Search

Regenerative sorption on porous solids can be used to separate many mixtures into their components, to remove an air pollutant present in very small concentrations, or to store a gas. The key to any regenerative system is finding a simple reaction with a small negative standard free energy change. This chapter summarizes typical rates for some of the solid-gas reactions. In these reactions, structural changes are often important, because diffusion takes place through the pores filled with a fluid, not through the solid phase due to its low diffusivity. The chapter proposes the continuum solid phase mass balances ratio called E in a natural way. The most common model which finds some application in this field is the moving boundary or the shrinking core model. The ratio of the initial rate is meaningful for fluid solid reactions with pore closing, because their rates decrease exponentially with time even for large Thiele moduli.

A process has been developed for the removal of carbon dioxide from a gaseous stream and for its subsequent transfer to another gas stream by thermal regeneration. The sorbent is a stable mixture of finely ground potassium carbonate with a high surface area commercial aluminum oxide. Seventy percent of the carbonate can be reacted reversibly with a negligible diffusional resistance in the particles. Potassium carbonate was chosen as the sorbent because the decomposition pressure of the bicarbonate is very low near room temperature. The decomposition pressure of sodium bicarbonate appeared to be too high, while the salts of rubidium or cesium require higher regeneration temperatures and are expensive. At low CO2 concentrations, sorption proceeded for a sufficiently long time so one could obtain rates from outlet concentrations of CO2, rather than from the more time consuming series of regeneration runs.

The hydrodynamics of a ‘fast bubble’ in a two-dimensional rectangular fluidized bed is studied numerically. The fast bubble (which moves faster than the fluidizing velocity) is produced by injecting gas through a jet for a short time into a bed at minimum fluidization. The numerical computations are in good agreement with the classical theories of Davidson, Jackson, and Murray and with experimental data, thus enhancing confidence in the computer modeling of the hydrodynamics of fluidization.

The objective of this investigation is to convert our {open_quotes}learning gas - solid - liquid{close_quotes} fluidization model into a predictive design model. The IIT hydrodynamic model computes the phase velocities and the volume fractions of gas, liquid and particulate phases. Model verification involves a comparison of these computed velocities and volume fractions to experimental values. As promised in the SIXTH TECHNICAL PROGRESS REPORT, January 1996, this report presents our measurements of granular temperature of Air Products catalyst in our new two dimensional bubble column with controlled jets that roughly approximates the Air Products methanol reactor. We have observed much higher granular temperatures with gas flow than was the case for liquid flow with Air Products catalyst. This means there is much better stirring caused by the high gas flow; clearly a desirable feature in a reactor. We have also observed that there exists a maximum in granular temperature at low solids loading. Operationally this means it may not be desirable to operate at very high catalyst concentrations due to reduced stirring at the high catalyst concentrations. This behavior is very similar to that observed by us in a circulating fluidized bed used to make gasoline from coal, a study tomore » appear in the AIChE Journal.« less

The objective of this investigation is to convert our ``learning gas solid-liquid`` fluidization model into a predictive design model. The IIT hydrodynamic model computes the phase velocities and the volume fractions of gas, liquid and particulate phase. Model verification involves a comparison of these computed velocities and volume fractions to experimental values. A hydrodynamic model for multiphase flows, based on the principles of mass, momentum and energy conservation for each phase, was developed and applied to model gas-liquid, gas-liquid-solid fluidization and gas-solid-solid separation. To simulate the industrial slurry bubble column reactors, a computer program based on the hydrodynamic model was written with modules for chemical reactions (e.g. the synthesis of methanol), phase changes and heat exchangers. In the simulations of gas-liquid two phases flow system, the gas hold-ups, computed with a variety of operating conditions such as temperature, pressure, gas and liquid velocities, agree well with the measurements obtained at Air Products` pilot plant. The hydrodynamic model has more flexible features than the previous empirical correlations in predicting the gas hold-up of gas-liquid two-phase flow systems. In the simulations of gas-liquid-solid bubble column reactors with and without slurry circulation, the code computes volume fractions, temperatures and velocity distributions for the gas, the liquid and the solid phases, as well as concentration distributions for the species (CO, H{sub 2}, CH{sub 3}0H, ... ), after startup from a certain initial state. A kinetic theory approach is used to compute a solid viscosity due to particle collisions. Solid motion and gas-liquid-solid mixing are observed on a color PCSHOW movie made from computed time series data. The steady state and time average catalyst concentration profiles, the slurry height and the rates of methanol production agree well with the measurements obtained at an Air Products` pilot plant.

