The water–gas shift (WGS) reaction is an attractive process for producing hydrogen gas from lignocellulosic biomass conversion applications.

We discuss the design and implementation of a preliminary software package written in Python 3 that is intended to represent complex multiphase reactors as networks of ideal continuous stirred tank reactors. This software also implements statistical design of experiments, uncertainty quantification, and global sensitivity analysis. These advanced features can provide important qualitative and quantitative insights into the effect of operating conditions and model parameters on predicted reactor performance. We demonstrate the utility of the program by modeling the vapor phase catalytic upgrading of bio-oil in a bubbling fluidized bed reactor.

We report results from a multiscale computational modeling study of biomass fast pyrolysis in an experimental laboratory reactor that combined the hydrodynamics predicted by a two-fluid model (TFM) with predictions from a finite element method (FEM) simulation of heat and mass transfer and chemical reactions within biomass particles. The experimental pyrolyzer consisted of a 2 in. (5.1 cm) diameter bubbling fluidized bed reactor (FBR) fed with milled pine pellets. The predicted FBR hydrodynamics included estimates of the residence times that the gas and biomass particles spend in the reactor before they exit. A single-particle FEM simulation was constructed on the basis of the geometry and heat transfer properties determined from optical and X-ray computed tomography measurements of wood and char particles collected from the experimental FBR, along with previously proposed pyrolysis reaction kinetics. Taken together, the combined TFM and FEM simulation results predicted net bio-oil yields at the reactor exit that agree well with experimental observations, without any arbitrary fitting parameters. The combined computational models also provided practical information about the most important reactor and feedstock parameters.

In this study, the results from laboratory measurements of the devolatilization kinetics of switchgrass in a rapidly heated fixed bed reactor flushed with argon and operated at constant temperatures between 600 and 800°C was reported. Results indicate that switchgrass decomposes in two sequential stages during pyrolysis: stage I involves the evaporation and devolatilization of water and extractives and stage II involves that of hemicellulose, cellulose, and lignin. The estimated global activation energy for stage II increased from 52.80 to 59.39kJ/mol as the reactor temperature was increased from 600 to 800°C. The maximum conversion of carbon, hydrogen, oxygen, sulfur, and nitrogen ranged from 0.68 to 0.70, 0.90 to 0.95, 0.88 to 0.91, 0.70 to 0.80, and 0.55 to 0.66, respectively. The retention of alkali and alkaline earth metal (AAEM) species in the solid char after complete pyrolysis was significantly higher than in the original feed, indicating the importance of AAEM species in subsequent char processing.

We present statistical analyses for highly resolved, short and long-time-scale experimental measurements of magnetic flotsam tracer particle motion in bubbling air fluidized beds of Geldart Group B particles. The observed axial tracer trajectories appear to exhibit Weibull distributions over both short and long time scales and suggest possible analogies with other complex dynamic processes. Comparisons with other recently published experimental and simulated flotsam mixing patterns reveal patterns consistent with our measurements, implying that the observed patterns may be general features of bubbling fluidized beds.

Simulating fast pyrolysis at the scale of single particles allows for the investigation of the impacts of feedstock-specific parameters such as particle size, shape, and species of origin. For this reason particle-scale modeling has emerged as an important tool for understanding how variations in feedstock properties affect the outcomes of pyrolysis processes. The origins of feedstock properties are largely dictated by the composition and hierarchical structure of biomass, from the microstructural porosity to the external morphology of milled particles. These properties may be accounted for in simulations of fast pyrolysis by several different computational approaches depending on the level of structural and chemical complexity included in the model. The predictive utility of particle-scale simulations of fast pyrolysis can still be enhanced substantially by advancements in several areas. Most notably, considerable progress would be facilitated by the development of pyrolysis kinetic schemes that are decoupled from transport phenomena, predict product evolution from whole-biomass with increased chemical speciation, and are still tractable with present-day computational resources.

We report results from a numerical study of multi-time-scale bistable dynamics for CO oxidation on a catalytic surface in a flowing, well mixed gas stream. The problem is posed in terms of surface and gas-phase submodels that dynamically interact in the presence of stochastic perturbations, reflecting the impact of molecular-scale fluctuations on the surface and turbulence in the gas. Wavelet-based methods are used to encode and characterize the temporal dynamics produced by each submodel and detect the onset of sudden state shifts (bifurcations) caused by nonlinear kinetics. When impending state shifts are detected, a more accurate but computationally expensive integration scheme can be used. This appears to make it possible, at least in some cases, to decrease the net computational burden associated with simulating multi-time-scale, nonlinear reacting systems by limiting the amount of time in which the more expensive integration schemes are required. Critical to achieving this is being able to detect unstable temporal transitions such as the bistable shifts in the example problem considered here. Our results indicate that a unique wavelet-based algorithm based on the Lipschitz exponent is capable of making such detections, even under noisy conditions, and may find applications in critical transition detection problems beyond catalysis.

