Splitting model and rapid simulation for large-scale natural gas pipeline networks

Numerical discrete methods must typically solve high-dimensional nonlinear partial differential equations when simulating large-scale natural gas pipeline networks. This leads to a sharp increase in computational complexity, resulting in reduced simulation speed. In response to these issues, this study focused on the development of a splitting model and rapid simulation techniques for large-scale pipeline networks. A novel simulation method named linearized lumped parameter model (LLPM) was proposed; by using Lanczos integrate and Taylor expansion methods, the LLPM considers the inertial and gravity terms. Combining with the linearized finite difference method, a simulation technique was constructed, in which the pipeline network is split into an independent node model and pipeline models. Additionally, the boundary conditions of each pipeline are decoupled from the network. This method not only accelerates the simulation efficiency by breaking down the high-dimensional network model into low-dimensional node and pipeline models, but also provides detailed parameter profiles along the pipeline, thus overcoming the limitations of traditional lumped parameter methods. Finally, to validate the proposed method, an actual offshore pipeline network with a total length of 930.36 km was simulated. The results reveal that, compared with the measurement data, the simulation error of the proposed method is 0.92%, and the calculation speed is 132.16 times faster than that of the discrete method. The findings of this study provide a valuable reference for the fast and accurate simulation of large-scale natural gas pipeline networks.