Global natural gas demand to 2025: A learning scenario development model

Abstract

Scenario development approaches are designed to deal with chaotic behaviors of complex systems and are widely used in the case of energy-related demand forecasting and policy planning. Building on traditional qualitative scenario models, a novel Learning Scenario Development Model (LSDM), incorporating qualitative and quantitative components, is proposed to generate different scenarios for global natural gas demand to 2025 in order to discover and compare the likely behavior of alternative future natural gas markets. This model, consists of five phases: 1) organize the fundamental data set, 2) investigate a data mining based pre-process procedure to initialize the quantitative dimension of the model, 3) select a set of procedures for forecasting global natural gas demand to 2025, referred to as the mixed model, 4) generate a reference case scenario (business as usual) using the mixed model, and 5) develop alternative scenarios (five in this study) applying a qualitative expert-based process. Unlike other scenario models, the LSDM is equipped with validation procedures that enable decision makers to develop alternative scenarios based on various input strategies to evaluate and simulate them. For the application of global natural gas demand, results suggest a gentle uptrend for the reference case (about 4232 bcm in 2025). The alternative scenarios considered support a continued increase for the global natural gas demand, but at different rates depending on the removal or addition of multiple natural gas suppliers (from 2013 to 2025, the scenarios considered display demand growth varying from 23.5% to 25%).