Multiphase Flow With Interphase eXchanges Research Articles

MFIX–DEM simulation of gas-solid flow dynamics in a dual fluidized bed system

Abstract To overcome the operational difficulties associated with bubbling and circulating fluidized bed gasifiers, a new gasification technology termed dual fluidized bed gasification (DFBG) has evolved. To strengthen the understanding of the gasification and combustion processes in the DFBG system, a thorough analysis of bed hydrodynamics is of paramount importance. Furthermore, the complexity of the hydrodynamics escalates while incorporating diverse fuels like biomass and coal, along with bed materials fluidized in a single reactor due to its nonlinearity and transience. As a result, the study of hydrodynamics in such a system is indispensable. The present study focuses on the discrete element model (DEM) simulation of the bed hydrodynamic behaviour within the gasifier of a DFBG system. The simulation was conducted using Multiphase Flow with Interphase eXchanges (MFiX) software. The influence of superficial velocity on the number of particles in a gasifier was analyzed using the 2-D discrete element model (DEM). In this numerical investigation, transient variation of pressure drop, axial and radial solid volume fraction, and particle velocity was evaluated, and the data were analyzed using the ParaView software. The simulation results of the gasifier indicated an increase in the bed pressure drop with the rise in inlet air velocity. The effective height of solid volume fraction also increased with the surge in superficial velocity. The solid velocity profile and streamlines were also investigated to understand its variation pertaining to the location within the fluidized bed.

MFiX-TFM simulation of hydrodynamics of a dual fluidized bed system

The abundance and replenishable nature of biomass make it a suitable fuel for power generation. Among the various technologies, dual fluidized bed gasification (DFBG) is a suitable technology to generate high-quality syngas from biomass. However, the complexity of hydrodynamics and heat transfer with bed geometry, flow conditions as well as feedstock demands effective research for the design of such system. This study, therefore, focuses on the simulation of hydrodynamics of a dual fluidized bed gasification system. The 2-D two-fluid model (TFM) simulations were performed using Multiphase Flow with Interphase eXchanges (MFiX) software by varying the superficial air velocities for both the gasifier and riser. The influence of superficial air velocity on static pressure, pressure drop, axial and radial voidage, solid velocities, and granular temperature was evaluated. From the simulations, it was observed that with the increase in superficial air velocities, there was an increase in the effective height of pressure drop (66%), voidage, and solid velocities of the gasifier. For the simulation of the riser, similar trends in the profile of hydrodynamic parameters were also observed. The optimum velocity for the gasifier and riser has been suggested to be between 0.15 and 0.35 m/s and 6–7 m/s, respectively.

Development of an equation-based parallelization method for multiphase particle-in-cell simulations

Manufacturers have been developing new graphics processing unit (GPU) nodes with large capacity, high bandwidth memory and very high bandwidth intra-node interconnects. This enables moving large amounts of data between GPUs on the same node at low cost. However, small packet bandwidths and latencies have not decreased, which makes global dot products expensive. These characteristics favor a new kind of problem decomposition called “equation decomposition” rather than traditional domain decomposition. In this approach, each GPU is assigned one equation set to solve in parallel so that the frequent and expensive dot product synchronization points in traditional distributed linear solvers are eliminated. In exchange, the method involves infrequent movement of state variables over the high bandwidth, intra-node interconnects. To test this theory, our flagship code Multiphase Flow with Interphase eXchanges (MFiX) was ported to TensorFlow. This new product is known as MFiX-AI and can produce near identical results to the original version of MFiX with significant acceleration in multiphase particle-in-cell (MP-PIC) simulations. The performance of a single node with 4 NVIDIA A100s connected over NVLINK 2.0 was shown to be competitive to 1000 CPU cores (25 nodes) on the JOULE 2.0 supercomputer, leading to an energy savings of up to 90%. This is a substantial performance benefit for small- to intermediate-sized problems. This benefit is expected to grow as GPU nodes become more powerful. Further, MFiX-AI is poised to accept native artificial intelligence/machine learning models for further acceleration and development.

