Vol 78 No 1 78105

Abstract

To address the time-scale mismatch between gas network dynamics (minutes to hours) and power grid response (milliseconds) in hybrid ships, which renders steady-state gas network models overly conservative under propulsion load fluctuations and limits gas-fired unit output, this paper proposes a dynamic gas network model via time-frequency transformation and a joint economic low-carbon dispatch framework. First, gas pipeline PDEs are reduced to frequency-domain algebraic equations, forming a dynamic circuit model with gas resistance, inductance, capacitance, and pressure sources. On a 4-node network, the method achieves a computational speedup of 4.6× over FDM (9.89 s) and 3.7× over FEM (7.9 s), with a runtime of 2.14 s (8 iterations) and mean absolute errors of 10⁻⁴, enabling fast dynamic power flow calculation. Second, the ε-constraint method solves the economic-carbon Pareto front, and a gas-electric closed-loop tightening mechanism yields a two-stage “rigid constraint-flexible regulation” strategy. Results show that applying dynamic constraints reduces gas-fired output by 6.7 % while increasing diesel output by 8.2-10.9 %, cutting total cost by 0.4 % at the expense of a 1.65 % emission increase – reflecting the inherent economic-environmental trade-off. Compared with the steady-state model, the dynamic model increases gas-fired generation share by 3.89 %, lowers cost by 0.13%, cuts emissions by 1.59 %, reduces gas-fired power fluctuation by 13.6-19.8 %, and decreases node pressure and flow fluctuations by approximately 72 % and 19-26 %, respectively, significantly enhancing operational stability and safety. The computation time of 55.7 s (5.65× that of the steady-state model) remains acceptable for day-ahead dispatch, and the method’s robustness is confirmed by comprehensive sensitivity analyses. This work provides theoretical and technical support for economic low-carbon operation of hybrid ship gas-electric systems.