The objective of this investigation is to convert our ``learning gas-solid-liquid`` fluidization model into a predictive design model. The IIT hydrodynamic model computes the phase velocities and volume fractions of gas, liquid and particulate phases. Model verification involves a comparison of these computed velocities and volume fractions to experimental values. The simulation of Air Product methanol reactors described in this paper are continuing. Granular temperatures and viscosities have been computed. Preliminary measurements of granular temperatures using the Air Product catalysts were obtained using our CCD camera.

The objective of this investigation is to convert our {open_quotes}learning gas-solid-liquid{close_quotes} fluidization model into a predictive design model. The IIT hydrodynamic model computes the phase velocities and the volume fractions of gas, liquid and particulate phases. Model verification involves a comparison of these computed velocities and volume fractions to experimental values. After a discussion of our research with the DOE-Air Products team in January, we decided to concentrate on the slurry configuration of interest to DOE-Air Products which has no recirculation of liquid. In such a system the gas is the continuous phase, rather than the liquid that we had used in our model in the past. We have built such a cold flow two dimensional plastic model. We have also changed our computer code. At the request of Air Products and DOE we have started a simulation of LaPorte RUN E-8.1 (1991) for production of methanol as described in the Air Products report sent to us. For isothermal operation, there is good mixing, and the preliminary results shown in this report indicate that we should obtain an agreement between the experiment and the simulations. A final report will be prepared upon completion of the simulation.

The overall objective of this investigation is to develop experimentally verified models for circulating fluidized bed (CFB) combustors. Sorption of S0{sub 2} with calcined limestone was studied in a PYROFLOW type CFB loop at conditions approximating those found in a CFB combustor. Initially the CFB loop contained 150 micron CaO particles of a density of 3.3 g/cm{sup 3} and air at 1143{degrees}K and 3.25 atm. Atzero time, air containing 600 ppm SO{sub 2}, was introduced into the riser bottom at 1143{degrees}K. The effect of gas velocity, sorbent inventory and inlet pressure on the sorption of SO{sub 2}, were studied isothermally by running our hydrodynamic code with the S0{sub 2} sorption conservation of species equation. At a velocity of 5m/sec., reported to be a typical velocity by PYROPOWER, there is reasonably good S0{sub 2} removal. At 10 m/sec the S0{sub 2} removal is poor. The best SO{sub 2}, removal is for a velocity of 5 m/s and a high bed inventory, initial bed height, H = 9m. Most of the S0{sub 2} is removed in the first two meters of the reactor. However, the S0{sub 2} removal is not complete at the bed outlet. This is due to mixing. At themore » left wall of the reactor (wall opposite the solids inlet) the S0{sub 2} removal was poor due to gas bypassing caused by the asymmetrical solids inlet. Simulation of the PYROPOWER loop with a symmetrical inlet gave us an order of magnitude improvement over the conventional PYROPOWER system. These results demonstrate the practical utility of the predictive model that we have developed over the last three years.« less

The overall objective of this investigation is to develop experimentally verified models for circulating fluidized bed (CFB) combustors. Sorption of S0[sub 2] with calcined limestone was studied in a PYROFLOW type CFB loop at conditions approximating those found in a CFB combustor. Initially the CFB loop contained 150 micron CaO particles of a density of 3.3 g/cm[sup 3] and air at 1143[degrees]K and 3.25 atm. Atzero time, air containing 600 ppm SO[sub 2], was introduced into the riser bottom at 1143[degrees]K. The effect of gas velocity, sorbent inventory and inlet pressure on the sorption of SO[sub 2], were studied isothermally by running our hydrodynamic code with the S0[sub 2] sorption conservation of species equation. At a velocity of 5m/sec., reported to be a typical velocity by PYROPOWER, there is reasonably good S0[sub 2] removal. At 10 m/sec the S0[sub 2] removal is poor. The best SO[sub 2], removal is for a velocity of 5 m/s and a high bed inventory, initial bed height, H = 9m. Most of the S0[sub 2] is removed in the first two meters of the reactor. However, the S0[sub 2] removal is not complete at the bed outlet. This is due to mixing. At themore » left wall of the reactor (wall opposite the solids inlet) the S0[sub 2] removal was poor due to gas bypassing caused by the asymmetrical solids inlet. Simulation of the PYROPOWER loop with a symmetrical inlet gave us an order of magnitude improvement over the conventional PYROPOWER system. These results demonstrate the practical utility of the predictive model that we have developed over the last three years.« less