Transient simulation of biomass combustion in a circulating fluidized bed riser

Interest in circulating fluidized bed (CFB) boilers as a power generation technology has sky-rocketed in recent years because of several advantages this technology offers over conventional boilers, such as increased gas-solid mixing, which results in higher combustion efficiency and the ability to use lower rank fuels. CFB combustors are operated at lower temperatures than conventional thermal power generation combustors, thus reducing NOx emissions, while SO2 emissions can be conveniently controlled through the addition of Ca-based sulfur sorbents within the combustor.This paper summarizes the modeling effort on a 50 kWth CFB combustor designed, built, and operated at CanmetENERGY in Ottawa, Canada. The numerical model employs the multiphase particle-in-cell (PIC) approach in the open-source Multiphase Flow with Interphase eXchanges (MFiX) Software Suite. The MFiX-PIC model parameters for the simulation are tuned against cold-flow experiments from CanmetENERGY using olivine sand as the inert bed material. It is shown that for the relatively coarse fluid meshes and large parcel sizes necessitated by the scale of the simulation, filter size dependent corrections to the drag law must be incorporated to ensure accuracy of the simulation results.The validated cold flow model is extended to simulate reacting flow with torrefied hardwood as the feedstock and to validate the combustion reaction scheme. The species concentrations at the riser outlet are compared against CanmetENERGY's experiments and show satisfactory agreement. The simulations demonstrate the ability of MFiX-PIC to accurately capture both the physics and chemistry of a CFB combustor at bench scales, which can be further extended to pilot- and industrial-scale systems.

Assessment of model parameters in MFiX particle-in-cell approach

The limitations in numerical treatment of solids-phase in conventional methods like Discrete Element Model and Two-Fluid Model have facilitated the development of alternative techniques such as Particle-In-Cell (PIC). However, a number of parameters are involved in PIC due to its empiricism. In this work, global sensitivity analysis of PIC model parameters is performed under three distinct operating regimes common in chemical engineering applications, viz. settling bed, bubbling fluidized bed and circulating fluidized bed. Simulations were performed using the PIC method in Multiphase Flow with Interphase eXchanges (MFiX) developed by National Energy Technology Laboratory (NETL). A non-intrusive uncertainty quantication (UQ) based approach is applied using Nodeworks to first construct an adequate surrogate model and then identify the most influential parameters in each case. This knowledge will aid in developing an effective design of experiments and determine optimal parameters through techniques such as deterministic or statistical calibration.

Influence of the pyrolysis and heterogeneous char reactions modeling in the simulation of sugarcane bagasse gasification in a bubbling fluidized bed reactor

The present work investigates different kinetic models for the pyrolysis and heterogeneous char reactions for numerical simulation of the sugarcane bagasse gasification in a bubbling fluidized bed reactor. A two-dimensional Euler-Euler approach is considered, and the reaction rate correlations are implemented in the open code MFIX (Multiphase Flow with Interphase eXchanges) via Fortran sub-routines. The impact of the considered kinetic models for pyrolysis and heterogeneous char reactions upon the prediction of the composition and temperature of the produced gas is evaluated. Six sets of kinetic correlations describing the reactions involved during the sugarcane bagasse gasification are used to perform the computational simulations. Results for the syngas composition obtained in all simulations are compared to available experimental data from literature. Axial profiles for the volumetric fractions of sugarcane bagasse and char, and gas temperature distribution along the reactor are analyzed. Based on the evaluations of this computational study, it is identified that the multiple components parallel competitive reactions pyrolysis model designated as RAN model is superior in the simulations of sugarcane bagasse gasification due to smaller deviations and better prediction of the char formed in the primary pyrolysis. Furthermore, it is shown that the Boudouard and steam gasification reactions modelling have a smaller impact over the predictive capability of the simulations than the char oxidation reaction. Finally, because the results tend to underpredict the O2 amount in the yielded gas, the present work suggests that an improved kinetic correlation for the sugarcane bagasse char oxidation can enhance the simulation predictive capability.

