Fitting Foam-Simulation-Model Parameters to Data: II. Surfactant-Alternating-Gas Foam Applications

Abstract

Summary Foam improves sweep in miscible and immiscible gas-injection enhanced-oil-recovery processes. Surfactant-alternating-gas (SAG) foam processes offer many advantages over coinjection of foam for both operational and sweep-efficiency reasons. The success of a foam SAG process depends on foam behavior at very low injected-water fraction (high foam quality). This means that fitting data to a typical scan of foam behavior as a function of foam quality can miss conditions essential to the success of an SAG process. The result can be inaccurate scaleup of results to field application. We illustrate how to fit foam-model parameters to steady-state foam data for application to injection of a gas slug in an SAG foam process. Dynamic SAG corefloods can be unreliable for several reasons. These include failure to reach local steady state (because of slow foam generation), the increased effect of dispersion at the core scale, and the capillary end effect. For current foam models, the behavior of foam in SAG depends on three parameters: the mobility of full-strength foam, the capillary pressure or water saturation at which foam collapses, and the parameter governing the abruptness of this collapse. We illustrate the fitting of these model parameters to coreflood data, and the challenges that can arise in the fitting process, with the published foam data of Persoff et al. (1991) and Ma et al. (2013). For illustration, we use the foam model in the widely used STARS (Cheng et al. 2000) simulator. Accurate water-saturation data are essential to making a reliable fit to the data. Model fits to a given experiment may result in inaccurate extrapolation to mobility at the wellbore and, therefore, inaccurate predicted injectivity: for instance, a model fit in which foam does not collapse even at extremely large capillary pressure at the wellbore. We show how the insights of fractional-flow theory can guide the model-fitting process and give quick estimates of foam-propagation rate, mobility, and injectivity at the field scale.