(159d) Radial and Axial Profiles of Solids Loading in a Gas-Solid Circulating Fluidized Bed

Numerous chemical processes are run in circulating fluidized beds (CFBs), including fluid catalytic cracking, combustion, and many others. This paper presents the results of an experimental program in which multiple diagnostic techniques were developed and applied to the flow in a CFB, providing solids volume fractions and their radial and axial distributions.

A pilot-scale gas-solid CFB facility, shown schematically in Figure 1, was designed and fabricated. Solids are fed from the 28-cm inside-diameter (ID) downcomer column through a metering valve and a standpipe into the riser engagement section. The annular engagement section at the riser's base has a fluidized bed surrounding a central 8.5-cm diameter air supply pipe. Motive air entrains particles from the fluidized bed and transports them up the 14-cm ID riser column to the particle disengagement section. The motive air exits the top of the disengagement section through cyclone separators, which return particles to the downcomer. The air and any remaining particles exit the cyclones and are vented to atmosphere through a HEPA filter baghouse. Fluidization air is supplied at locations throughout the CFB.

The riser has a total uniform length of 5.77 m, or an aspect ratio of L/D ~ 41. This length measurement excludes the 54.6-cm high engagement section, the top of which defines the axial origin z = 0. The annular design of both the engagement and disengagement sections ensures that the flow in the vertical riser is as axisymmetric as possible. The riser is extensively grounded and the inlet air humidified to reduce triboelectric effects.

The riser is loaded with equilibrium fluid catalytic cracking (FCC) catalyst particles of density 1250 kg/m³ and Sauter mean diameter 65 microns. Solids flux is measured using a diverter valve section that allows fast capture, weighing, and return of particles to the system.

Differential Pressure (DP) measurements are made by instrumenting the flow loop using electronic transducers. Sixteen transducers are installed at 30.5-cm intervals along the riser. Reference pressures are acquired at one location in the riser, at the tops of the disengagement and downcomer, and on the air supply and outlet lines. Sintered metal discs (10-micron pore size) protect the transducers from contamination. These in-line filters limit the frequency response of the transducers to about 1 Hz. The DP signals are converted to volume-averaged solids loading using the hydrostatic assumption.

The Gamma Densitometry Tomography (GDT) system consists of a 100-mCi ¹³⁷Cs source and an array of 8 NaI(Tl) scintillation detectors. The source produces a fan-shaped beam that passes through the riser to the detector array, where the gamma intensity along each distinct ray is measured. The source and detectors are mounted on a vertical traverse that allows measurement along the riser axis, and the detectors are mounted on a traverse that allows lateral displacement of the detectors, with the source fixed, to achieve improved spatial resolution.

Attenuation of monoenergetic gamma photons is given by I = I_oe^-mL, where I is the measured intensity, I_o is the unattenuated ?empty? intensity, m is the attenuation coefficient, and L is the path length through the solid material. For measurement of multiphase mixtures for which the attenuation coefficient m is not known, the amount of attenuating material in the beam path can be determined by using a ratio between empty and full measurements. For these experiments the values of measured, full, and empty intensity were taken from the peak intensity region of the energy spectrum measured using a multichannel analyzer.

A generalized Abel transform (Shollenberger, 1997) is used to convert the path-averaged solids volume fraction into a radial solids volume fraction profile in the circular domain. Figure 3 includes representative radial profiles of solids fraction measured using GDT.

A 16-electrode Electrical Impedance Tomography (EIT) system was developed and is described in detail by Tortora (2004). Electrodes are supplied with a 100 kHz, 5 V driving frequency, and impedance is measured between electrode pairs.  Reconstruction is performed using an optimization code to determine the best values of local solids fraction to yield the measured impedances, using the Rayleigh mixture model to relate impedance to solids fraction. Figure 3 includes representative radial profiles of solids volume fraction measured using EIT.