Computational fluid dynamic simulation modeling of carbon capture using polyethylenimine impregnated protonated titanate nanotubes

AbstractA comprehensive computational fluid dynamic (CFD) model of CEES‐developed polyethylenimine impregnated protonated titanate nanotubes (PEI‐PTNTs) was developed using the Multiphase Flow with Interphase eXchanges (MFiX) package to evaluate the performance of the PEI‐PTNTs in a 1‐MW pilot‐scale carbon capture reactor developed by the National Energy Technology Laboratory (NETL). In this CFD model, the momentum, continuity, and energy transport equations were integrated with the first‐order chemistry model for chemical kinetics of heterogeneous reactions to predict the adsorption of CO2 onto amine‐based sorbent particles and the reactor temperature. Based on the amount of the CO2 adsorption obtained in the small‐scale experiment, the coefficients for the chemical reaction equations of PEI‐PTNTs are adjusted. The adjusted PEI‐PTNTs model is applied to the simplified numerical model of 1‐MW pilot‐scale carbon capture system, which is calibrated through the comparison between our simulation results and the results provided by NETL. This calibrated CFD model is used for selecting the optimized flow rate of the gas phase. Our study shows that the optimized gas flow rate to absorb 100% CO2 without loss is 1.5 kg/s, but if higher absorption rate is preferable despite some loss of CO2 absorption in the reactor, a higher flow rate than 1.5 kg/s can be selected.

Optimization of a Cyclone Using Multiphase Flow Computational Fluid Dynamics

Abstract The U.S. Department of Energy National Energy Technology Laboratory's (NETL) 50 kWth chemical looping reactor (CLR) has an underperforming cyclone, which was designed using empirical correlations. To improve the performance of this cyclone using computational fluid dynamics (CFD)-based modeling simulations, four critical design parameters including the vortex tube radius and length, barrel radius, and the inlet width and height were optimized. NETL's open source multiphase flow with interphase exchange (MFiX) CFD code has been used to model a series of cyclones by systematically varying the geometric design parameters. To perform the optimization process, the surrogate modeling and sensitivity analysis followed by the optimization capability in nodeworks was used. The basic methodology for the process is to employ a statistical design of experiments (DOE) method to generate sampling simulations that fill the design space. Corresponding CFD models are then created, executed, and postprocessed. A response surface is created to characterize the relationship between input parameters and the quantities of interest (QoI). Finally, the CFD-surrogate is used by an optimization method to find the optimal design condition based on the objective and constraints prescribed. The resulting optimal cyclone has a larger diameter and longer vortex tube, a larger diameter barrel, and a taller and narrower solids inlet. The improved design has a predicted pressure drop 11 times lower than the original design while reducing the mass loss by a factor of 2.3.

Effect of contact force modeling parameters on the system hydrodynamics of spouted bed using CFD-DEM simulation and 2k factorial experimental design

In this study, the effects of contact force modeling parameters, which included the particle–particle friction coefficient (A), spring constant (B), ratio of the tangential spring constant to normal spring constant (C), normal restitution coefficient (D), and tangential restitution coefficient (E) presenting in the linear spring-dashpot contact model on the hydrodynamics profile of spouted bed reactor, were investigated via computational fluid dynamics coupled with discrete element method (CFD-DEM) using Multiphase Flow with Interphase eXchange (MFIX). First, the base case simulation was validated with the experimental data. Next, the 2k factorial experimental design and an analysis of variance (ANOVA) were conducted to identify the significant main and interaction contact force modeling parameters. Four response variables were observed: the translation kinetic energy of particles, the rotational kinetic energy of particles, the bed expansion, and the standard deviation of pressure drop. Based on these results, the particle–particle friction coefficient, spring constant, and normal restitution coefficient were the major contact force modeling parameters that affected the system hydrodynamics in the spouted bed system. In addition, the effects of main and interaction contact force modeling parameters were summarized, which enhanced our knowledge of the modeling parameters on system hydrodynamics. Contact force modeling parameters must be carefully selected when simulating the spouted bed reactor or related gas-solid multiphase flow process, such as fluidized bed reactor.

