in the Schooner-Ketch assets and broken down by facies to identify relationships which allowed permeability to be derived. Spatial dimensions for the channels and deltas were derived from analogues such as the Kosi Fan, India, The Kutai Lakes, Kalimantan, Indonesia, Gum Hollow Delta, USA, Pyrenees, Spain and the Carboniferous, Kentucky, USA. In parallel chemo-stratigraphic and palynologic analysis were carried out on cuttings to revise the stratigraphic architecture and delineate the Base Permian unconformity and improve the resolution within the reservoir. The reservoir units are now divided into the Bolton, Upper Ketch formation (A&B), Lower Ketch formation (A-E) based on chemostratigrahy and the Cleaver formation based on marine bands. Chemostratigraphy provides the large scale stratigrahpic framework used for reservoir modeling. At a finer reservoir sub-unit scale we utilise fluvial sequence stratigraphic concepts (Aitkin & Flint, 1995) to identify, interpret, understand and predict reservoir communication and connectivity. The facies model was used to derive porosity, permeability and saturation models which were then upscaled for incorporation into reservoir simulation to history match the field and predict locations of infill wells. The key characteristic of the model design is a strong deterministic control, adopted in preference to relying on the highly stochastic approaches which are often adopted in heterogeneous reservoirs with low net-to-gross. Probabilistic algorithms were required to fill-in small-scale detail in the model, but the architectural connectivity is effectively determined by the chemo- and sequence-stratigraphic work and the sizing of channel bodies from analogues. Particular attention was paid to the upscaling algorithms, to ensure the small-scale architecture was captured implicitly in the simulation, removing the need to run impractically large models. The simulation produced good history matches for almost all wells honoring existing data without having to apply significant 'model edits'. The limited need for model build iteration is attributed to the strong deterministic control on the model design. In addition volumetric gas in place calculations were run to compare with existing material balance models and P/Z plots. Development drilling commenced in June 2006 and two extended reach development (ERD) wells have been completed to date. The first, KA07 which re-drilled into the pre-existing KA02 compartment encountered over 700' net pay in 2400' of section and produced at an initial rate of over 60 mmcfd and has been producing at rates above 45 mmcfd since October 2006. The second KA08 which re-drilled the KA04 compartment targeted an incised valley fill in the UKB and encountered >800' net pay in 1300' drilled. KA09 is planned to target an incised valley fill in the KA01 and KA03 compartment shortly.