Scale Reactor Research Articles

Scaling a xeno-free fed-batch microcarrier suspension bioreactor system from development to production scale for manufacturing XF hMSCs

Background & Aim With over 800 clinical trials investigating the use of mesenchymal stem/stromal cells (hMSCs) for regenerative medicine, there is an urgent need for the development of economical biomanufacturing processes capable of generating billions to trillions of cells per manufacturing lot in order to meet the demand of future commercial applications. Innovative manufacturing technologies, such as suspension bioreactors, show great promise in reaching commercially-viable working volumes; however, scalability of cell production within such systems remain a challenge, hindering widespread adoption of this culture system for hMSCs. The Quality by Design (QbD) approach was used to develop a scalable xeno-free (XF) hMSC bioreactor process that maintains the final cell population doubling level (PDL) within the recommended range (16-20) to ensure product quality, even as the bioreactor culture volume is scaled to hundreds of liters. The strategic XF bioprocess was designed using high volume XF cell banks, an optimized XF fed-batch media system, and XF microcarriers combined with a scalable bioreactor system to meet design criteria and streamlined production at different culture scales. Methods, Results & Conclusion In this study, the ability to scale XF hBM-MSC expansion in low shear single-use, Vertical-Wheel suspension bioreactors (PBS Biotech) was evaluated in systems of different scale including: small scale (0.1L), development scale (3L), pilot scale (15L), and production scale (50L) reactors. Cell densities greater than 500,000 cells/mL (with >95% viability) were achieved within 4 to 5 days of fed-batch culture at each bioreactor scale. Scalability of the bioreactor culture process was supported by comparable nutrient and waste metabolite profiles, cell growth curves, and pH observed in each system. Furthermore, harvested cells from each bioreactor scale demonstrated comparable critical hMSC quality attributes including angiogenic cytokine (FGF, HGF, IL-8, TIMP-1, TIMP-2, and VEGF) secretion, tri-lineage differentiation potential, and inducible immunomodulatory potential (as measured by functional IDO activity) compared to control cells of similar PDL cultured in 2D. The data generated by this study demonstrates the expansion of XF hBM-MSCs in a scalable bioreactor culture system, providing significant time and cost savings as a plug-and-play standardized system for translational researchers and product developers in the regenerative medicine, tissue engineering, and cell therapy fields.

A multimodular pseudoheterogeneous model framework for optimal design of catalytic reactors exemplified by methanol synthesis

Model-based design of chemical reactors via dynamic optimization can be performed using a recently developed approach called Multi-Level Reactor Design (MLRD). As a start, the MLRD method, its extensions and applications over the last ten years are reviewed compactly. In the main part of this contribution, a further extension to the model framework applied within this method allowing for the rigorous consideration of intrapellet transport processes in heterogeneously catalyzed systems is introduced. Accordingly, a system of coupled differential equations describing the mass and energy balance of both, reactor and catalyst pellet is set up. Balance equations are kept strictly distinguished from kinetic approaches, leaving the latter to be easily exchangeable. This leads to a multimodular model structure used to incorporate different models for diffusion flux, heat transport and pressure drop. The type of the extended model structure is “pseudoheterogeneous”, i.e., the domain in which chemical reactions take place is shifted from a commonly applied pseudohomogeneous mixed phase to the catalyst interior while a solid phase on reactor scale is not modelled. Applying the extended and modularized method to methanol synthesis as a case study reveals considerable concentration and temperature gradients inside the catalyst. These are neglected implicitly in pseudohomogeneous approaches which can lead to a violation of the allowed range of operating conditions for a given catalyst. The extended MLRD model framework using a pseudoheterogeneous model overcomes this drawback and allows for apparatus independent reactor design while meeting process specific constraints in both the bulk gas and the catalyst phase. Further, a systematic investigation applying 27 combinations of diffusion flux, heat transport and pressure drop models reveals that a certain performance can be reached with any combination, while the deduced reactor design differs significantly.

