Determination of residual gas saturation and gas-water relative permeability in water-driven gas reservoirs.

Abstract

The research on Determination of Residual Gas Saturation and Gas-Water Relative Permeability in Water-Driven Gas Reservoirs is divided into four stages: literature research, core-flooding experiments, development and application of a new technique for reservoir simulation. Overall, all stages have been completed successfully with several breakthroughs in the areas of Special Core Analysis (SCAL), reservoir engineering and reservoir simulation technology.Initially, a literature research was conducted to survey all available core analysis techniques and their individual characteristics. The survey revealed that there are several core analysis techniques for measuring residual gas saturation (Sgr) and hence, the lack of a commonly agreed method for measuring Sgr. The often-used core analysis techniques are steady-state displacement, co-current imbibition, centrifuge and counter-current imbibition. In this research, all centrifuge tests were performed with a decane-brine system to investigate the possibility of replacing gas with a 'model fluid' to minimise errors due to gas compressibility. Furthermore, Sgr is a function of testing temperature and pressure, types of fluid, wettability, viscosity, flow rate and overburden pressure. Consequently, large uncertainties are associated with measured Sgr and the recoverable gas reserves for water-driven gas reservoirs.Due to the lack of a common method for measuring Sgr, the first important step is to clarify which is the most representative core analysis technique for measuring Sgr. In Stage 2 of the research, core analysis experiments were performed with uniform fluids and ambient temperature. In the core flooding experiments, four different sets of core plugs from various gas reservoirs were selected to cover a wide range of permeability and porosity. Finally, all measured Sgr from the various common core analysis techniques were compared.The evidence suggested that steady-state displacement and co-current imbibition tests are the most representative techniques for reservoir application. Steady-state displacement also yields the complete relative permeability (RP) data but it requires long stabilisation times and is costly.In the third stage, a new technique was successfully developed for determining both Sgr and gas-water RP data. The new method consists of an initial co-current imbibition experiment followed by the newly developed correlation (Mulyadi, Amin and Kennaird correlation). Co-current imbibition is used to measure the end-point data, for example, initial water saturation (Swi) and Sgr. The MAK correlation was developed to extend the co-current imbibition test by generating gas-water relative permeability data. Unlike previous correlations, MAK correlation is unique because it incorporates and exhibits the formation properties, reservoir conditions and fluid properties (for example, permeability, porosity, interfacial tension and gas density) to generate the RP curves. The accuracy and applicability of MAK correlations were investigated with several sets of gas-water RP data measured by steady-state displacement tests for various gas reservoirs in Australia, New Zealand, South-East Asia and U.S.A. The MAK correlation proved superior to previously developed correlations to demonstrate its robustness.The purpose of the final stage was to aggressively pursue the possibility of advancing the application of the new technique beyond special core analysis (SCAL). As MAK correlation is successful in describing gas water RP in a core plug scale, it is possible to extend its application to describe the overall reservoir flow behaviour. This investigation was achieved by implementing MAK correlation into a 3-D reservoir simulator (MoReS) and performing simulations on a producing field.The simulation studies were divided into two categories: pre and post upscaled application.The case studies were performed on two X gas-condensate fields: X1 (post upscaled) and X2 (pre upscaled) fields. Since MAK correlation was developed for gas-water systems, several modifications were required to account for the effect of the additional phase (oil) on gas and water RP in gas-condensate systems. In this case, oil RP data was generated by Corey's equations. Five different case studies were performed to investigate the individual and combination effect of implementing MAK correlation, alternative Swi and Sgr correlations and refining porosity and permeability clustering. Moreover, MAK correlation has proven to be effective as an approximation technique for cell by cell simulation to advance reservoir simulation technology.