Design, simulation, and construction of a test rig for organic vapors

A blow-down wind tunnel for real-gas applications has been designed, validated by means of dynamic simulation, and then built. The facility is aimed at characterizing an organic vapor stream, representative of the expansion taking place in organic Rankine cycle (ORC) turbines, by independent measurements of pressure, temperature, and velocity. The characterization of such flows and the validation of design tools with experimental data, which are still lacking in the scientific literature, is expected to strongly benefit the performance of future ORC turbines. The investigation of flow fields within industrial ORC turbines has been strongly limited by the unavailability of calibration tunnels for real-gas operating probes, by the limited availability of plants, and by restricted access for instrumentation. As a consequence, it has been decided to design and realize a dedicated facility, in partnership with a major ORC manufacturer. The paper thoroughly discusses the design and the dynamic simulation of the apparatus, presents its final layout, and shows the facility “as built”. A straight-axis planar convergent-divergent nozzle represents the test section for early tests, but the test rig can also accommodate linear blade cascades. The facility implements a blow down operating scheme, due to high fluid density and operating temperature, which prevent continuous operation because of the prohibitive thermal power required. A wide variety of working fluids can be tested, with adjustable operating conditions up to maximum temperature and pressure of 400 C and 50 bar, respectively. Despite the fact that the test rig operation is unsteady, the inlet nozzle pressure can be kept constant by a control valve. In order to estimate the duration of the setup and experimental phase, and to describe the time evolution of the main process variables, the dynamic plant operation, including the control system, has been simulated. Design and simulation have been performed with both lumped-parameter and 1D models, using siloxane MDM and hydrofluorocarbon R245fa as the reference working fluids, described by state-of-the-art thermodynamic models. Calculations show how experiments may last from 12 seconds up to several minutes (depending on the fluid and test pressure), while reaching the experimental conditions requires few hours, consistently with the performance of daily-based experiments. Moreover, the economic constraints have been met by the technical solutions adopted for the plant, allowing the construction of the facility.