Analytical sensitivity analysis of radial natural gas networks

Abstract

Given manifold advantages of natural gas in clean energy substitution and promotion of renewable energy penetration, the scale of natural gas networks (NGN) and the number of natural gas users will continue to increase, resulting in more complex coupling characteristics among system state parameters. To facilitate coupling characteristics analysis, this paper proposes an analytical sensitivity analysis (ASA) method for radial NGN, which establishes analytical expressions between sensitivity, system state, and topology parameters. The steady-state model of NGN is established, considering different equilibrium nodes: reference nodes and slack nodes. The Newton-Raphson method is used to solve this model. Then, analytical expressions, including the sensitivity between gas pressure and reference node pressure, the sensitivity between gas pressure and gas load, and the sensitivity between gas pressure and the ratio of injected flow at slack node and reference node, are derived. The branch's sensitivity contribution factor (SCF), which comprises branch's gas flow and characteristic parameter, is defined. Finally, we conduct a case study based on Belgium NGN to verify the proposed method. The comparison with the perturbation method shows most of the analytical sensitivity results have an error of less than 3%. Results in different scenarios indicate that both the number of reference and slack nodes and their positions will influence sensitivity results. Coupling characteristics in NGN can be adjusted by changing branches' SCF. Two ways to change SCF are installing a gas storage tank (GST) and adjusting branches' characteristic parameters. The optimum installation position of GST and the characteristic parameter having the greatest influence on sensitivity results can be accurately obtained based on ASA, which is verified in the case study.