Summary
The ability to predict well deliverability is a key issue for the development of gas/condensate reservoirs. We show in this paper that well deliverability depends mainly on the gas relative permeabilities at both the endpoint and the near-wellbore saturations, as well as on the reservoir permeability. We then demonstrate how these parameters and the base capillary number can be obtained from pressure-buildup data by using single-phase and two-phase pseudopressures simultaneously. These parameters can in turn be used to estimate gas relative permeability curves. Finally, we illustrate this approach with both simulated pressure-buildup data and an actual field case.
Introduction and Background
In gas/condensate reservoirs, a condensate bank forms around the wellbore when the bottomhole pressure (BHP) falls below the dewpoint pressure. This creates three different saturation zones around the well. Close to the wellbore, high condensate saturation reduces the effective permeability to gas, resulting in severe well productivity decline (Kniazeff and Nvaille 1965; Afidick et al. 1994; Lee and Chaverra 1998; Jutila et al. 2001; Briones et al. 2002). This decline is reduced at high gas rates and/or low capillary forces, which lower condensate saturation in the immediate vicinity of the wellbore, resulting in a corresponding increase in the gas relative permeability. This is called the capillary-number effect, positive coupling, viscous stripping, or velocity stripping (Boom et al. 1995; Henderson et al. 1998, 2000a; Ali et al. 1997a; Blom et al. 1997). High gas rates, on the other hand, induce inertia (also referred to as turbulent or non-Darcy flow effects), which reduces productivity. Well productivity is thus a balance between capillary number and inertia effects (Boom et al. 1995; Henderson et al. 1998, 2000a; Ali et al. 1997a, 1997b; Blom et al. 1997; Mott et al. 2000.).
Well-deliverability forecasts for gas/condensate wells are usually performed with the help of numerical compositional simulators. Compositional simulation requires fine gridding to model the formation of the condensate bank with the required accuracy (Ali et al. 1997a). Non-Darcy flow and capillary-number effects (Mott 2003) are accounted for through empirical correlations, which require inputs such as the base capillary number (i.e., the minimum value required to see capillary-number effects), the reservoir absolute permeability, and the relative permeability curves. These are usually determined experimentally, but laboratory measurements at near-wellbore conditions are very difficult and expensive to obtain. An alternative, as shown in this paper, is to obtain them from well-test data.
Well-test analysis is recognized as a valuable tool for reservoir surveillance and monitoring and provides estimates of a number of parameters required for reservoir characterization, reservoir simulation, and well-productivity forecasting. In gas/condensate reservoirs, when the BHP is below the dewpoint pressure, the effective permeability to gas in the near-wellbore region and at initial liquid saturation can be estimated with single-phase pseudopressures (Al-Hussainy et al. 1966) and a two- or three-region radial composite well-test-interpretation model (Chu and Shank 1993; Gringarten et al. 2000; Daungkaew et al. 2002), whereas the reservoir absolute permeability may be determined with two-phase steady-state pseudopressures (Raghavan et al. 1999; Xu and Lee 1999). In this paper, we show that well-test analysis can provide additional parameters, such as the gas relative permeabilities at both the endpoint and the near-wellbore saturations and the base capillary number. These in turn can be used to generate estimated relative permeability curves for gas.
Evaluation of gas condensate reservoir behavior using velocity dependent relative permeability during the numerical well test analysis