RESULTS

Table 1 lists the four flow conditions examined in this work. Superficial gas velocity and solids flux were varied. Figure 2 shows representative radial profiles of solids volume fraction for each flow condition measured at z/D = 12. As is common in these flows, a core-annular flow structure is indicated, with higher solids volume fraction at the walls and lower in the center.  

Figure 2 shows axial profiles of the volume-averaged solids volume fraction determined by GDT, EIT, and DP for each of the four flow conditions given in Figure 3.

The DP profiles in Figure 3 were constructed by picking the DP points recorded simultaneously with the GDT data and interpolating between the nearest two axial DP locations. Conditions (U_g and G_s) were held nominally constant within each series of runs. Each of the four runs shown in Figure 3 was run over a several day period. Run conditions were always brought back to the same nominal settings.

Figure 2. Radial profiles of solids loading as measured by GDT and EIT methods. Flow conditions are given in Table 1.

Figure 3. Axial profiles of solids loading as measured by DP (DP in legends), GDT, and EIT methods. Flow conditions are given in Table 1.

DISCUSSION

GDT, EIT and DP-determined solids loadings are shown in Figure 3. DP is volume averaged, while GDT and EIT are area-averaged. For the purposes of this comparison, all were time-averaged for five minutes. The data of Figure 3 indicate that the DP-determined values are higher than the GDT and EIT values low in the riser. This is as predicted by Louge and Chang (1990), since the flow is not fully developed near the base of the riser resulting in significant gradients in solids loading.

The data show similar trends to those of Schlichthaerle and Werther (1999), even though the present solids fluxes are much higher and the axial region scanned extends much further up the riser. The Schlichthaerle and Werther data indicate that even at low solids flux the effect of solids loading gradients is still important at the base of the riser. Louge (1990) also presents data showing similar behavior at even higher solids fluxes (up to 600 kg/m²·s).

CONCLUSIONS

Experiments were performed to compare the solids volume fraction measured by GDT, EIT, and DP techniques over a range of CFB operating conditions. The GDT and EIT results show good agreement for both radial and axial solids-volume-fraction profiles. The present data support the analysis of Louge (1990) and the experimental data of Schlichthaerle (1999) and extend this comparison to higher solids fluxes and axial distances along the riser. The DP technique is complicated low in the riser by the effects of solids flux and gradients of solids volume fraction and thus overpredicts the true solids loading. Higher in the riser the DP data come into general agreement with those of GDT and EIT.

ACKNOWLEDGEMENTS

Sandia is a multiprogram laboratory operated by Sandia Corporation, a Lockheed Martin Company, for the United States Department of Energy's National Nuclear Security Administration under Contract DE-AC04-94AL85000. The authors acknowledge the support of the U. S. Department of Energy Industrial Technologies Program, Brian Valentine, contract sponsor (for Sandia), and the National Science Foundation, Cyrus Aidun, contract sponsor (for U. Michigan). We also acknowledge the assistance of John Oelfke for his technical support and Kim Shollenberger for the gamma system setup and early implementation.

REFERENCES

Louge, M., and Chang, H., 1990, ?Pressure and Voidage Gradients in Vertical Gas-Solid Risers,? Powder Technology, 60, 197-201.

Schlichthaerle, P., and Werther, J., 1999, ?Axial Pressure Profiles and Solids Concentration Distributions in the CFB Bottom Zone? Chemical Engineering Science, 54, 5485-5493.

Shollenberger, K. A., Torczynski, J. R., Adkins, D. R., O'Hern, T. J., and Jackson, N. B., 1997, ?Gamma-Densitometry-Tomography of Gas Holdup Spatial Distribution in Industrial-Scale Bubble Columns,? Chemical Engineering Science, 52, 2037-2048.

Tortora, P. R., 2004, ?Electrical-Impedance Tomography for the Quantitative Measurement of Solids Distributions in Gas-Solid Riser Flows,? Ph.D. dissertation, University of Michigan, Mechanical Engineering Department.