Experimental Evaluation of Napier Grass Gasification in an Autothermal Bubbling Fluidized Bed Reactor

Air gasification of Napier grass (NG) was studied with the target of producing combustible synthesis gas to be used in direct combustion for power generation. A small-scale autothermal bubbling fluidized bed gasifier was used to investigate the effect of reactor temperature, equivalence ratio (ER), and static bed height (SBH) on gasification performance and combustibility of the producer gas. The main generated species in syngas were identified through gas chromatography (GC) analysis. Minimum fluidization conditions were determined at different levels of SBH. Experiments carried out with two intentions of first, to achieve the highest composition of combustible species to ensure the maximum Lower Heating Value (LHV) of syngas and second, to obtain a high performance process with maximum yield of syngas and minimum residues. The results showed that the temperature and ER have significant effects on syngas yield and composition. SBH was found have a substantial effect on the production of H2 and CO. The results from this study was compared to other gasification studies from literature which have evaluated biomass gasification in bubbling fluidized bed reactors with different scales but almost similar method of experimentation. The purpose of verification was to demonstrate the effect of different reactor scales and heating characteristics on the results.

Plasma etching of high aspect ratio features in SiO2 using Ar/C4F8/O2 mixtures: A computational investigation

Plasma etching of high aspect ratio (HAR) features, typically vias, is a critical step in the fabrication of high capacity memory. With aspect ratios (ARs) exceeding 50 (and approaching 100), maintaining critical dimensions (CDs) while eliminating or diminishing twisting, contact-edge-roughening, and aspect ratio dependent etching (ARDE) becomes challenging. Integrated reactor and feature scale modeling was used to investigate the etching of HAR features in SiO2 with ARs up to 80 using tri-frequency capacitively coupled plasmas sustained in Ar/C4F8/O2 mixtures. In these systems, the fluxes of neutral radicals to the wafer exceed the fluxes of ions by 1–2 orders of magnitude due to lower threshold energies for dissociation compared with ionization. At low ARs (<5), these abundant fluxes of CFx and CxFy radicals to the etch front passivate the oxide to form a complex which is then removed by energetic species (ions and hot neutrals) through chemically enhanced reactive etching, resulting in the formation of gas phase SiFx, COx, and COF. As the etching proceeds into higher ARs, the fractional contribution of physical sputtering to oxide removal increases as the fluxes of energetic species to the etch front surpass those of the conduction constrained CFx and CxFy radicals. The instantaneous etch rate of oxide decreases with increasing aspect ratio (ARDE effect) due to decreased fluxes of energetic species and decreased power delivered by these species to the etch front. As the etch rate of photoresist (PR) is independent of AR, maintaining CDs by avoiding undercut and bowing requires high SiO2-over-PR selectivity, which in turn requires a minimum thickness of the PR at the end of etching. Positive ions with narrow angular distributions typically deposit charge on the bottom of low AR features, producing a maximum in positive electric potential on the bottom of the feature. For high AR features, grazing incidence collisions of ions on sidewalls depositing charge produce electric potentials with maxima on the sidewalls (as opposed to the bottom) of the feature.

Experimental data based combinatorial kinetic simulations for predictions of synergistic catalyst mixtures

Abstract Physical mixtures of catalysts can sometimes show synergistic activity which exceeds the sum of the activities of the same amount of catalysts employed separately. However, identifying such synergistic mixtures is non-trivial. Not all mixtures display synergy, and the number of combinations that are possible (even for binary mixtures of equal portions) scales very rapidly, and thus they would be costly to screen experimentally. In this work, we show that it is possible to predict synergistic mixtures using combinatorial kinetic simulations based on experimental data collected on individual catalysts. The data was first collected for conditions-of-interest under low conversions (also called near differential conditions, such that each condition approximates a small volume in a non-differential conditions reactor) to build a library of kinetic models (one model for each catalyst). This data was then used for combinatorial kinetic simulations of non-differential reactor scale conversions, and successfully predicted the qualitative behavior of two synergistic physical mixtures. The capability was utilized in the present work in the context of converting CO, C3H6, and NO species in simulated car exhaust, but is a general approach. We provide equations to estimate the costs of the screening method as well as the combinatorial kinetics method and show that the costs of identifying synergistic physical mixtures become much lower with the combinatorial kinetics method even before ten catalysts of interest. The cost of using prediction by the combinatorial kinetics method continue to become lower (relative to screening) when more catalysts are added -- and even more so if mixtures beyond two components, or consideration of multiple catalyst ratios, is of interest.