The nanostructure of diesel particulates has been shown to impact its oxidation rate and burnout trajectory. Additionally, this nanostructure can evolve during the oxidation process, furthering its influence on the burnout process. For this study, exhaust particulates were generated on a light-duty diesel engine with conventional diesel fuel, biodiesel, and intermediate blends of the two at a single load-speed point. Despite the singular engine platform and operating point, the different fuels created particulates with varied nanostructure, thereby greatly expanding the window for observing nanostructure evolution and oxidation. The physical and chemical properties of the particulates in the nascent state and at partial oxidation states were measured in a laboratory reactor and by high-resolution transmission electron microscopy as a function of the degree of oxidation in O2. X-ray photoacoustic spectroscopy analysis, thermal desorption, and solvent extraction of the nascent particulate samples reveal a significant organic content in the biodiesel-derived particulates, likely accounting for differences in the nanostructure. This study reports the nanoscale structural changes in the particulate with biofuel blend level and during O2 oxidation as observed by high-resolution transmission electron microscopy and quantitated by fringe analysis and Brunnauer–Emmet–Teller total surface area measurements. It was observed that initial fuel-related differences in the lamella lengths, spacing, and curvature disappear when the particulate reaches approximately 50% burnout. Specifically, the initial ordered, fullerenic, and amorphous nanostructures converge during the oxidation process and the surface areas of these particulates appear to grow through these complex changes in internal particle structure. The specific surface area, measured at several points along the burnout trajectory, did not match the shrinking core projection and in contrast suggested that internal porosity was increasing. Thus, the appropriate burnout model for these particulates is significantly different from the standard shrinking core assumption, which does not account for any internal structure. An alternative burnout model is supported by high-resolution transmission electron microscopy image analysis.

The NO2 oxidation kinetics and burning mode for diesel particulate from light-duty and medium-duty engines fueled with either ultra low sulfur diesel or soy methyl ester biodiesel blends have been investigated and are shown to be significantly different from oxidation by O2. Oxidation kinetics were measured using a flow-through packed bed microreactor for temperature programmed reactions and isothermal differential pulsed oxidation reactions. The burning mode was evaluated using the same reactor system for flowing BET specific surface area measurements and HR-TEM with fringe analysis to evaluate the nanostructure of the nascent and partially oxidized particulates. The low activation energy measured, specific surface area progression with extent of oxidation, HR-TEM images and difference plots of fringe length and tortuosity paint a consistent picture of higher reactivity for NO2, which reacts indiscriminately immediately upon contact with the surface, leading to the Zone I or shrinking core type oxidation. In comparison, O2 oxidation is shown to have relatively lower reactivity, preferentially attacking highly curved lamella, which are more reactive due to bond strain, and short lamella, which have a higher proportion of more reactive edge sites. This preferential oxidation leads to Zone II type oxidation, where solid phase diffusion of oxygen via pores contributes significantly to slowing the overall oxidation rate, by comparison.

We propose a wavelet-based scheme that encodes the essential dynamics of discrete microscale surface reactions in a form that can be coupled with continuum macroscale flow simulations with high computational efficiency. This makes it possible to simulate the dynamic behavior of reactor-scale heterogeneous catalysis without requiring detailed concurrent simulations at both the surface and continuum scales using different models. Our scheme is based on the application of wavelet-based surrogate time series that encodes the essential temporal and/or spatial fine-scale dynamics at the catalyst surface. The encoded dynamics are then used to generate statistically equivalent, randomized surrogate time series, which can be linked to the continuum scale simulation. We illustrate an application of this approach using two different kinetic Monte Carlo simulations with different characteristic behaviors typical for heterogeneous chemical reactions.

The impact of bubbling bed hydrodynamics on temporal variations in the exit tar yield for biomass fast pyrolysis was investigated using computational simulations of an experimental laboratory-scale reactor. A multi-fluid computational fluid dynamics model was employed to simulate the differential conservation equations in the reactor, and this was combined with a multi-component, multi-step pyrolysis kinetics scheme for biomass to account for chemical reactions. The predicted mean tar yields at the reactor exit appear to match corresponding experimental observations. Parametric studies predicted that increasing the fluidization velocity should improve the mean tar yield but increase its temporal variations. Increases in the mean tar yield coincide with reducing the diameter of sand particles or increasing the initial sand bed height. However, trends in tar yield variability are more complex than the trends in mean yield. The standard deviation in tar yield reaches a maximum with changes in sand particle size. The standard deviation in tar yield increases with the increases in initial bed height in freely bubbling state, while reaches a maximum in slugging state. (C) 2015 Elsevier Ltd. All rights reserved.

We report results from a computational study of the transition from bubbling to slugging in a laboratory-scale fluidized-bed reactor with Geldart Group B glass particles. For simulating the three-dimensional fluidized-bed hydrodynamics, we employ MFiX, a widely studied multi-phase flow simulation tool, that uses a two-fluid Eulerian-Eulerian approximation of the particle and gas dynamics over a range of gas flows. We also utilize a previously published algorithm to generate bubble statistics that can be correlated with pressure fluctuations to reveal previously unreported details about the stages through which the hydrodynamics progress during the bubbling-to-slugging transition. We expect this new information will lead to improved approaches for on-line reactor diagnostics, as well as new approaches for validating the results of computational fluidized-bed simulations with experimental measurements.