Cohesive particles (Geldart Group C powders) are fine particles generally less than 30 microns in size. Interparticle forces are large relative to inertial forces in these particles, and cause clumping, sticking, and channeling when attempts are made to fluidize them. These solids do not flow easily through pipes, and bridge extremely easily. The objectives of the work in this program were (1) to develop a hydrodynamic model which can be applied to cohesive solids, and (2) to obtain data in a large-scale (30-cm-diameter) riser to test the model. The work was divided into six tasks: Task 1. Preparation of a Project Work Plan; Task 2. Hydrodynamic Model Development; Task 3. Determination of Rheological Properties for Incorporation into the Model; Task 4. Small-Scale Flow Tests; Task 5. Large-Scale Flow Tests; and Task 6. Comparison of Model With Data. The work was conducted by the Institute of Gas Technology (IGT) in collaboration with the Illinois Institute of Technology (IIT). This work combined the expertise of IIT in model development, with the large-scale experimental capabilities of IGT. IIT researchers developed the hydrodynamic model in the program, while the large-scale data were generated by IGT. Following the preparation of the Project Work Plan inmore » Task 1, work was started on the development of a two-dimensional hydrodynamic model to enable the behavior of cohesive solids in a dilute-phase riser to be simulated. In Task 2, two hydrodynamic models were developed based on the kinetic theory model of granular flow. The models were used to predict data presented in the literature, as well as data generated in Task 5 of this study. In Task 3, rheological data on cohesive oil shale with an average particle size of approximately 12 microns was obtained using a unique device called a cohetester.« less

Steady flows influenced by walls cannot be described by inviscid models. Flows in circulating fluidized beds have significant wall effects. Particles in the form of clusters or layers can be seen to run down the walls. Hence modeling of circulating fluidized beds (CFB) without a viscosity is not possible. However, in interpreting Equations (8-1) and (8-2) it must be kept in mind that CFB or most other two phase flows are never in a true steady state. Then the viscosity in Equations (8-1) and (8-2) may not be the true fluid viscosity to be discussed next, but an Eddy type viscosity caused by two phase flow oscillations usually referred to as turbulence. In view of the transient nature of two-phase flow, the drag and the boundary layer thickness may not be proportional to the square root of the intrinsic viscosity but depend upon it to a much smaller extent. As another example, liquid-solid flow and settling of colloidal particles in a lamella electrosettler the settling process is only moderately affected by viscosity. Inviscid flow with settling is a good first approximation to this electric field driven process. The physical meaning of the particulate phase viscosity is described in detail inmore » the chapter on kinetic theory. Here the conventional derivation resented in single phase fluid mechanics is generalized to multiphase flow.« less

Steady flows influenced by walls cannot be described by inviscid models. Flows in circulating fluidized beds have significant wall effects. Particles in the form of clusters or layers can be seen to run down the walls. Hence modeling of circulating fluidized beds (CFB) without a viscosity is not possible. However, in interpreting Equations (8-1) and (8-2) it must be kept in mind that CFB or most other two phase flows are never in a true steady state. Then the viscosity in Equations (8-1) and (8-2) may not be the true fluid viscosity to be discussed next, but an Eddy type viscosity caused by two phase flow oscillations usually referred to as turbulence. In view of the transient nature of two-phase flow, the drag and the boundary layer thickness may not be proportional to the square root of the intrinsic viscosity but depend upon it to a much smaller extent. As another example, liquid-solid flow and settling of colloidal particles in a lamella electrosettler the settling process is only moderately affected by viscosity. Inviscid flow with settling is a good first approximation to this electric field driven process. The physical meaning of the particulate phase viscosity is described in detail inmore » the chapter on kinetic theory. Here the conventional derivation resented in single phase fluid mechanics is generalized to multiphase flow.« less