Multiscale multiphase phenomena in bubble column reactors: A review

This article presents a state-of-the-art review focusing on the current understanding of the multiscale multiphase phenomena inside Bubble Column Reactors (BCRs). Although many reviews are available on BCRs, little attention has been devoted to summarizing multiscale multiphase phenomena, which are common fundamental issues encountered in their applications. These issues range from the microscale of single bubble dynamics to the mesoscale of bubble swarms and up to the macroscale of the reactor. Understanding these phenomena in all relevant scales can help the rational design, scale-up and optimization of BCRs. The microscale bubble dynamics including the bubble shape and motion, the relevant forces involved and the single bubble mass transfer, is summarized. At the mesoscale, the hydrodynamics of a bubble swarm is influenced by the bubble-bubble or bubble-liquid interactions and hence the overall transport properties of a bubble swarm are not linearly related to that of a single bubble. The bubble swarm effect and the bubble breakage and coalescence mechanisms are discussed in detail. In the end, the macroscale or reactor scale dynamics is strongly governed by the interplay between microscale and mesoscale phenomena, but more research focusing on mesoscale phenomena will be particularly needed for improving our understanding of BCRs.

Avalanche mechanism for runaway electron amplification in a tokamak plasma

The avalanche of runaway electrons is thought to pose a significant obstacle to the success of reactor scale devices such as ITER. As a result, a significant effort has been devoted toward quantifying both the threshold for the initiation of the avalanche of runaway electrons and the efficiency of the avalanche mechanism. In this work, these two quantities are computed utilizing a guiding-center formulation with large-angle collision operators of varying physics fidelity. The use of a guiding-center formulation, while computationally more costly compared to bounce-averaged approaches, provides a conceptually straightforward means of incorporating tokamak geometry. It is found that while the avalanche threshold is only weakly impacted by toroidal geometry for fully ionized low-Z plasmas, it can be significantly impacted if high-Z impurities are present. Furthermore, it is shown that the efficiency of the avalanche mechanism depends sensitively on the impurity content, the charge state of the underlying impurities, and the radial profile of the seed electron population. Finally, the commonly employed Møller secondary source term used to model the generation of secondary electrons is shown to yield avalanche growth rates and thresholds in good agreement with a more complete conservative large-angle collision operator.

Pore engineering of hierarchically structured hydrodemetallization catalyst pellets in a fixed bed reactor

Probing the effects of hydrodemetallization (HDM) catalyst pore structures is of great importance. In this work, the pore structures of hierarchically structured HDM catalyst pellets are optimized based on a Random Spheres Model at the reactor scale. Understanding of the metal deposition behaviors and the concentration profiles along the reactor indicates strong macropore-dependent catalyst pellet utilization and mesopore-dependent bed utilization, respectively. Subsequently, relationships of the effective diffusivity with the macropore structure and the concentration gradient along the reactor with the mesopore structure are proposed, and the effects of the macro-/mesopore diameters and porosity are studied to maximize the exit conversion and catalyst lifetime. Our optimized results not only are superior to previously optimized results, but also provide rational interpretations for previous experimental observations. The insights revealed here could shed new light on the rational pore engineering of the hierarchical structured HDM catalysts.

Kinetic analysis of methane production from anaerobic digestion of water lettuce (Pistia stratiotesL.) with waste sludge

AbstractBACKGROUNDIncreasing demand on energy sources has motivated research studies on renewable energy originating from biomass. Water lettuce (Pistia stratiotes L.) as an invader lignocellulosic biomass with high crop yield can be utilized in anaerobic digestion to produce biomethane. Substrate and inoculum concentrations, and temperature are important parameters for kinetic analysis of methane production that was performed with water lettuce and waste sludge for the first time. Kinetic analysis can be used in reactor design, scale up and identification of optimum conditions.RESULTSIn the present study, anaerobic digestion of water lettuce with waste sludge as inoculum was performed in a batch system. Substrate concentration [30, 40 and 50 g total solid (TS) L−1], waste sludge concentration (3.4 and 6.8 g TS L−1) were investigated at different digestion temperatures (35, 45, 55 and 65 °C). In the studied range, the highest biogas yield [321 mL g–1 volatile solid (VS) with 72.5% of methane content) was reached at substrate concentration of 30 g TS L−1 and waste sludge concentration of 6.8 g TS L−1 at 35 °C. The Cone model fitted well to the actual methane production compared to the modified Gompertz model. The effect of digestion temperature on methane production potential was described by the Ratkowsky model (R2 = 0.91) and optimum digestion temperature was found as 45 °C. Experimental methane yields were 81.5–232.7 mL methane g–1 VS which were 27.7–79.0% of theoretical maximum methane yield. Energy conversion efficiency was reached up to 78.4%.CONCLUSIONThe results revealed that anaerobic digestion of water lettuce with waste sludge for high methane production rate was accomplished at certain substrate and inoculum concentrations and temperature. © 2019 Society of Chemical Industry

Is reactor multiphase thermal-hydraulics a mature field of engineering science?

