Well-Test Analysis of Horizontal Wells in Gas/Condensate Reservoirs

Abstract

Summary Gas/condensate reservoirs usually exhibit complex flow behaviors owing to the buildup of condensate banks around the wells when the bottomhole pressure drops below the dewpoint pressure. The formation of this liquid saturation can lead to a severe loss of well productivity and, therefore, lower gas recovery. Several studies have examined various ways to minimize the pressure drop in order to reduce liquid dropout and related problems. One solution implemented over the past decade is the use of horizontal wells. There is a lack of published knowledge on the flow behavior of horizontal wells in gas/condensate reservoirs. The limited studies in this area (Muladi and Pinczewski 1999; Dehane et al. 2000; Harisch et al. 2001) focused on well performance rather than on well-test behavior. There has been no evidence of condensate dropout effects in published horizontal-well-test data. This paper presents preliminary results from a study aimed at establishing an understanding of the near-wellbore well-test behavior in horizontal wells in gas/condensate reservoirs, with a focus on the existence of different mobility zones caused by condensate dropout. We used a 3D fully compositional model to develop derivative shapes to be expected from horizontal-well-test data in gas/ condensate reservoirs below the dewpoint under various conditions. We then analyzed actual well-test data that exhibit such derivative characteristics, using a uniform flux horizontal well with wellbore storage and skin model and appropriate reservoir boundaries. The condensate drop effects in the production tests have been accounted for through changes in the values of the total skin effect. Finally, we used a compositional model to verify the results obtained from conventional well-test analysis. It was found that condensate deposit near the wellbore yields a well-test composite behavior similar to what is found in vertical wells, but superimposed on horizontal-well behavior, which makes it much more complex.