This topical report on Sulfur Control'' presents the results of work conducted by the Institute of Gas Technology (IGT), the Illinois Institute of Technology (IIT), and the Ohio State University (OSU) to develop three novel approaches for desulfurization that have shown good potential with coal and could be cost-effective for oil shales. These are (1) In-Bed Sulfur Capture using different sorbents (IGT), (2) Electrostatic Desulfurization (IIT), and (3) Microbial Desulfurization and Denitrification (OSU and IGT). The objective of the task on In-Bed Sulfur Capture was to determine the effectiveness of different sorbents (that is, limestone, calcined limestone, dolomite, and siderite) for capturing sulfur (as H{sub 2}S) in the reactor during hydroretorting. The objective of the task on Electrostatic Desulfurization was to determine the operating conditions necessary to achieve a high degree of sulfur removal and kerogen recovery in IIT's electrostatic separator. The objectives of the task on Microbial Desulfurization and Denitrification were to (1) isolate microbial cultures and evaluate their ability to desulfurize and denitrify shale, (2) conduct laboratory-scale batch and continuous tests to improve and enhance microbial removal of these components, and (3) determine the effects of processing parameters, such as shale slurry concentration, solids settling characteristics, agitation rate,more » and pH on the process.« less

The overall objective of this investigation is to develop experimentally verified models for circulating fluidized bed (CFB) combustors. Spectral analysis of CFB data obtained at Illinois Institute of Technology shows that the frequencies of pressure oscillations are less than 0.1 Hertz and that they increase with solids volume fraction to the usual value of one Hertz obtained in bubbling beds. These data are consistent with the kinetic theory interpretation of density wave propagation.

The overall objective of this investigation is to develop experimentally verified models for circulating fluidized bed (CFB) combustors. Spectral analysis of CFB data obtained at Illinois Institute of Technology shows that the frequencies of pressure oscillations are less than 0.1 Hertz and that they increase with solids volume fraction to the usual value of one Hertz obtained in bubbling beds. These data are consistent with the kinetic theory interpretation of density wave propagation.

Rigorous methods of kinetic theory were used to derive particular phase viscosities and granular conductivities. This new kinetic theory predicted flow behavior and oscillations in a complete loop of a CFB. The results were compared to computations with imposed gas phase turbulence in the riser. The computations were repeated for production of synthesis gas from char.

Rigorous methods of kinetic theory were used to derive particular phase viscosities and granular conductivities. This new kinetic theory predicted flow behavior and oscillations in a complete loop of a CFB. The results were compared to computations with imposed gas phase turbulence in the riser. The computations were repeated for production of synthesis gas from char.

This Topical Report on Shale Oil Dedusting'' presents the results of a research program conducted by the Illinois Institute of Technology (IIT, Chicago) to determine the suitability and effectiveness of the lamella electrosettler -- a novel solid-liquid separation device -- for removing fine shale particles from shale oil via the application of an electric field. The work was conducted by IIT from November 1989 through December 1990 as a subcontractor to the Institute of Gas Technology. The overall objective of the larger program was to develop the Pressurized Fluidized-Bed Hydroretorting (PFH) Process for EasternOil Shales.'' The subtask undertaken by IIT was part of a larger task entitled Testing of Process Improvement Concepts.'' The lamella electrosettler has been shown to be an effective method for separating fine particulate (including colloidal) matter from a liquid using the application of an electric field. Using the walls of the settler as electrodes and during continuous operation, solids migrate preferentially toward one of the electrodes and become concentrated in the refuse stream. The product stream is clarified of particulates. The success of the process depends upon the physical properties of the solids and liquids being tested. A sample with a high specific conductance is notmore » suitable for separation in the lamella electrosettler. The liquid begins to heat up under the influence of the electric field and, eventually, may short. Also, under these conditions, the particles cannot maintain a charge. The high conductivity of the shale oil samples tested rendered them unsuitable for further testing in the lamella electrosettler.« less