The progress in multiphase science and technology has been driven to a considerable extent by the problems arising from nuclear reactor thermal-hydraulics. However, the needs for obtaining solutions to a variety of practical questions have had both pros and cons. The former have been reflected in the role of needs as an obvious research stimulation factor, the latter have been due to fact that the urgency of producing results often limited systematic investigations of the underlying physics. Consequently, the maturity of this field has been frequently questioned in the past, in particular as compared to the single-phase fluid mechanics and heat transfer.The objective of this paper is to present an overview of selected issues which, in the author’s personal opinion, illustrate the current state of knowledge in regard to various aspects of multiphase physics and its application to reactor thermal-hydraulics (TH). The discussed issues include: theoretical fundamentals of multiphase fluid mechanics and heat transfer, formulation and limitations of closure laws, transition from correlations to models, consistency vs. accuracy (or validation vs. tuning), and challenges associated with the multicomponent/multiphase modeling of core degradation phenomena.Examples shown to illustrate selected issues come from several studies performed by the author and his collaborators and associates. They cover a broad range of scales, from micro-scale phenomena at, or below, the individual bubble level, to the scale of reactor fuel assemblies.Given the complexity of the multiphase thermal-hydraulics, the author is aware that a lot of room is still available for different opinions on the current state of affairs in this field. So, the overall goal of the paper is to stimulate constructive criticism and future discussions among members of the nuclear TH community.

Performance Evaluation of Continuous Flow Aerobic Granular Sludge Configurations

Increasing the efficiency and capacity of existing wastewater treatment plants can be carried out by using intensive biological processes. One of the currently studied biological solutions consists in using aerobic granular sludge in order to achieve both organics and nutrients removal simultaneously in one tank and with high efficiency. Aerobic granular sludge is currently used at full scale in sequential batch reactors, research for identifying the optimal solutions for continuous flow systems being carried out worldwide. The paper summarizes the results obtained for two continuous flow configurations with aerobic granular sludge, in terms of organics and nutrients removal for synthetic wastewater in laboratory conditions. Both experimental setups led to wastewater treatment efficiencies, with values ranging between 80 and 99% for COD, 85 and 99% for BOD5, 52 and 98% for NH4+ and 5 to 87% for TP.

Operational performance and biofoulants in a dynamic membrane bioreactor

In this study, a mathematical model was developed to have a better understanding of the process and be used in future reactor scale models to predict its process performance. This model utilizes the Activated Sludge Model NO.1 (ASM1) framework and incorporates bioprocesses of formation and degradation of soluble microbial products (SMP) and extracellular polymeric substances (EPS). Simulation result shows the model could very well predict the bioreactor performance. The average error of COD, BOD and NH3-N removal efficiency was 0.48, 0.28 and 1.18%, respectively.

Multi-scale modeling and simulation of a novel membrane reactor (MR)/adsorptive reactor (AR) process

The objective of this study is the modeling and simulation of a novel process, consisting of a membrane reactor (MR)/adsorptive reactor (AR) sequence that intensifies equilibrium-limited reactions and associated separation processes. A comprehensive, multi-scale, multiphase, steady-state/dynamic, computational fluid dynamics (CFD)-based process model is developed, that quantifies the many underlying complex physicochemical phenomena occurring at the pellet, and reactor scales. Model simulations are then carried-out, to study the influence of the individual units’ operating and design parameters on overall system performance, and to better understand this simultaneous reaction/separation, steady-state/cyclic process. Simulation-based optimization studies are then carried out to identify acceptable ranges for both operating and design parameters, so as to meet pre-defined process performance specifications.

Understanding nitrogen recovery from wastewater with a high nitrogen concentration using microbial electrolysis cells

This study was aimed at understanding the effect of applied voltage, catholyte and reactor scale on nitrogen recovery from two different organic wastes (digestate and pig slurry) by means of microbial electrolysis cell (MEC) technology. For this purpose, MEC sizes of 100, 500 and 1000 mL were tested at applied voltages of 0.6, 1 and 1.4 V using either a phosphate-buffered solution or NaCl solution as the catholyte. By increasing the reactor size from 500 to 1000 mL, a decrease in the ammonia recovery efficiency from 47 to 42% was observed. The results also showed that the phosphate-buffered solution is preferable as the catholyte and that the voltage applied does not have a noticeable effect on current production and ammonia recovery. Low biodegradability of the wastes was identified as the main bottleneck.

