Abstract
Summary: During the sequential transportation of hot and cold fluids in long-distance pipelines, convective transport dominates the temperature evolution of the pipe fluid. Accurate numerical simulation provides crucial support for engineering operation and scheduling. Aiming at the severe numerical dissipation and resultant nonphysical broadening of temperature interfaces in the commercial software Stoner Pipeline Simulator (SPS) and the traditional method of characteristics (MOC) with linear interpolation (MOC-Linear), we propose a numerical computational framework that integrates the MOC with locally adaptive piecewise Akima interpolation. The framework discretizes the computational grid based on a characteristic velocity of 2.5 m/s, ensuring the Courant-Friedrichs-Lewy (CFL) number stays strictly below unity to satisfy the numerical stability criterion. Two typical operating conditions are designed for method validation: first, a step fluctuation of flow velocity (1.5 m/s→1.0 m/s) under a constant temperature field, for verifying numerical stability; second, a step change of temperature (45°C→35°C) under a constant flow velocity field, for evaluating antidissipation performance. An additional temperature step condition for large-diameter water transportation is included to verify the general applicability of the method across different fluid media and pipe diameters. The results show that under the flow velocity fluctuation condition, the temperature results from SPS, MOC-Linear, and MOC-Akima are fully consistent with no spurious oscillations, verifying the robustness of the proposed framework for smooth temperature fields. Under the temperature step condition, MOC-Akima effectively suppresses numerical dissipation: After 20 hours of simulation, the interface width is narrowed by approximately 70% compared with the 12.0 km yielded by MOC-Linear and by more than 90% compared with the 38.0 km yielded by SPS. Furthermore, MOC-Akima is insensitive to timestep variations and maintains a sharp temperature interface even with a relatively large timestep, striking a favorable balance between computational cost and accuracy. This numerical framework provides a reliable solution for temperature interface tracking in long-distance sequential transportation pipelines.

This publication has 19 references indexed in Scilit: