Automated Global Optimization of Commercial SAGD Operations

Abstract The economics of a SAGD operation depend on several factors including the oil rate, cumulative production/well pair (recovery), and cumulative SOR. Complete economic optimization of the process would need to consider all these and possibly additional factors. To date, so far as the authors are aware, economic optimization of a commercial SAGD operation has been carried out by trial and error methods, using many simulations in an attempt to identify the optimum approach to steam use (steam optimization). There is considerable interest in automated ‘goal seeking’ optimization approaches which can largely eliminate the tedious and repetitive work involved in running a large number of simulations to find an optimum case. These approaches seek to find a global optimum of some objective function which is related to the economic criteria for the project. Because of the size, and run time requirements, of commercial SAGD simulation models, ‘brute force’ optimization approaches, where hundreds of runs are made, are neither practical nor elegant. More sophisticated approaches are required. The authors have developed techniques which allow both the automation of the manual approach, and the optimization of this automation. That is to say, the optimization approach used not only finds the global optimum for a given objective function but also finds this global optimum in an optimal fashion. In the work described here, we have chosen to use cumulative SOR as a surrogate for the economic objective function to be optimized. For demonstration purposes the commercial SAGD model used has been restricted to a single well pair. Introduction Steam-Assisted Gravity Drainage (SAGD) has proven itself as a commercial heavy oil and bitumen recovery technology in the Athabasca and Cold Lake regions of Alberta (Komery et al., 1999; Butler, 1997; AED, 2004; Yee and Stroich, 2004; Scott, 2002). SAGD, displayed in cross-section in Figure 1, was developed by Butler (1997) while at Imperial Oil in the late 1970s and consists of two substantially parallel horizontal well bores, one positioned above the other. Typically, the inter-well bore distance is between 5 and 10 m, the wells are between 500 and 1000 m long, and inter-well pair spacing is between 90 and 120 m. (Singhal et al., 1998; Komery et al., 1999). The production well is located a few meters above the base of pay. Steam injected into the top well enters the steam chamber, flows convectively throughout the chamber, and eventually loses its latent heat to the oil sand along the edge of the chamber. As a consequence, the oil's temperature rises, its viscosity falls, and under the action of gravity, it flows either vertically from the top of the chamber or down the edges of the chamber towards the production well located at the bottom of the chamber. To start the process, in common practice, steam is circulated within the injector and producer. In effect, they act as line heat sources which conductively heat the regions around and between the wells. After the oil between the wells is sufficiently warm, the wells are converted over to SAGD mode.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.