Integration of phase distribution from gamma-ray tomography technique with monolith reactor scale modeling

In this work, a monolith reactor model was developed to study the effect of phase distribution on the performance of the monolith reactors by integrating phase distribution data from the experiments. To obtain the phase distribution at the reactions conditions, gamma ray computed tomography (CT) was used. Effect of gas density, surface tension of liquid and operating conditions on the phase distribution were studied using gamma-ray computed tomography. Experiments were conducted in Taylor flow regime. With the increase in gas density, uniformity increases. Surface tension has little effect on the distribution in the investigated conditions. Liquid with lower surface tension and gas with lower density has higher cross-sectional liquid saturation. Further, the monolith reactor model with uniform phase distribution and actual distribution were compared. It was found that at higher velocities both gives the same reactor performance irrespective of the maldistribution, however at low velocities, they differ significantly due to the maldistribution. If the maldistribution is present at both low and high velocities, catalyst utilization plays a major role, thus it is recommended to operate at higher velocities where catalyst utilization is high due to high mass transfer.

Comparison between different strategies for the realization of flashing-light effects – Pneumatic mixing and flashing illumination

A major limitation for the phototrophic cultivation of microalgae in photobioreactors is the low culture density which results in expensive harvesting and downstream processes. It is often mentioned that intensive mixing of the culture can improve the efficiency of light utilization by realizing flashing-light effects which enhance the growth of cell cultures. Alternatively, flashing light sources could provide a way to realize this effect without the need of supplying additional energy for fast mixing. While the effects of mixing have been investigated at various reactor scales, experiments with flashing illumination have been mostly conducted in small geometries. In addition, few studies have been conducted for photobioreactors with larger light paths, being characteristic for the production scale. By means of numerical simulations, we evaluate pneumatic mixing and flashing illumination with regard to their ability to realize flashing-light effects in a 5 cm diameter bubble column filled with a suspension of Chlamydomonas reinhardtii. To the best of our knowledge, a numerical comparison between the effects of flashing illumination and pneumatic mixing on the culture growth has not been reported in literature.A thorough comparison of the two methods requires a robust numerical tool which integrates the computation of the fluid flow and the light field, as well as the growth kinetics of algal cells. In the present work, we compute the three-dimensional flow field in a bubble column photobioreactor tracking also micrometer particles in order to simulate the movement of algae. The spectral light field is computed by solving the three-dimensional Radiation Transfer Equation (RTE) at different wavelengths and biomass concentrations. The coupling of the flow and light fields enables the computation of the spatio-temporal light exposure of individual algae cells, which is used to estimate the respective dynamic photosynthesis reaction rates under different operating conditions of the bubble column reactor.We found, numerically, that the contribution of pneumatic mixing alone is negligible in comparison to flashing-light effects, for the investigated operating conditions. In contrast, illumination with flashing LED leads to an increase of the growth rate up to a factor of 2.5 at flashing frequencies higher than 50 Hz in a PBR with industrially relevant operating conditions. Thereby, a proper selection of the duty cycle is not only needed to prevent photoinhibition but also to maximize the effects of flashing light at the same time. According to these results, the utilization of flashing LED sources can provide a mostly unused way to improve photobioreactor productivity in the future.

Application of Taguchi’s experimental design method for optimization of Acid Red 18 removal by electrochemical oxidation process

Background: Electro-oxidation is developed as an electrochemical method to overcome the problems of the conventional decolorization technologies and is an appropriate alternative for the treatment of colored wastewater from various industries. The purpose of this study was to evaluate the efficiency of the electrochemical oxidation process in removal of chemical oxygen demand (COD) and Acid Red 18 (AR18) dye from aqueous solutions. Methods: In this research, a laboratory scale of electro-coagulation reactor for the treatment of synthetic wastewater was made and studied. The effects of different variables including pH, current density, dye concentration, and electrolysis time were investigated. The experiment steps were designed by DesignExpert 10 software using the selected variables. Finally, the dye and COD analysis was performed by spectrophotometer. The optimization was performed using Taguchi fractional factorial design during the removal of dye and COD. Results: Maximum removal of dye (89%) and COD (72.2%) were obtained at pH=3, current density=20 mA/cm2 , initial dye concentration=100 mg/L, and reaction time=45 min. ANOVA test showed a significant relationship between statistical model and test data. Also, the results indicate that the distribution of the residues of the model was normal. Conclusion: By designing experiments through Taguchi method, the removal process will be optimized and by decreasing the number of experiments, the optimal conditions for pollutant removal will be prepared. The results suggest that the Electro-oxidation system is a very suitable technique for the enhancement of wastewater treatment.

