Experimental validation and CFD modeling study of biomass fast pyrolysis in fluidized-bed reactors Q. Xue a,⇑ , D. Dalluge b,c , T.J. Heindel a,b , R.O. Fox a , R.C. Brown a,b,c a Department of Chemical and Biological Engineering, Iowa State University, 2114 Sweeney Hall, Ames, IA 50011, USA b Department of Mechanical Engineering, Iowa State University, 2025 Black Engineering Building, Ames, IA 50011, USA c Center for Sustainable and Environmental Technologies, Iowa State University, Ames, IA 50011, USA article info Article history: Received 15 December 2011 Received in revised form 27 February 2012 Accepted 29 February 2012 Available online 15 March 2012 Keywords: Biofuel Computational fluid dynamics (CFD) Multiphase flows Thermochemical conversion Variable particle density abstract In this work, an Euler–Euler multiphase computational fluid dynamics (CFD) model, which couples a bio- mass particle pyrolysis model with a multi-fluid hydrodynamics model for gas–particle flow, is used to describe a biomass pyrolysis process, and model predictions are compared to experimental data pro- duced in a lab-scale fluidized-bed reactor. A parametric study of operating conditions was also per- formed. The kinetic model is based on superimposed hemicellulose, cellulose, and lignin reactions. General biomass feedstock can be represented through the initial mass composition with respect to the three components. The gas–particle flow is modeled with a multi-fluid description (gas, sand, bio- mass) derived from the kinetic theory of granular flows. The predicted product yields at the reactor outlet are presented and compared with the experimental measurements for both pure cellulose and red oak pyrolysis, and encouraging quantitative agreement is achieved. The model is then applied to investigate the effect of various operating conditions on the pyrolysis product yields in the reactor. Results indicate that biomass particle size and superficial gas velocity influence tar yield and residence time considerably with a fixed bed height. For the range of operating temperature studied, the model captures the trend of biomass decomposition versus temperature and shows an optimal temperature of about 500 °C for bio- oil production as reported in the literature. Different biomass feedstocks are also simulated and model shortcomings are discussed. Ó 2012 Elsevier Ltd. All rights reserved. 1. Introduction In 2010, global biofuel production reached 28 billion US gallons, and biofuel accounted for 2.7% of the world’s transportation fuel [1]. Concerning the transportation fuel dependence and security on fossil fuels and the environmental impacts of greenhouse gas emissions associated with fossil fuels, many countries and regions have been driven to rapidly develop and deploy second-generation biofuels derived from lignocellulosic biomass that does not com- pete directly with food supplies as first-generation corn ethanol. For example, the stated objective of the US Department of Energy is to develop biomass technologies to make cellulosic biofuels cost competitive with petroleum-based fuels and support the EISA goal of a minimum 36 billion gallons of renewable fuels per year, including advanced and cellulosic biofuels and biomass-based die- sel by 2022 [2]. The European Union began to increase biodiesel production more rapidly in 2005 and has mandated that biofuels account for 10% of transportation fuel use by 2020 [3]. Stephen et al. [4] studied that fuel made from wood could become a competitive commercial alternative to fuel made from corn by 2020 if the wood biofuel industry is supported. Among technolo- gies for converting cellulosic biomass to biofuels, thermochemical conversion techniques (pyrolysis and gasification) show great promise for being economically comparative with conventional petroleum derived gasoline and diesel [5–7]. Particularly, the fast pyrolysis process in fluidized-bed reactors has attracted many researchers [8–15]. Considerable progress has been achieved for both gasification and pyrolysis routes for biofuels production [16], however, some technical challenges still need to be solved to enable large-scale industrialization of these processes [17]. Empirical correlations and lab-scale or pilot-scale experiments have shown limitations for chemical engineering process design in terms of cost, scale- up, and optimization of operating conditions. Measurements of reacting dense gas–solid flows in a three-dimensional full-scale reactor are very challenging and restricted with today’s experi- mental techniques. With the continuous advancement of state- of-the-art computational fluid dynamics (CFD) methods, however, detailed CFD models for the system should allow for the design and 0016-2361/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.fuel.2012.02.065 ⇑ Corresponding author. Tel.: +1 515 294 7642; fax: +1 515 294 2689. E-mail addresses: ernest.xue@gmail.com, qlxue@iastate.edu (Q. Xue), ddalluge@ iastate.edu (D. Dalluge), theindel@iastate.edu (T.J. Heindel), rofox@iastate.edu (R.O. Fox), rcbrown@iastate.edu (R.C. Brown). Fuel 97 (2012) 757–769 Contents lists available at SciVerse ScienceDirect Fuel journal homepage: www.elsevier.com/locate/fuel