A Dynamic Modeling of Pipeline Networks for Dense Compressible Fluids Tuned With Time Series of Plant Data

We approached a phenomenological model for dynamic simulation of large pipeline networks carrying dense compressible fluids. In a typical application, the thermodynamic condition of the fluid is slightly supercritical, temperature is near environmental levels, pressure is high and the network size reaches hundreds of kilometers. Our model relies on decomposing the network on tube and vertex contexts. In the first context, continuity and momentum equations are written for each tube, jointly with a simplified flow relationship between density and pressure. These equations are treated via a finite element formalism. In the second context, network connectivity is taken into account through transient vertex mass balances. The set of tube-vertex time differential equations is integrated numerically. Model parameters are tuned with quasi-stationary time series of data via reconciliation procedures. Our case study is a 485km line for ethylene transportation between petrochemical plants. Results are useful for engineering, training, monitoring and inventorial applications.