Both gas permeability and in situ closure stresses computed from log analysis are confirmed by pre-fracture well tests and analysis of fracture stimulation operations on the No. 3 Deep Seam 30-4 in Mesa County. Colorado, one of three wells in a Red Mountain tight gas sands project funded by the Gas Research Institute. The log analysis provides porosity, water saturation, permeability, moduli, and in situ closure stress from an integration of open hole and cased hole geophysical well logs. Log analysis in the completed interval indicates an average gas permeability of 44 microdarcies as compared to 31 microdarcies from subsequent well test analysis. The in situ stress profile from log analysis indicates only moderately weak barriers above and below the completed zone, and this was later confirmed by the fracture modeling performed using the pressure versus time results of the well performed using the pressure versus time results of the well stimulation, which indicated no apparent constraints on vertical growth.
Permeabilities and in situ stresses needed for reservoir characterization and for hydraulic fracture design can be determined from the analysis of a combination of available open hole and appropriate cased hole logs. The available open hole data included resistivity, gamma ray, density, and compensated neutron logs. Acoustic velocities (Vp and Vs) are also needed for moduli and stress calculations, and cased hole full wave acoustic logs were recommended to provide these data. Prior to the full wave acoustic logs, provide these data. Prior to the full wave acoustic logs, cement bond logs were run on all three wells in the Red Mountain project to determine if good cement bond existed across the zones of interest. Presence of a microannulus, which was found in one well, causes strong casing acoustic signals that mask the compressional signals. in this well, good full wave acoustic log data were obtained by logging with the casing pressured up to 1500 psi. To obtain the necessary accuracy for pre-fracture analysis, quality control is a must.
Three Red Mountain wells were drilled in the Piceance Basin in Mesa County, Colorado (Figure 1) as part of a program funded by the Gas Research Institute to develop the program funded by the Gas Research Institute to develop the technology required to produce gas commercially from deeply buried coal. When these experiments were completed, the wells were used to conduct experiments in the shallower Mesaverde sandstones to develop improved fracture stimulation technology as a part of the Gas Research Institute's Tight Gas Sands program. To aid in choosing which well to use in these experiments, an initial formation evaluation was made of the shallower Mesaverde section in all three wells. The No. 3 Deep Seam was chosen because of its central location (Figure 2) and because of a potentially gas productive sand. In addition, this well has relatively thick shaly intervals both above and below the zone of interest that might contain fracture growth.
Another aspect of the initial evaluation was determining what additional logs would be needed in order to aid in designing the fracture stimulation experiments. The situation that existed was similar to that of an old well in which a deeper, primary reservoir zone has been depleted, and shallower reservoirs are workover candidates, but original logging data are incomplete. What was lacking, in this case, was acoustical information needed to calculate moduli and the stresses for fracture design models. The acoustical information needed in these calculations are compressional and shear velocities (Vp and Vs). The full wave acoustic log (FWAL) has been used fairly routinely in open hole logging for determining Vp and Vs, but the quantitative use of full wave acoustic logs run through casing, however, is not nearly so routine.
In order to get the accurate waveforms necessary for Vp and Vs determinations, it is necessary to have an acoustical coupling between the pipe and the cement. For this reason, cement bond logs were run first to determine if sufficient bonding was present to obtain reliable full wave acoustic logs.