Modeling Average Pressure and Volume Fraction of a Fluidized Bed Using Data-Driven Smart Proxy

Simulations can reduce the time and cost to develop and deploy advanced technologies and enable their rapid scale-up for fossil fuel-based energy systems. However, to ensure their usefulness in practice, the credibility of the simulations needs to be established with uncertainty quantification (UQ) methods. The National Energy Technology Laboratory (NETL) has been applying non-intrusive UQ methodologies to categorize and quantify uncertainties in computational fluid dynamics (CFD) simulations of gas-solid multiphase flows. To reduce the computational cost associated with gas-solid flow simulations required for UQ analysis, techniques commonly used in the area of artificial intelligence (AI) and data mining are used to construct smart proxy models, which can reduce the computational cost of conducting large numbers of multiphase CFD simulations. The feasibility of using AI and machine learning to construct a smart proxy for a gas-solid multiphase flow has been investigated by looking at the flow and particle behavior in a non-reacting rectangular fluidized bed. The NETL’s in house multiphase solver, Multiphase Flow with Interphase eXchanges (MFiX), was used to generate simulation data for the rectangular fluidized bed. The artificial neural network (ANN) was used to construct a CFD smart proxy, which is able to reproduce the CFD results with reasonable error (about 10%). Several blind cases were used to validate this technology. The results show a good agreement with CFD runs while the approach is less computationally expensive. The developed model can be used to generate the time averaged results of any given fluidized bed with the same geometry with different inlet velocity in couple of minutes.

Impacts of solid stress model on MP-PIC simulation of a CFB riser with EMMS drag

The cluster-based EMMS (energy minimization multi-scale) drag model in combination with MP-PIC (multi-phase particle in cell) method can successfully simulate the fluidization behavior in CFBs (circulating fluidized beds). The solid normal stress model is very important to MP-PIC simulations, but still lacks thoroughly investigation, especially in CFB riser simulation. In this work, the effects of parameters from the solid normal stress model have been investigated by using 2D simulations of a laboratory-scale CFB riser with MFIX (Multiphase Flow with Interphase eXchanges) open-source code. The results show that the role of the solid stress parameters is more significant when the solid approaches close packing state. Based on the comparison between simulation results and experimental data and the elapsed simulation time, optimal values for the solid stress model parameters were obtained. The thickness of the third direction for 2D simulations has also been investigated. Increase of the thickness entails additional computational costs with little effect on the accuracy and the recommended value of thickness is the diameter of one single particle.

Device-scale computational fluid dynamics modeling of carbon dioxide absorption using encapsulated sorbents

Micro-encapsulated carbon dioxide sorbent (MECS) is considered as a promising material for carbon capture system because of high capture rate, low fabrication cost, and efficient energy consumption. To accelerate the development of MECS technology for industrial deployment, device-scale computational fluid dynamics (CFD) simulations are performed to study hydrodynamics and CO2 absorption behavior in a conceptual fluidized bed reactor. The two-fluid model solver, built in the Multiphase Flow with Interphase eXchanges (MFIX), is adopted to solve the multiphase flow hydrodynamics by representing both fluid and solids as continuous phases. Filtered models are implemented to capture the sub-grid effects that normally cannot be directly considered because MECS particles are too small to be explicitly resolved at such a large scale. Using established CFD tools, this study aims to characterize MECS behavior and predict its carbon capture performance in a conceptual device-scale absorber. The MECS particle size and gas flow rate have significant impact on CO2 capture efficiency. The CO2 capture efficiency decreases with increasing gas flow rate. Smaller MECS particles lead to better CO2 absorption and gas-solid homogeneity.

Development of a Spouted Bed Reactor for Chemical Looping Combustion

We have investigated a novel gas/solid contacting configuration for chemical looping combustion (CLC) with potential operating benefits. CLC configurations are typically able to achieve high fuel conversion efficiencies at the expense of high operating costs and low system reliability. The spouted fluid bed (SB) was identified as an improved reactor configuration for CLC, since it typically exhibits high heat transfer rates and offers the ability to use lower gas flows for material movement compared to bubbling beds (BB). Multiphase Flow with Interphase eXchanges (MFIX) software was used to establish a spouted fluid bed reactor design. An experimental setup was built to supplement the model. The experimental setup was also modified for testing under high temperature, reacting conditions (1073–1273 K). The setup was operated in either a spouted fluid bed or a bubbling bed regime to compare the performance attributes of each. Results for the reactor configurations are presented for CLC using a mixture of carbon monoxide and hydrogen as fuel. Compared to the bubbling bed, the spouted fluid bed reactor achieved an equivalent or better fuel conversion at a lower pressure drop over the material bed. The spouted fluid bed design represents a viable configuration to improve gas/solid contacting for efficient fuel conversion, lower energy requirements for material movement and increase operational robustness for CLC. The research laid the groundwork for future research into a multi-phase reacting flow CLC system. The system will be developed from computational fluid dynamic modeling and pilot-scale testing to expedite the development of CLC technologies.

