Experimental Validation of CFD Hydrodynamic Models for Catalytic Fast Pyrolysis

Authors: Bruce D. Adkins ^a, Neeti Kapur ^a, Travis Dudley ^a, Stephen Webb ^b and Peter Blaser ^a
a Inaeris Technologies LLC, 13001 Bay Park Road, Pasadena, TX 77507, USA
^b CPFD Software LLC, 10899 Montgomery NE, Suite A, Albuquerque, NM 87111, USA

Source: Adkins, B. D.; Kapur, N.; Dudley, T.; Webb, S.; Blaser, P. Experimental Validation of CFD Hydrodynamic Models for Catalytic Fast Pyrolysis. Powder Technology 2017, 316, 725–739.

Abstract: Hydrodynamics for catalyst fluidization and circulation for Catalytic Fast Pyrolysis (CFP) were investigated using cold flow experiments and used to validate predictions from computational fluid dynamics (CFD) simulations. Proprietary fresh and equilibrium CFP catalyst were tested in lab-scale fixed fluidized bed (FFB) and circulating fluidized bed (CFB) setup. Bed expansion and pressure drop were recorded for fresh catalyst in the fixed fluidized bed setup for a range of gas velocities (0–0.04 m/s) which encompassed minimum fluidization and bubbling regimes. Catalyst circulation in the circulating fluidized bed setup on the other hand, was quantified using mass holdup in the mock reactor, particle size distribution for this holdup and pressure profile for a range of transport gas velocities (0.094–0.561 m/s, 10–60 SLPM) and catalyst feedrates (3, 6, 12 kg/h; G = 0.46, 0.92, 1.84 kg/m²-s) for a single-inlet and single-outlet mock reactor configuration. Flow patterns and gas jet penetration through the mock reactor provided qualitative observations that were used for evaluating CFD models. CFD code based on CPFD method (Barracuda Virtual Reactor®) was used to simulate these gas-solid fluidized beds based on three-dimensional grids. Grid independence was established for both fixed and circulating fluidized bed models such that an average grid volume of 23–29 mm³ was sufficient to effectively predict quantitative and qualitative observables. Several CFD parameter sets were tested to evaluate combinations of drag models, blended acceleration model, particle-wall interactions and collision and stress models. Parker-drag model with blended acceleration and advanced settings for particle-wall interactions, collision and stress parameters is determined to provide the best agreement with experimental results.