Well test analysis in tight gas reservoirs

The increasing global energy demand, rising energy prices and declining conventional gas reserves all call for increasing exploitation of gas reserves from unconventional sources, most notably gas trapped in tight formations. This study describes an investigation of some of the key issues related to welltest interpretation in tight gas condensate formations, which are often overlooked or ignored. A single-well radial model was constructed with measured rock and fluid properties using ECLIPSE300. This compositional simulator was used to generate synthetic pressure transient data which were subsequently interpreted in PanSystem welltesting package. The impact of test duration (drawdown, buildup), phases present (gas, condensate), type of fluid system (lean, rich) and reservoir heterogeneity (homogeneous, layered) were investigated. The results indicate that in the absence of wellbore storage effects, a short drawdown period ensures that a short buildup is sufficient to compute reservoir parameters from radial flow regime. For long drawdown periods, the subsequent buildup period can take months to reach infinite-acting radial flow. The time needed to dissipate layering effects (such as a high-perm streak) is frequently excessively long in tight reservoirs. Also, a number of simulations with and without rate-dependent relative permeability curves were performed to investigate the positive impact of velocity (coupling effect) on two-phase skin due to condensate banking. The results reveal that the conventional two-phase pseudopressure function results in more difficult interpretations for obtaining reservoir parameters and is also unable to capture the positive coupling, which is more pronouned than negative inertia especially for richer fluid systems. Finally, the application of the probe radius concept yielded remarkably accurate reservoir pressure profiles versus distance as compared to those obtained by compositional simulations. This data could be used to calculate reasonably accurate rate-dependent relative permeability curves, thus improving long-term well deliverability predictions.