Sensitivity study of a full-scale industrial spray-injected fluidized bed reactor

The industrial fluidized bed reactor (FBR) described here is designed to convert an aqueous solid laden stream into a consistent granular product. The FBR is heated to 650 °C and chemically reducing conditions are achieved by using steam as the fluidizing gas and coal as a source of carbon. The objective of this study is to use the MFiX (Multiphase Flow with Interphase eXchanges) two-fluid model as a screening tool to investigate the effects of off-nominal conditions to establish performance guidelines. This information will be helpful in assessing whether the identified boundaries of the system are steep or gradual. Initially, 13 independent operating parameters were sampled in a one-at-a-time manner using high, low, and base values and ranked against a set of performance criteria indicative of defluidization. Five operating parameters were found to have the largest effect (bed particle size, bed particle density, coal particle size, spray feed flow rate, and fluidizing gas flow rate) on three quantities of interest—bed differential temperature, low solids velocity, and bed voidage. Latin-hypercube sampling was used to generate a minimal number of random values for various combinations of input parameters. The spray feed flow rate was the most significant parameter affecting the temperature differences, followed by the coal particle and bed particle size. The bed particle size had the largest effect on the low velocity of the bed particles, with the coal particle size as a contributing second effect. The high solid packing shows the coal particle size with the largest effect, and significant secondary contributions from the feed flow rate and fluidizing gas flow rate. The fits to the five-dimensional Gaussian Process models were 0.7797, 0.8664, and 0.9440 for the temperature, velocity, and solids packing, respectively.

Computational Study of Sugarcane Bagasse Pyrolysis Modeling in a Bubbling Fluidized Bed Reactor

This work investigates computationally the modeling of sugarcane bagasse pyrolysis in a bubbling fluidized bed reactor. An Euler-Euler multiphase approach, as invoked by the open code MFIX (Multiphase Flow with Interphase eXchanges), is adopted, and the simulations are carried out in a two-dimensional Cartesian domain. While several pyrolysis kinetic models have been developed for wood, coal, and generic biomass, generally using TGA analysis, no such model has been specifically tested or adapted to simulate sugarcane bagasse pyrolysis in fluidized bed reactors through CFD. In the present study, seven pyrolysis kinetic models available in the literature are implemented as MFIX user-supplied routines, and evaluated within given operational temperature ranges. Initially, well established wood pyrolysis results are used to validate the implementations. Following validation, six kinetic schemes are employed to simulate sugarcane bagasse pyrolysis. Results for the products distribution, formation reaction rate ...

Effects of heat exchanger tubes on hydrodynamics and CO2 capture of a sorbent-based fluidized bed reactor

In virtual design and scale up of pilot-scale carbon capture systems, the coupled reactive multiphase flow problem must be solved to predict the adsorber's performance and capture efficiency under various operation conditions. This paper focuses on the detailed computational fluid dynamics (CFD) modeling of a pilot-scale fluidized bed adsorber equipped with vertical cooling tubes. Multiphase Flow with Interphase eXchanges (MFiX), an open-source multiphase flow CFD solver, is used for the simulations with custom code to simulate the chemical reactions and filtered sub-grid models to capture the effect of the unresolved details in the coarser mesh for simulations with reasonable accuracy and manageable computational effort. Previously developed filtered models for horizontal cylinder drag, heat transfer, and reaction kinetics have been modified to derive the 2D filtered models representing vertical cylinders in the coarse-grid CFD simulations. The effects of the heat exchanger configurations (i.e., horizontal or vertical tubes) on the adsorber's hydrodynamics and CO2 capture performance are then examined. A one-dimensional three-region process model is briefly introduced for comparison purpose. The CFD model matches reasonably well with the process model while provides additional information about the flow field that is not available with the process model.

