The paper describes the procedure that was adopted to develop operating guidelines for the cyclic steam stimulation process in a water underlain oil sands reservoir. The study consisted of three parts:
The risks associated with fracturing into water sands were quantified by conducting a stochastic simulation;
Numerical simulations were conducted to obtain correlations among the cumulative production volumes and six selected operating variables;
These correlations were subsequently embodied in the economic model to maximize the profits under different constraints.
The study concluded that rather than a unique value, a wide range of operating conditions maximized the project profitability. It recommended that a cautious approach in exploitation of these resources was warranted since the penalty for selecting non-optimal conditions was not high.
BP Canada Resources Limited, in partnership with Petro Canada Inc. is operating a cyclic steam project in the oil sands deposits of Cold Lake, Alberta. The deposits occur in several different units and some of the rich sands are underlain by a thick water leg. These sands were excluded from the initial development of the project because of uncertainty and the need for piloting. However, with falling oil prices, the exploitation of these highly oil rich sands became essential since these were expected to yield the highest returns. As a consequence, one satellite of twenty two wells was drilled in 1986 to exploit these sands. The operation, although considered as experimental, was expected to produce bitumen at commercial rates. This required that the operating strategy should not only maximize the cash flow but also be robust to the risk of rupturing into the water leg and plant imposed constraint of heat recovery capacity.
The operating strategy for the original satellites was developed by using numerical simulation studies in conjunction with an economic model. The method was incapable of determining the optimal conditions under a set of constraints. The strategy represented a local rather than a global maximum because only a unidimensional search in the variable space was carried out.
The optimization procedures were modified in the present study to overcome the above stated shortcomings.
The study contains three parts. The first part deals with quantification of expected loss in production due to fracturing into the water sand. It describes the details of the Monte Carlo simulation study which was conducted to determine the probability distribution of the production losses associated with cyclic steam stimulation of the water-underlain reservoir. The second part contains the description of numerical simulation studies which were carried out to develop relationships among the production performance indicators and the operating variables. Statistical techniques were used to reduce the number of dependent variables. The development of an economic model and the optimization procedures used to obtain the optimal strategy are described in the last part.
The Wolf Lake Project exploits the Cold Lake oil sands deposits in north-western Alberta (see Figure 1). The project is designed to produce 1300 m3/d of bitumen from the Clearwater formation of the Manville Group. The present study was aimed at developing the operating guidelines for the production maintenance satellite drilled in the Lower Grand Rapids member which lies above the Clearwater Formation.
P. 197^
