The North Dauphin Island Gas Field, located in Alabama State waters, was discovered in 1990 by Arco and has produced 64 BCF of gas during its primary production phase. This field was productive from the Dauphin Sand, a Late Miocene sandstone located at an approximate depth of 1,850 feet TVDSS. The field's reservoir has both high porosity and permeability and it has been produced from five boreholes, two near-vertical and three horizontals. One of the near-vertical wells was sidetracked after the original borehole encountered unexpectedly poor reservoir.
In preparation for converting the North Dauphin Island Field and two adjacent smaller Dauphin Sand fields to the first offshore gas storage facility in the US, all available geologic and geophysical data were re-examined. This storage conversion project will result in up to 21 new deviated injection and withdrawal wells being drilled, in addition to several sub-vertical monitoring wells. The targeted reservoir performance for this facility is 50 Bcf high deliverability with injection and withdrawal rates of 1 Bcfd+. Because of the substantial investment required, a complete understanding of the reservoir is vital. The objective of this study was to utilize all available geological and geophysical data to perform a complete reservoir characterization and simulation to determine the optimum design and placement of horizontal injection/withdrawal boreholes and monitoring wells.
During its primary production phase, a set of parallel 2D seismic lines was acquired over much of the field. This data includes 40 towed dual streamer lines, with approximately 300 ft line spacing. Shallow mud gas detrimentally affected some of the seismic signature making normal reservoir characterization of the Dauphin Sand difficult.
As part of this study the original parallel 2D seismic data were re-interpreted using 3D tools and methods. This was accomplished by first loading 40 individual 2D seismic lines as one 3D survey into a geophysical workstation. A shift in the
CDP numbering on some of the lines was necessary to load them as a 3D survey and still maintain a correct special relationship. 3D tools, such as seismic variance and voxel picking, were applied during the interpretation of this pseudo 3D survey. Additionally, the flattening of progressively deeper stratigraphic units proved to be instrumental in obtaining a clearer image of the reservoir. This enhanced interpretation was then used for a more accurate static model and reservoir simulation.