Exploitation of Deeply Buried Coalbed Methane Using Different Hydraulic Fracturing Techniques in the Piceance Basin, Colorado and San Juan Basin New Mexico

Abstract

ABSTRACT This paper reports on observations of hydraulic fracturing in coal seams gained in research activities, conducted by Resource Enterprises, Inc. and sponsored by the Gas Research Institute in their Western Cretaceous Coal Seam Project. The fracture treatments span a period from 1986 through 1988 and involve activities in the Piceance and San Juan Basins. Findings resulting from this work are that hydraulic fracture designs must reflect the lithology of the bounding strata as this affects fracture height growth and fracture geometry. Where coal seams are adjacent to sandstone, diagnostics show that hydraulic fracture growth occurs in both the sandstone and the coal and fracturing pressures are less than 1.1 psi/ft. It was also demonstrated that a linear fracture fluid could be used effectively to place proppant in the coal zone if the sandstone overlies the coal. Much higher treating pressures were encountered in fracture treatments where coals were bound by shale. Diagnostics showed the fracture to be contained primarily to the coals and the high treating pressures may have resulted in part from a complex fracture geometry. Several possible reasons for this behavior are listed. High strength casing was found to be useful in accommodating the higher treating pressures and it allowed a larger treatment to be pumped where many jobs have had to be terminated prematurely when the burst strength of the tubulars was reached. Higher injection rates and the use of solid fluid loss material are recommended to create sufficient width in the complex geometry and thereby increase the likelihood of a successful treatment. Caution is advised in the use of conventional interpretations of Nolte-Smith (log net pressure vs. lot time) plots because of the abnormally high treating pressures associated with coal and the potentially-complex fracture geometry that have been observed in mineback experiments are well beyond the scope of the simple Nolte-Smith technique.