Environmental Division Rensselaer Polytechnic Institute Fundamentals of Environmental

Environmental Division Rensselaer Polytechnic Institute Fundamentals of Environmental Systems Engineering Optimal design and operation of a Draft Tube Spouted Bed Reactor for a photocatalytic process David Follansbee, John Paccione, Lealon Martin Tuesday, November 6, 2007 Department of Chemical & Biological Engineering The Martin Research Group

Rensselaer Polytechnic Institute Outline • • • Motivation for process Process Model Parameters and Problem statement Results Conclusion and Future Work Department of Chemical & Biological Engineering The Martin Research Group

Traditional photocatalytic Reactors Rensselaer Polytechnic Institute • Photocatalytic slurry reactors • Batch configuration • Photocatalyst particle separation • Photocatalyst loading limitations • Photocatalytic fixed bed reactors • • Cross sectional area limitations Longer reactor length for increase throughput High pressure drops Mass transfer and kinetics are coupled • Photocatalyst coating of reactor walls • Cross sectional and mass transfer limitations Department of Chemical & Biological Engineering The Martin Research Group

Rensselaer Polytechnic Institute Motivation for DTSMB • Decoupling of mass transfer from kinetics • Continual degradation of contaminant and regeneration of photocatalyst Draft tube UV Clean water outlet • Counter-current design • Photocatalyst immobilized on large, dense particles Dirty Water inlets Jet flow Department of Chemical & Biological Engineering The Martin Research Group

Rensselaer Polytechnic Institute Process block diagram Gfa yo Gp xo . WUV Photo Reactor Gfa yi Gfd εD DA M εA Target Parameters Performance variables G xip Draft tube Packed bed reactor Design Parameters Dt HA Gp xo Department of Chemical & Biological Engineering The Martin Research Group Key design variables Gfd . WPump

Rensselaer Polytechnic Institute Annular bed Model Assumptions: adsorption: 1. Langmuir Counter current contact 2. Constant fluid properties 3. Costant particle size and density Mass load : Gp xi Mass balance: Gp xo GA yo DA M Log mean concentration difference: HA GA Height: Gp xi GA yi yi GA yi A. Y. Khan. Titanium dioxide coated activated carbon: Masters thesis, University of Florida, 2003. V. Manousiouthakis and L. L. Martin. Computers & Chemical Engineering, 28(8): 1237– 1247, July 2004. Department of Chemical & Biological Engineering The Martin Research Group HA H Gp xo

Rensselaer Polytechnic Institute Draft tube model Assumptions • Only non-accelerating portion of bed Gp Mass flowrate of fluid: Gf. D Mass flow rate of particles: Fluid-particle interphase drag coefficient: Dt Ht εD Slip velocity: Gp Pressure Drop Gf. D Z. B. Grbavcic, R. V. Garic, D. V. Vukovic, D. E. Hadzismajlovic, H. Littman, M. H. Morgan, and S. D. Jovanovic. Powder Technology, 72(2): 183– 191, Oct. 1992. Department of Chemical & Biological Engineering The Martin Research Group

UV model (Intensity, Power, and Kinetics) • • • Modeled as a PFR Pseudo first order reaction No mass transfer limitations Rensselaer Polytechnic Institute Gp xo DUV HUV I Io Intensity (Lambert-Beer Law): Mass flow rate: Adsorption coefficient: Rate equation: Power required: Gp xi . WUV Department of Chemical & Biological Engineering The Martin Research Group

Operation limitations and specifications Rensselaer Polytechnic Institute • Mass flowrate can not exceed an upper limit where particles will not settle in annular bed 1. Gp<(1 - mf)Aa pva(max) • Voidage in the draft tube has to be above a critical collapsing voidage and below 1 2. vc< D<1 1. The fluid velocity has to be great enough to ensure transport of particles 1. u 1. 5 vt Department of Chemical & Biological Engineering The Martin Research Group

