Abstract: The accurate prediction and fine evaluation of operating costs are important means for oil enterprises to promote lean management and cost control and reduction. Taking deep and ultra-deep oil and gas fields in Western China as examples, two operating cost prediction methods have been proposed, i.e., cost component-based, and principal component model-based prediction methods, which provides robust evaluation tools and decision-making foundation for development plan design, financial budgeting, production and operation optimization, and oil and gas development strategy formulation. The study results confirm that both methods exhibit strong practicability and reliability. The cost component-based operating cost prediction method starts from the components of operating costs, focuses on cost quota, and selects common modes, common alternative modes, or deep modes for operating cost prediction based on specific situations. Two guarantee mechanisms ensuring the rational prediction results rely on the benchmarking-based cost quota auxiliary decision-making method and the understanding on the law of operating cost variation with burial depth. In addition, the rational selection and accurate definition of quotas are the key to the effective application of this method. When historical quotas are involved, a three-year historical average value is generally recommended. The principal component model-based prediction method starts with technical and economic indicators influencing operating costs, covering geological, development, and operation factors, and conducts macro-level operating cost prediction through multi-factor dimensionality reduction and regression based on historical samples. The accuracy of principal component model depends on the comprehensiveness and representativeness of historical operating cost samples.

Key words: oil and gas field, operating cost, prediction, cost quota, principal component, deep to ultra-deep formation