A Probabilistic Analysis of the Economics of Oil Recovery From the Athabasca Tar Sands

Abstract

ABSTRACT A general-purpose probabilistic economic evaluation model for petroleum investment ventures has been developed. The model’s operation is demonstrated by considering an actual field in Alberta, where an economic evaluation of the feasibility of producing the Athabasca tar sands is conducted. An oil recovery method (COFCAW) is applied, and necessary input data are selected. Other oil recovery methods may be evaluated by modifying the input data. The model allows the use of five continuous input probability distributions as well as a discrete input distribution. Input for the continuous distributions is entered as a three point estimate (low, medium, and high), whereas input for the discrete distribution is entered as a number of values and their respective cumulative probabilities of occurrence. The maximum number of values allowed is nine. The model calculates possible production rates which are directly dependent on the calculated recoverable reserves. The production rates are used to calculate total revenue; this is followed by calculating direct and indirect cost items, taxable income, taxes, earnings after tax, non tax deductible items, cash outflows, and net cash flows. These net cash flows are used to calculate three profitability measures: the discounted cash flow rate of return, a modified rate of return, and the net present value. The results obtained are summarized as follows: the discounted cash flow rate of return, the modified rate of return, and net present value calculated had average values of 35%, 30%, and $10.5 million, respectively. A sensitivity study on the input variables was conducted to identify the critical variables whose input values have an impact on the results of the model. Among the most critical variables are well operating expenses, inflation rate on the operating expenses, bitumen selling price, and inflation rate on that price. It is also observed that the modified rate of return is the least sensitive to input data variations of all profitability measures considered.