Multiscale CFD modelling and analysis of TBR behavior for an HDS process: Deviations from ideal behaviors

For the hydrotreatment process of petroleum cuts, there is a great diversity and level of Trickle Bed Reactors (TBR) models, in which it is very common to use important simplifications, such as the consideration of plug flow; these simplifications restrict the analysis of the multiphysics and multiscale nature of the phenomena that take place in such systems. The analysis of the phenomena that take place at different scales in the TBRs is very scarce; moreover, the qualitative and quantitative determination of the information that passes from one to another scale and its influence on the mass transport nature in these systems is even more difficult to find. With the purpose of analyzing the multiphysic and multiscale nature of a reactor for an HDS process, three Computational Fluid Dynamics (CFD) models were studied at three different scales, corresponding to: (i) Heterogeneous micropores model (HMM) for a particle with explicit representation of the internal surface microstructure of the catalyst, (ii) Pseudohomogeneous catalyst particle model (PCM), and (iii) Reactor scale model (RSM). The objective sought is to establish the validity of some of the assumptions commonly made in the simulation of these types of reactors and develop an analysis to determine the importance of establishing bridges of information between the phenomena that take place at different scales. The results show that despite the small dimensions of the reactor, the plug flow assumptions might not be valid, due to the presence of strong radial gradients. It was also determined that the models that do not take into account the communication between scales (catalysts to reactor scale) differ substantially from those that do, which can be verified by the differences found for the effectiveness factors, of around 35%.

Effect of organic loading rate on anaerobic digestion of pig manure: Methane production, mass flow, reactor scale and heating scenarios

The effect of organic loading rate (OLR) with total solid (TS) control (3%–8%) on the performance of anaerobic digestion of pig manure (PM) using completely stirred anaerobic reactor was investigated. Based on the lab data, how OLR affects mass flow, construction scale and heating requirement in a farm-scale biogas plant was calculated. And three scenarios of typical reactor-heating technology were comparatively analyzed. The optimal OLR was 1.89 g volatile solid (VS)/(L.d) with methane yield of 438.38 mL/gVS in the lab condition. The lower OLR, the larger amount of water and energy consumption, lower methane production and larger amount of liquid digestate was observed. Thus, the reactor with low OLR was suitable in tropical regions with the main target of disposing PM and fertilizer production. High OLR has advantage in the investment, but owns risk of instable process for a long-term run. In our study, among the three heating supply scenarios, biogas boiler was the best option for the designed biogas plant with the given breeding scale under moderate OLR. Combined heat and power (CHP) has potential advantage for the biogas plant under high OLR.

Effect of heat losses on monolithic reactors for VOC abatement

In this contribution a theoretical study of the thermal effects due to heat losses on monolithic reactors for the catalytic combustion of VOC is carried out.To describe the performance of the reactor under steady-state non-isothermal non-adiabatic conditions a heterogeneous 1D Multiple Channel Model is considered, where the monolith is represented by several zones in the form of square concentric rings, each one represented by a square section single channel. Internal and external mass transfer diffusional limitations are taken into account as well as external heat transfer resistance. Heat transfer by conduction through the solid (cordierite) is considered along the cross-section coordinate of the monolith, i.e., from the central zones to peripheral zones at lower temperatures.Although an efficient thermal insulation is selected, the channels of the periphery can operate several degrees colder that the channels of the central zones, these lower temperatures are associated to non-complete VOC conversions, which are emitted into the atmosphere. Studies of parametric sensitivity with respect of inlet temperature reveals that heat losses become more relevant when the reactor is operating near the Emission Limit Value.In relation to design considerations, as the reactor scale decreases and the external area/volume ratio increases, the effect of heat losses on VOC emissions is magnified. The inlet temperature set point should be high enough to prevent high VOC emissions due to lower inlet VOC concentration in a heat loss scenario. This results in higher energetic requirements to ensure VOC complete conversion.To face the reactor performance issues related with heat losses to the environment, optimized designs were proposed aiming to increase the residence times in peripheral zones, which effectively enhance the reactor performance.