Uncertainty quantification in production forecast for shale gas well using a semi-analytical model

Abstract

Abstract The productivity of shale gas well is often with high uncertainty because of the uncertainties in the characterization of formation properties, fracture properties, gas adsorption, and flow mechanisms. This paper provides an efficient method to probabilistically forecast shale gas production by combining the Markov chain Monte Carlo method (MCMC) and a semi-analytical model. A trilinear flow model is used to predict shale gas production with the consideration of gas desorption and multiple flow mechanisms. The parameters in the model are sampled with the MCMC. A workflow is proposed to predict the gas production and characterize the uncertainties. To make the study results helpful for the field use, a field case from a shale gas field in Southwestern China is applied in the analysis. In this case, we chose ten uncertain parameters to study their effects on eventual ultimate recoveries. Shale gas production is shown to be closely related to the properties of formation, fracture, and flow mechanisms. The fracture half-length and BHP have strong effects on gas production, particularly the production within 5 years. BHP also influences the production after 5 years because of the gas PVT properties and gas adsorption. The results also show that enough iteration number is needed to get a reasonable uncertainty quantification. The sensitivity analysis shows that at least 2000 iterations are required for this case. After that, the probable production could be predicted with a range rather than only one value, and P10, P50, and P90 can be obtained. For the case studied in this paper, there is 90% probability that the EUR for a well is ranging from 0.62 × 108 to 1.48 × 108 m3.

Keywords

• Uncertainty quantification
• Shale gas
• Semi-analytical model
• Production forecast
• Monte Carlo