The success of petroleum development projects depends to a large degree on well construction costs. Well construction cost estimates often contain high levels of uncertainty. In many cases, these costs are estimated using deterministic methods that do not reliably account for uncertainty, leading to biased estimates. The objectives of this work were to improve the reliability of deterministic well construction cost estimates by incorporating external probabilistic methods into the estimation process.
The method uses historical well cost estimates and actual costs to develop probabilistic correction factors that can be applied to future well cost estimates. These factors can be applied to the entire well cost or to individual cost components. Application of the methodology to estimation of well construction costs for horizontal Barnett shale wells in the Fort Worth basin resulted in estimated well cost estimates that were well-calibrated probabilistically. Overall, average estimated well cost using this methodology was significantly more accurate than average estimated well cost using deterministic methods.
The methodology is intended for use by individual engineers to improve their well construction cost estimation and by management to externally revise and improve well cost estimates for budgeting purposes. Systematic use of this methodology will provide for more accurate and efficient allocation of capital for drilling campaigns, which could have significant impacts on reservoir development and profitability.