A data-driven approach for the optimization of future two-level hydrogen supply network design with stochastic demand under carbon regulations

Abstract

Reducing emission of conventional vehicle exhaust has many benefits for developing low-carbon cities. The transition from on-road conventional passenger cars to Hydrogen Fuel Cell Vehicles (HFCVs) may enhance air quality and health significantly. Aiming at constructing low-cost and environmentally friendly hydrogen supply networks (HSNs) in urban areas, a basic model that considers hydrogen plant and refueling station decisions is presented for the Two-level Hydrogen Supply Network Design (THSND) with stochastic demand. By coupling the basic model and carbon regulations, four extended models are established to evaluate carbon reduction on total cost of HSNs. A data-driven approach that does not require the full information of target areas is developed to solve the models. It introduces a data acquisition method of Points of Interests (POIs), a clustering-based method to select alternative refueling stations, a trip simulation method to determine demand uncertainty, and the application of solvers. The developed data-driven approach is shown to be robust and effective in terms of applicability and generalizability according to the simulations. Meanwhile, the best suitable decision result on plants and stations is both cost-effective and environmentally successful under the model considering carbon cap and trade. The internal relationship between the models indicates that the hybrid carbon emission policy is superior to naive carbon regulations.