An Efficient Multiscale Method for the Simulation of In-situ Conversion Processes

Abstract

Numerical modeling of the in-situ conversion process (ICP) is a challenging endeavor involving thermal multiphase flow, compositional pressure/volume/temperature (PVT) behavior, and chemical reactions that convert solid kerogen into light hydrocarbons, which are tightly coupled to temperature propagation. Our investigations of grid-resolution effects on the accuracy and performance of ICP simulations have demonstrated that ICP-simulation outcomes-specifically, chemical-reaction rates, kerogen-accumulation profiles, and oil-/gas-production rates, may exhibit relatively large errors on coarse grids. Coarse grids are attractive because they deliver favorable computational performance. We have developed a novel multiscale modeling method for simulating ICP that reduces numericalmodeling errors and reproduces fine-scale-simulation results on relatively coarse grids. The method uses a two-scale solution method, in which the reaction kinetics of the solids is solved locally on a fine-scale grid, with interpolated temperatures obtained from coarse-grid simulations of thermal flow and fluid transport.We demonstrate the accuracy and efficiency of our multiscale method with representative 1D models. It is shown that the method delivers accurate solutions for key ICP performance indicators with very little computational overhead compared with corresponding coarse-scale models. The robustness of the multiscale method has been verified over a number of physical-parameter ranges with a limited-scope sensitivity study. Numerical results show that the multiscale method consistently improves the simulation results and matches the fine-scale reference results closely.Governing Equations. The mass-conservation equation for the ICP problem is written for each component by combining the

Adaptive Local-Global Multiscale Simulation of the In-situ Conversion Process

“…One can view the nearwellbore upscaling method reported in Aouizerate et al (2012) as a step in this direction. Li et al (2015a) describe a preliminary effort to apply multiscale methods to ICP modeling.…”

“…The implementation of the MM for ICP is performed at a relatively high level with the built-in scripting language that accompanies the dynamic-modeling package. There are multiple options to improve the method further: For ICP, an obvious extension is to combine this work with the multiscale treatment of solid reactions discussed in a separate paper (Li et al 2015a). For more-general (also nonthermal) applications, one can investigate dynamically generating pseudorelative permeability (and where appropriate) capillary pressure curves, with fine-scale results.…”

Modelling in-situ upgrading of heavy oil using operator splitting method

“…where P is the pressure, K the rock permeability and k r,p and μ p are the relative permeability and viscosity of phase p. The impact of the solid saturation on the permeability is given by a simple heuristic exponential relationship [10]:…”

“…The coefficient a determines how strongly the permeability varies with the change of solid saturation. Typically, a has a value between 5 and 50 [10]. We assume that the thermally unstable chemical entities decompose with first-order kinetics.…”

“…We note that values for the pre-exponential factors and activation energy are smaller than those reported by [25]. The rock properties are shown in Table 4 and the fluid properties in Table 5, adapted from [10,20,21]. The viscosity of the gas phase is given by: These data are adapted from [20] and the viscosity of the liquid phase is given by [26]: …”