This Topical Report on ``Shale Oil Dedusting`` presents the results of a research program conducted by the Illinois Institute of Technology (IIT, Chicago) to determine the suitability and effectiveness of the lamella electrosettler -- a novel solid-liquid separation device -- for removing fine shale particles from shale oil via the application of an electric field. The work was conducted by IIT from November 1989 through December 1990 as a subcontractor to the Institute of Gas Technology. The overall objective of the larger program was to develop the ``Pressurized Fluidized-Bed Hydroretorting (PFH) Process for EasternOil Shales.`` The subtask undertaken by IIT was part of a larger task entitled ``Testing of Process Improvement Concepts.`` The lamella electrosettler has been shown to be an effective method for separating fine particulate (including colloidal) matter from a liquid using the application of an electric field. Using the walls of the settler as electrodes and during continuous operation, solids migrate preferentially toward one of the electrodes and become concentrated in the refuse stream. The product stream is clarified of particulates. The success of the process depends upon the physical properties of the solids and liquids being tested. A sample with a high specific conductance is notmore » suitable for separation in the lamella electrosettler. The liquid begins to heat up under the influence of the electric field and, eventually, may short. Also, under these conditions, the particles cannot maintain a charge. The high conductivity of the shale oil samples tested rendered them unsuitable for further testing in the lamella electrosettler.« less

The hydrodynamics of fluidization for two thin ''two-dimensional'' fluidized beds, one containing a centrally located rectangular obstacle and the other containing three square tubes roughly approximating a portion of a fluidized bed combustor (FBC), was solved using the FLUFIX/MOD2 computer program. The monolayer energy dissipation erosion model extended to include solids viscous stress was compared with the Finnie erosion model. Hydrodynamic and erosion convergence studies were performed for the first bed. Computed erosion rates from the monolayer energy dissipation erosion model compare more favorably for both beds with available erosion data than those from the Finnie erosion model. 27 refs., 15 figs.

A Green's matrix for a one dimensional diffusion equation for composite media with Neumann type boundary conditions has been constructed using orthogonal functions. This Green's matrix has been used to construct a solution to a two dimensional steady-state composite media flow problem using multiplicative properties of the problem. The analytic solution is applied to the problem of fluid flow in a fuel cell gas compartment with distributors near the inlet and outlet ports. It is useful in the design of distributors for a better purge efficiency. The Green's function is also useful in various problems of heat and mass transfer.

A method of transpiration heating which offers a means for reducing heat loss and controlling temperature for a variety of fuel cell systems is described. The theoretical temperature distribution in a molten carbonate fuel cell is plotted for various gas flow rates and power outputs of the fuel cell. (WHK)

The object of this investigation was to develop an experimentally verified hydrodynamic model to predict solids circulation around a jet in a fluidized bed gasifier. Hydrodynamic models of fluidization use the principles of conservation of mass, momentum and energy. To account for unequal velocities of solid and fluid phases, separate phase momentum balances are developed. Other fluid bed models used in the scale-up of gasifiers do not employ the principles of conservation of momentum. Therefore, these models cannot predict fluid and particle motion. In such models solids mixing is described by means of empirical transfer coefficients. A two dimensional unsteady state computer code was developed to give gas and solid velocities, void fractions and pressure in a fluid bed with a jet. The growth, propagation and collapse of bubbles was calculated. Time-averaged void fractions were calculated that showed an agreement with void fractions measured with a gamma ray densitometer. Calculated gas and solid velocities in the jet appeared to be reasonable. Pressure and void oscillations also appear to be reasonable. A simple analytical formula for the rate of solids circulation was developed from the equations of change. It agrees with Westinghouse fluidization data in a bed with a draft tube.more » One dimensional hydrodynamic models were applied to modeling of entrained-flow coal gasification reactors and compared with data. Further development of the hydrodynamic models should make the scale-up and simulation of fluidized bed reactors a reality.« less

The overall objective of this investigation is to develop experimentally verified models for circulating fluidized bed (CFB) combustors. The purpose of these models is to help American industry, such as Combustion Engineering, design and scale-up CFB combustors that are capable of burning US Eastern high sulfur coals with low SO{sub x} and NO{sub x} emissions. In this report, presented as a technical paper, solids distributions and velocities were computed for a PYROFLOW circulating fluidized bed system. To illustrate the capability of the computer code an example of coal-pyrite separation is included, which was done earlier for a State of Illinois project. 24 refs., 20 figs., 2 tabs.