(729d) Thermal Conversion of Biomass: Pyrolysis Reaction Networks for Optimal Reactor Design Conference: AIChE Annual MeetingYear: 2014Proceeding: 2014 AIChE Annual MeetingGroup: Computing and Systems Technology DivisionSession: Modeling and Computation in Energy and Environment Time: Thursday, November 20, 2014 - 4:09pm-4:27pm Authors: Zygourakis, K., Rice University Gooding, A. A., Rice University Biochar is charcoal generated for intentional soil amendment by pyrolyzing sustainable biomass feedstocks. In addition to providing a method for carbon sequestration, properly engineered charcoals can increase the water holding and cation exchange capacities of soils, improving the ability of plants to survive under drought conditions and reducing fertilizer runoff into watersheds. Fertilizer runoff has become a serious problem, because as much as 70% of fertilizer applied to crop fields is leached into the groundwater or lost to streams and rivers, eventually leading to large hypoxic “dead” zones in the world’s oceans (including the Gulf of Mexico). One desirable quality of reactors used for the thermochemical conversion of biomass is that they be “feedstock-blind”: capable of producing consistent products even when the feedstocks or the feedstock properties vary. To achieve a satisfactory level of reproducibility, we must be able to accurately control the operation of a reactor so that each biomass load will be processed with the temperature program required to produce biochar with the desired properties. Biomass pyrolysis involves a complicated series of parallel and sequential reactions that include dehydration, depolymerization and decomposition of the primary biomass components (hemicellulose, cellulose, lignin), devolatilization, and condensation. Therefore, the distribution of the pyrolysis products as well as the chemical and physical properties of the produced biochar will depend on the complete temperature history of the biomass particles. Clearly, the design of optimal reactors for biomass pyrolysis requires a detailed kinetic model for every feedstock. To address this challenge, we will first present a systematic procedure that can be used to develop lumped models for biomass pyrolysis. While many earlier studies based their analysis on thermogravimetric experiments with constant heating ramps, our procedure uses a sequence of step-isothermal experiments to identify the potential pseudo-components of each biomass feedstock. Dynamic simulations and nonlinear least-squares methods are then used to test several reaction networks and select the candidate that best fits the experimental data. We applied this method to develop pyrolysis models for slash pine and eucalyptus wood obtained from a managed forest that belongs to Rice University. Our results showed that a network of first-order parallel and sequential reactions with as many as five solid pseudo-components must be used to accurately describe the pyrolysis dynamics of these woods. The kinetic parameters of the individual reactions and their heats of reaction were also estimated using thermogravimetric and differential scanning calorimetry data. The pyrolysis reaction networks were then used to simulate the transient operation of fixed-bed and rotary kiln or auger reactors used for slow pyrolysis of biomass. Reactors of the second type are usually characterized by (a) biomass residence times that may be three or more orders of magnitude longer than the residence times of gas, and (b) very high solid to gas mass ratios, conditions that allow for simplification of their modeling equations. Transient mass balances were developed for all the gaseous and solid species. The solid phases were assumed to be continuous media that exchange heat with the flowing gas, decompose and release volatile products into the gas phase. Fixed-bed reactor models assumed a stationary solid phase (where spatial variations are induced solely by temperature gradients) and two-dimensional PDEs with axial convection and two dispersion terms for the gases. Two-dimensional PDEs were used to describe the transient energy balances for the solid and gas phases, with heat exchange through the reactor wall since this is the most common method for heating these reactors. Finally, the moving bed reactor models considered plug flow of the solid phases. Due to the significant heat effects of pyrolysis reactions, numerical solutions of the modeling equations revealed that large and prolonged axial and radial temperature gradients develop in these reactors. As a result, biochar particles located in different locations of a fixed-bed reactor may experience widely different temperature histories and, thus, achieve different conversions. Even prolonged exposure to high pyrolysis temperatures may not be enough to eliminate these differences and guarantee that the produced biochar will have uniform properties. A lab-scale fixed-bed reactor was used to validate the fixed-bed reactor models. The reactor had a total volume of 1 liter, was heated by placing it inside a furnace and was equipped with four thermocouples to monitor the temporal and spatial temperature distributions. Axial and radial temperature gradients often exceeding 100oC per inch were recorded in this small reactor when the furnace temperature was set to 500oC. Different spatial and temporal temperature patterns were observed for runs with pine and eucalyptus wood particles, a strong indication that pyrolysis models should be feedstock-specific. We will close this presentation by discussing the implications of these results for the design and optimal operation of “feedstock-blind” reactors.