300 likes | 525 Views
Simulation of the efficiency of hydrogen recombiners as safety devices. Ernst-Arndt Reinecke, Stephan Kelm, Wilfried Jahn, Christian Jäkel, Hans-Josef Allelein. International Conference on Hydrogen Safety September 12-14, 2011, San Francisco, CA. Overview.
E N D
Simulation of the efficiency of hydrogen recombiners as safety devices Ernst-Arndt Reinecke, Stephan Kelm, Wilfried Jahn, Christian Jäkel, Hans-Josef Allelein International Conference on Hydrogen Safety September 12-14, 2011, San Francisco, CA
Overview Simulation of the efficiency of hydrogen recombiners as safety devices • Passive auto-catalytic recombiner (PAR) • Goal of the numerical study • Scenario investigated • Model approach • Results
Commercial PARs in Nuclear Power Plants Vendors • AECL, Canada • AREVA, France/Germany • NIS, Germany Source: Siempelkamp
Operational boundary conditions NPP containment • large temperature and density gradients • large natural convection loops • large geometry (20,000-70,000 m³, typical length scales 5-50 m) • steam-inertized in early accident phase Typical H2 and FC applications • significant smaller scales • different thermal hydraulic conditions PAR applicability ? NPP H2 & FC GOAL:
Experiment Source: CEA • GARAGE facility at CEA/France • single vehicle private garage (~40 m³) • concentration measurement at ~60 pos. Test 1: He release (~2 g/s) for ~2 min data recently published:Gupta et al., Int J Hydrogen Energy 34 (2009) 5902–5911
Approach Scenario based on GARAGE experiment (CEA) • Simulation of the helium release and distribution scenario and validation against experimental data • Replace helium by hydrogen and verify the calculated distribution • Add a PAR to the scenario and compare mitigated/unmitigated scenario
Coupled Modeling Approach Micro Scale Meso Scale Macro Scale REKO-DIREKT(in-house) ANSYS CFX
Output: T, yi, m natural convection Chimney chemical (catalytical) reaction mass/heat transfer Catalystsection Input: T, yi, p PAR model: REKO-DIREKT
REKO-DIREKT REKO-DIREKT CFX- Outlet gas temperature- Outlet gas composition- Mass flow through PAR REKO-DIREKT CFX T / °C yH2 / Vol.-% CFX REKO-DIREKT- Inlet gas temperature- Inlet gas composition- Pressure
Unmitigated release - setup • Physical Model: • Half Symmetry • RANS equations • Ideal gas equation of state • Isothermal • SST-model incl. buoyancy prod. & dissipation • Injection • He: 240 g (1,99 g/s) • H2: 120 g (0,99 g/s) • Wall functions at inner walls • Vent: Outlet Boundary
Mitigated release - setup • Physical Model: • Half Symmetry • RANS equations • Ideal gas equation of state • SST-model incl. buoyancy prod. & dissipation • Injection • H2: 120 g (0,99 g/s) • Wall functions at inner walls • Fixed Wall Temperature • Temperature dependent properties • No heat radiation • Vent: Outlet Boundary
120 s 240 s 800 s Comparison mitigated/unmitigated scenario
Performance (estimates) Processor • 1 CPU Quadcore I7-860, 2.8 GHz • Open Suse Linux 11.3 • CFX 12.1 Calculation time (1200 s) • unmitigated scenario: ~10 d • mitigated scenario: ~40 d • REKO-DIREKT: ~6 min • more time steps • more gas components (H2+O2+N2+H2O)
Conclusions (1/2) Goal • investigate the applicability of PAR from NPP containment to typical H2&FC application • significant differences in operational boundary conditions • first study based on GARAGE experiment, performed with ANSYS-CFX and REKO-DIREKT Results • H2 injection of 1.5 m³, flammable cloud was removed within 10 minutes • hot exhaust plume promotes the transport of hydrogen rich gas mixture towards the PAR inlet
Conclusions (2/2) Next steps • parameter variation • injection rate, location, and direction • PAR design and number • geometry of the enclosure • consideration of possible PAR ignition scenarios • validation against mitigation experiments in new multi-compartment facility, currently under construction at JÜLICH