Effects Of Reservoir Heterogeneities On Heavy Oil Recovery By Steam-Assisted Gravity Drainage

Abstract

Abstract Previous studies of the SAGD process have been based on homogeneous reservoir models. In this paper, new experiments are described which simulate heterogeneous reservoirs. These include reservoirs with thin sha1e layers and reservoirs containing horizontal layers of different permeabilities. The results show that a short horizontal barrier does not affect the general performance greatly. A long horizontal barrier decreases the production rate but, in some configurations, not nearly as much as might be expected. It is observed that faster production is found when a higher permeability layer is above a lower permeability layer than when the conditions are reversed. Introduction The production of heavy oils by gravity drainage to a horizontal well with the continuous injection of steam above follows from the existing steam stimulation and steam flooding practices. The process has been recognized as steam-assisted gravity drainage (SAGD). Rapid development in this area is being encouraged by improved horizontal well drilling technology which allows more favourable economics. Butler et al.(2,3,4) carried out early experimental and theoretical studies on this area. Their theoretical predictions agreed well with the homogeneous reservoir model experiments. No reservoir is completely homogeneous, and the degrees of heterogeneity can vary significantly, even within the same field. It is necessary that preliminary experiments be performed to study the effects of reservoir heterogeneities. The research work presented here is limited to two types of reservoirs which are considered to be representative and close to the real field conditions:reservoirs with thin shale layers,reservoirs containing horizontal layers of different permeabilities. Laboratory studies were conducted using a visual reservoir model saturated with Cold Lake bitumen. The steam injection pressure employed was 153 kPa. Two different well configurations have been investigated.. The first consisted of a vertical circulation steam injector which was perforated near the top of the formation with bitumen produced at the bottom; the second employed a similar configuration to the first, bur with steam introduced slightly above the production well at the base of the formation. Photographs were taken at specified time intervals and the temperature profile was recorded continuously to provide a means for comparisons and eventually for selecting an optimal well configuration for different reservoir conditions. Experimental The two-dimensional reservoir model is that which has been described by Butler et al.(1) only with minor modifications. The schematic experimental set-up is illustrated in Figure 1. The model was scaled to the field conditions by using the method described by Butler et a1(2). The scaling parameters are shown in Table 1. One hour for the model is equivalent to 1.5 years in the field. The vertical circulating steam injector(1) was employed to initiate the communication between the injection and production well. The injection well location was adjusted vertically to allow for different steam injection locations. Figures 2a and 2b show how this vertical steam injector works. The porous materials used in the reservoir model were 2 mm and 3 mm glass beads. In order to obtain a uniformly random packing of the porous pack, the reservoir model was secured on a vibrating table.