Simulation of the Entire Process of an Interbasin Water Transfer Project for Flow Routing

The flow routing process plays a crucial role in underpinning the execution of real-time operations within interbasin water transfer projects (IWTPs). However, the water transfer process within the supplying area is significantly affected by the time lag of water flow over extended distances, which results in a misalignment with the water demand process in the receiving area. Hence, there is an imperative need to investigate the flow routing patterns in long-distance water transfer processes. While MIKE11(2014 version) software and the Muskingum method are proficient in simulating flow routing within a water transfer network, they fall short in addressing issues arising from mixed free-surface-pressure flows in water transfer pipelines. This study enhanced the capabilities of the MIKE11(2014 version) software and the Muskingum method by introducing the Preissmann virtual narrow gap method to tackle the challenge of simulating mixed free-surface-pressure flows, a task unattainable by the model independently. This approach provides a clear elucidation of hydraulic characteristics within the water transfer network, encompassing flow rates and routing times. Furthermore, this is integrated with the Muskingum inverse method to compute the actual water demand process within the supplying area. This methodology is implemented in the context of the Han River to Wei River Diversion Project (HTWDP). The research findings reveal that the routing time for the Qinling water conveyance tunnel, under maximum design flow rate conditions, is 12.78 h, while for the south and north main lines, it stands at 15.85 and 20.15 h, respectively. These results underscore the significance of the time lag effect in long-distance water conveyance. It is noteworthy that the average errors between simulated and calculated values for the south and north main lines in the flow routing process are 0.45 m3/s and 0.51 m3/s, respectively. Compared to not using the Preissmann virtual narrow gap method, these errors are reduced by 59.82% and 70.35%, indicating a significant decrease in the discrepancy between simulated and calculated values through the adoption of the Preissmann virtual narrow gap method. This substantially improves the model’s fitting accuracy. Furthermore, the KGE indices for the flow routing model are all above 0.5, and the overall trend of the reverse flow routing process closely aligns with the simulated process. The relative errors for most time periods are constrained wit