Rensselaer Polytechnic Institute Test System • • Reactive Red degradation 2 mm catalyst particles Ti. O 2/AC photocatalyst composites Si. O 2 substrate Department of Chemical & Biological Engineering The Martin Research Group

Design Parameters Rensselaer Polytechnic Institute � p 2507 kg/m 3 f 1000 kg/m 3 f 1. 119*10 -3 Ns/m 2 Dt 1 in DA 6 in DUV 2 in Dp 2 mm Ht 2. 5 m HUV 1. 22 m vterminal 0. 257 m/s g 9. 81 m/s 2 At AA AUV Department of Chemical & Biological Engineering The Martin Research Group

Rensselaer Polytechnic Institute Model Constants Umf 0. 0205 m/s mf 1. 74*106 kg/m-4 mf 0. 447 vc 0. 87 -0. 9418 c 1 0. 9984 c 2 -0. 06014 Z. B. Grbavcic, R. V. Garic, D. V. Vukovic, D. E. Hadzismajlovic, H. Littman, M. H. Morgan, and S. D. Jovanovic. Hydrodynamic modeling of vertical liquid solids flow. Powder Technology, 72(2): 183– 191, Oct. 1992. Department of Chemical & Biological Engineering The Martin Research Group

Rensselaer Polytechnic Institute System Parameters k 0. 00833 s-1 C. ﾊM. So, M. ﾊY. Cheng, J. ﾊC. Yu, and P. ﾊK. Wong I 180 W/m 2 C. ﾊM. So, M. ﾊY. Cheng, J. ﾊC. Yu, and P. ﾊK. Wong 300 m-1 M. ﾊNazir, J. ﾊTakasaki, and H. ﾊ Kumazawa KA 602430 ppm-1 A. ﾊY. Khan. Titanium dioxide coated activated carbon xt kgcon/kgpar A. ﾊY. Khan. Titanium dioxide coated activated carbon 0. 272 Kla 0. 00615 s-1 9. 24 $/k. Wh Department of Chemical & Biological Engineering The Martin Research Group

Rensselaer Polytechnic Institute Problem Statement Given: • Adsorptive mass transfer rates • Contaminant degradation rates • The annular flowrate and inlet concentration • Target concentration Minimize yi 10 ppm yo 1 ppm Gf. A 0. 5 GPM Department of Chemical & Biological Engineering The Martin Research Group

Rensselaer Polytechnic Institute Schematic of Algorithm Physical Properties Design Parameters Sensitivity Analysis Operation specs Sensitivity Analysis Interval analysis Minimizing objective function Math Model Department of Chemical & Biological Engineering The Martin Research Group Optimal design and operating conditions

Rensselaer Polytechnic Institute Results Department of Chemical & Biological Engineering The Martin Research Group

Rensselaer Polytechnic Institute Results cont. Department of Chemical & Biological Engineering The Martin Research Group

Rensselaer Polytechnic Institute Results cont. Department of Chemical & Biological Engineering The Martin Research Group

Optimal Design and Operation Rensselaer Polytechnic Institute HA 52. 65 in UV Gp 0. 06 kg/s Gf 5 -25 GPM D 0. 922 -0. 986 0. 5 -0. 9 $/hr Department of Chemical & Biological Engineering The Martin Research Group

Rensselaer Polytechnic Institute Conclusion • Height of annular bed is insensitive to change in mass flowrate. • Operating at a low mass flowrate (<0. 1 kg/s) allows for the most robust performance. • For the test system of Ti. O 2/AC UV cost is high • Motivates for optimization of catalyst properties i. e. density, UV adsorption, and kinetics • Model must be experimentally validated Specifically the kinetics and mass transfer models Department of Chemical & Biological Engineering The Martin Research Group

Acknowledgements • • • Rensselaer Polytechnic Institute Dr. Howard Littman Dr. Joel Plawsky Dr. David Dziewulski (DOH and SUNY school of Public health) Martin Research Group RPI funding Department of Defense Department of Chemical & Biological Engineering The Martin Research Group

Rensselaer Polytechnic Institute Department of Chemical & Biological Engineering The Martin Research Group