Cost Estimation Proxy Models for Economic Evaluations in Petroleum Projects

A Case Study from the Onshore Gas Field in the Southern Coastal Tanzania Basin

Abstract

Petroleum economic evaluation involves estimating revenues from forecasted production profiles and field costs including capital expenditures (CAPEX), drilling expenses (DRILLEX), and operating expenses (OPEX). The existing cost-estimating tool requires several inputs making it time-intensive and difficult to use with few data during the early stages of projects. Majority of the previously developed time saving cost estimations proxy models rely on unrealistic assumptions that include uniform operational costs for different fields with a different number of wells, casings, and drilled depths. This work focused at developing proxy models that consider the variability of the development costs with different parameters. The developed models benefited from a three-step approach for CAPEX, DRILLEX, and OPEX estimations based on datasets from three wells from a gas field in southern coastal Tanzania. Firstly, cost sensitivity analysis was performed using QUE$TOR v15.1.0.18, a cost estimating commercial software to determine the most influential field parameters of the field costs. Secondly, the field cost models were generated based on historical cost data from the gas field using multivariable regression analysis with the help of Statistical Package for the Social Sciences software (SPSS) v22.0. Lastly, errors analysis was done for checking the predictive reliability of the models. Based on the analysis, the CAPEX and OPEX were found to be strongly linearly dependent on the size of processing facilities, number of producing wells in the gas field and production capacity, respectively. On the other hand, a nonlinearity relation was revealed on of the DRILLEX which was strongly dependent on drilled well depth and number of installed well casing. Results show that the developed models are useful and their reliability becomes robust when more data is used. A stochastic modeling approach was further recommended for the models to incorporate uncertainties associated with the parameters used to quantify the cost estimates.