We report results from computational simulations of an experimental, lab-scale bubbling bed biomass pyrolysis reactor that include a distributed activation energy model (DAEM) for the kinetics. In this study, we utilized multiphase computational fluid dynamics (CFD) to account for the turbulent hydrodynamics, and this was combined with the DAEM kinetics in a multi-component, multi-step reaction network. Our results indicate that it is possible to numerically integrate the coupled CFD-DAEM system without significantly increasing computational overhead. It is also clear, however, that reactor operating conditions, reaction kinetics, and multiphase flow dynamics all have major impacts on the pyrolysis products exiting the reactor. We find that, with the same pre-exponential factors and mean activation energies, inclusion of distributed activation energies in the kinetics can shift the predicted average value of the exit vapor-phase tar flux and its statistical distribution, compared to single-valued activation-energy kinetics. Perhaps the most interesting observed trend is that increasing the diversity of the DAEM activation energies appears to increase the mean tar yield, all else being equal. These findings imply that accurate resolution of the reaction activation energy distributions will be important for optimizing biomass pyrolysis processes. (C) 2015 Elsevier B.V. All rights reserved.

We summarize a global kinetic model for a commercial automotive Lean NOx Trap (LNT) catalyst derived from laboratory flow reactor measurements at Oak Ridge National Laboratory (ORNL). The experimental measurements were made according to a modified version of a publicly shared protocol developed for characterizing the dynamic responses of LNT catalysts over a temperature range of 150–550 °C under lean-rich cycling with CO, H2, and C3H6 as the reductants. The resulting model includes three NOx storage sites. The present model also includes reactions for oxygen storage, water gas shift, steam reforming, NOx reduction, and N2O and NH3 generation. When implemented in 1D integral reactor simulations of long-period cycling, the model was found to accurately predict the observed outlet concentrations of CO, H2, C3H6, NO, NO2, NH3, and N2O.

We present a computationally efficient, one-dimensional simulation methodology for biomass particle heating under conditions typical of fast pyrolysis. Our methodology is based on identifying the rate limiting geometric and structural factors for conductive heat transport in biomass particle models with realistic morphology to develop low-order approximations that behave appropriately. Comparisons of transient temperature trends predicted by our one-dimensional method with three-dimensional simulations of woody biomass particles reveal good agreement, if the appropriate equivalent spherical diameter and bulk thermal properties are used. We conclude that, for particle sizes and heating regimes typical of fast pyrolysis, it is possible to simulate biomass particle heating with reasonable accuracy and minimal computational overhead, even when variable size, aspherical shape, anisotropic conductivity, and complex, species-specific internal pore geometry are incorporated.

The fuel economy and emissions of conventional and hybrid buses equipped with emissions aftertreatment were evaluated via computational simulation for six representative city bus drive cycles. Both series and parallel configurations for the hybrid case were studied. The simulation results indicated that series hybrid buses have the greatest overall advantage in fuel economy. The series and parallel hybrid buses were predicted to produce similar carbon monoxide and hydrocarbon tailpipe emissions but were also predicted to have reduced tailpipe emissions of nitrogen oxides compared with the conventional bus in higher speed cycles. For the New York bus cycle, which has the lowest average speed among the cycles evaluated, the series bus tailpipe emissions were somewhat higher than they were for the conventional bus; the parallel hybrid bus had significantly lower tailpipe emissions. All three bus power trains were found to require periodic active diesel particulate filter regeneration to maintain control of particulate matter. Plug-in operation of series hybrid buses appears to offer significant fuel economy benefits and is easily employed because of the relatively large battery capacity that is typical of the series hybrid configuration.

An investigation of the impact of rail pressure and biodiesel fueling on exhaust particulate agglomerate morphology and primary particle (soot) nanostructure was conducted with a common-rail turbocharged direct injection diesel engine. The engine was operated at steady state on a dynamometer running at moderate speed with both low (30%) and medium–high (60%) fixed loads, and exhaust particulate was sampled for analysis. The fuels used were ultra-low sulfur diesel and its 20% v/v blends with soybean methyl ester biodiesel. Fuel injection occurred in a single event around top dead center at three different injection pressures. Exhaust particulate samples were characterized with transmission electronic microscopy imaging, scanning mobility particle sizing, thermogravimetric analysis, Raman spectroscopy, and X-ray diffraction analysis. Particulate morphology and oxidative reactivity were found to vary significantly with both rail pressure and biodiesel blend level. Higher biodiesel content led to an increase in the primary particle size and oxidative reactivity but had no impact on nanoscale disorder in the as-received samples. For particulates generated with higher injection pressures, the initial oxidative reactivity increased, but there was no detectable correlation with primary particle size or nanoscale disorder.