Particles climbing along a vertically vibrating tube: numerical simulation using the Discrete Element Method (DEM)

It has been reported experimentally that granular particles can climb along a vertically vibrating tube partially inserted inside a granular silo. In this paper, we use the Discrete Element Method (DEM) available in the Multiphase Flow with Interphase eXchanges (MFIX) code to investigate this phenomenon. By tracking the movement of individual particles, the climbing mechanism was illustrated and analyzed. The numerical results show that a sufficiently high vibration strength is needed to form a low solids volume fraction region inside the lower end of the vibrating tube, a dense region in the middle of the tube, and to bring the particles outside from the top layers down to fill in the void. The results also show that particle compaction in the middle section of the tube is the main cause of the climbing. Consequently, varying parameters which influence the compacted region, such as the restitution coefficient, change the climbing height.

CFD-DEM modeling the effect of column size and bed height on minimum fluidization velocity in micro fluidized beds with Geldart B particles

The fluidization behavior of Geldart B particles in micro fluidized beds is investigated numerically using Computational Fluid Dynamics coupled with Discrete Element Method (CFD-DEM) available in the open-source Multiphase Flow with Interphase eXchanges (MFIX) code. The effects of different bed inner diameters (D) of 8mm, 12mm, 16mm and various initial static bed heights (H) were examined. It is found that both decreasing the column diameter and increasing the bed height in a micro fluidized bed increases the minimum fluidization velocity (Umf). The observed overshoot in pressure drop that occurs before the onset of fluidization decreases in magnitude with increasing column diameter, however there is less sensitivity to bed height. Overall, the numerical results agree qualitatively with existing theoretical correlations and experimental studies. The simulations show that both column diameter and particle-wall friction contribute to the variation in minimum fluidization velocity. These two factors are coupled and hard to separate. The detailed influences of wall friction on minimum fluidization velocity are then investigated for a prescribed column diameter of 8mm by varying the wall friction from 0 to 0.4.

Numerical study of gas–solid drag models in a bubbling fluidized bed

ABSTRACTBubbling fluidized beds find application mainly in power conversion industries. For design, dimensioning, and operation of fluidized bed equipment, the understanding of multiphase gas–solid flows is of great importance. The use of computational fluid dynamics in the simulation of gas–solid systems is limited by the complexity of mathematical models, which rely on a series of empirical or theoretical correlations. In the present work, the code Multiphase Flow with Interphase eXchanges (MFIX) was employed to simulate flows in a bubbling fluidized bed and to compare results predicted using different gas–solid drag models. A two-fluid model with kinetic theory of granular flows (TFM-KTGF) was employed, in which gas–solid drag correlations, such as Gidaspow, Hill-Koch-Ladd, or Syamlal and O’Brien, were applied to model momentum transfer between phases. The results predicted were compared with each other and with experimental results from the literature. It was found that the results predicted using each model differ much. The Gidaspow and Hill-Koch-Ladd models yielded bubbles with shapes more similar to the experiments.

Hydrodynamics of gas–solids flow in a bubbling fluidized bed with immersed vertical U-tube banks

We apply a two-fluid model (TFM) from the open-source code Multiphase Flow with Interphase eXchanges (MFIX) to investigate hydrodynamics in a gas–solids fluidized bed with immersed vertical tubes. The cut-cell method implemented in MFIX is used to fully resolve the flow around vertical U-tube banks. Simulations are performed in a bed diameter of 0.145m with square and triangular tube arrangements, for inlet gas velocities of U0/Umf=2.3, 4.5 and 6.8. Simulation results are compared with experimental results from the literature and show very good agreement for the bubble size. The efficiency of vertical tubes in reducing bubble size depends upon inlet gas velocity and tube arrangement. Reduction in bubble size is due to the vertical tubes preventing bubble coalescence and promoting bubble splitting. In-bed vertical tubes result in uniform distribution of bubbles within the bed with increase in bubble frequency. The bubble frequency is higher within the bed for square tube arrangements. For a bed with vertical tubes, the bubble shape is generally elongated, which results in high bubble rise velocity. Axial solid velocity and solids circulation patterns are significantly affected by the vertical tubes, where triangular tube arrangements rarely show any solids circulating zone.