Field Observations of Gas-Condensate Well Testing

Advisors

Roland N. Horne

Abstract

The deliverability of rich gas wells producing below the dew-point pressure is impacted severely due to condensate banking around the wellbore. Condensate banking also complicates the pressure transient test analysis due to multiphase flow and mixture composition change. The three-zone method to compute the two-phase pseudopressure gives more accurate estimates of reservoir properties than the single-phase pseudopressure method, because it accounts for the composition change in the reservoir. The three flow regions are: inner region 1 where gas and condensate flow simultaneously, middle region 2 where gas and condensate are present but only gas is mobile and outer region 3 where only gas is present.

In this work, three pressure build-up tests for a Middle Eastern gas-condensate well were matched to responses calculated with a compositional simulator. First, the PVT test results of the reservoir fluid were simulated using the Peng-Robinson equation of state. Then, the generated fluid model was included in a radial compositional model that considered the three flow regions. The compositional model was used to match the pressure build-up tests by modifying the reservoir properties. Then, the generated pressure buildup tests were analyzed using three pseudopressure techniques for comparison: the single-phase pseudopressure, the two-phase steady-state pseudopressure and the two-phase three-zone pseudopressure.

Results indicated that all three pseudopressure techniques were capable of estimating permeability accurately. However, variations in estimates of skin factor were observed. The three-zone pseudopressure approach was able to consistently estimate skin factor accurately because it always represented the pressure-saturation relationship around the wellbore correctly. On the other hand, both the single-phase pseudopressure and the two-phase steady-state pseudopressure were unable to estimate skin factor accurately. The skin effect due to liquid dropout caused the skin factor estimations by the single-phase pseudopressure approach to be high. Additionally, the steady-state method overestimated the pressure-saturation relationship which ultimately caused the skin estimates to be low for the steady-state pseudopressure approach.