This reference is for an abstract only. A full paper was not submitted for this conference.

Abstract

Decline in conventional light crudes and current high demand for oil is making exploitation of heavy oil resources an important part of the portfolio of major oil companies. Although the volume of heavy oil resources is vast, of the order of 2.5 - 3 trillion barrels, their development is still very challenging. From a technical perspective, one of the major issues is how to optimally heat the reservoir and thus decrease the high viscosity of the deposits in-situ. In this work, a mathematical model of heating up the heavy oil reservoir by means of a condensable fluid is presented. It is assumed that the condensation mechanism of the injected fluid is first-contact condensation (FCC). The model formulation entails coupling the conservation equations and fluid-property relationships to a temperature-dependent fluid condensation model, resulting in a non-linear partial differential equation (PDE). Although the PDE is solved numerically with the implicit (Crank-Nicolson) finite difference scheme, whose convergence is demonstrated, an analytic steady-state solution is used as a check on the numerical solution. Using steam as the condensable fluid, parametric study is performed, examining the effects of injection-time, rock properties, and amount of condensate on the temperature profile. With the steam injected from reservoir bottom and assuming an insulating overburden and underburden, the results provide upper bounds for temperature distribution in reservoirs exposed to FCC. Furthermore, results of the simulations performed highlight the existence of a distinct constant temperature condensation zone, whose temperature however increases with injection-time. Prior to reaching a steady-state, the condensation zone is always sandwiched between two convection/conduction zones that are characterised by significant temperature gradients. With or without oil production, the trajectory of the condensation front does not conform to the moving-boundary theory. In the examples studied, 50-metre thick Athabasca-type heavy oil deposit initially at 0 oC and exposed to 250 oC steam, a strong correlation is observed between the temperature of the condensation zone and the condensable fraction of injected steam. For instance, after 100 days of heating with steam, the temperature of the condensation zone reaches 50 oC on allowing 20% of the injected mass of steam to condense. Similarly, with 50 and 100% steam condensation, the temperature of the condensation zone reaches 110 and 200 0C respectively. For the case of ‘non-condensable’ steam-like heating fluid, a condensation zone is not discernible, and the instantaneous reservoir temperature is much lower. Although it is clear that the relationship between the temperature of the condensation zone and the amount of condensable steam is not linear, it is also evident that latent heat plays a dominant role in the heating process. When the condensate is allowed to drain, the temperature of the condensation zone remains the same, but the lower bound of the condensation zone moves closer to the injection plane. The position of the lower bound is thus dependent on the rate of condensate drainage. This study, in addition to providing better insights into the dominant reservoir mechanisms during thermal enhanced oil recovery, such as the dynamics of steam-chamber development in steam-assisted gravity drainage processes, the proposed model is useful for screening and preliminary design of thermal flooding projects and field data validation.

This content is only available via PDF.