STUDIES ON THE RESIDENCE TIME DISTRIBUTION OF SOLIDS IN A SWIRLING FLUIDIZED BED

The Multi-Parameter Two-Layer (MPTL) mathematical model was developed in this work specifically to model the Residence Time Distribution (RTD) of particles In
a continuous system of swirling fluidized bed reactor. The model consists of parallel layers. The top layer is a stirred tanks-in-series model and represents the
conventional fluidized bed. Meanwhile, the bottom layer obeys the general recycle model and represents the swirling motion at the bottom layer of the bed. The Laplace
transformation and convolution integral techniques are used to derive explicit fiu the RTD functions of the stined tanks-in-series model and general
r e c ~ dde l .Th e pro& model has six indepmdindependent parameters - recycle fraction(PI, M elaye r flow mte fiaction (w), recycle layer volume iiaction ( y, ), number
of tanks in the main flow line of the recycle layer (q), number of tanks in the recycle line (n,) and number of tanlts in the top layer (n,). The RTD experiments were
amduckd at difbea particle sizes and bed weights. TZle bed material used in the eqmkm@l wo& is spherical plastic beads with a diameter dp = 2.99mntatuid, =3.85anr. Duting hydmdymnics study, it is found that bed pressure & ~ p
w, increases with air velocity and bed weight. Besides, the smaller bed particle gives a higher pressure drop for a given bed. The effects of parameters on the RTD
function ~ ( 0a)re studied and the model is shown to be highly versatile and capable of representing widely different mixing conditions depending on the system variables, By best-fitting of the model response to the experimental data, the model parameters can be evaluated. The experimental result of solid RTD shows that the bed
performance varies from one-layer to two-layer bed as the bed weight increased,One-layer bed can be modeled by having number of stirred tanks n, = 4. p , ,an d y, ranging from 0.8 to 0.83, 0.9 to 1.0 and 0.75 to 1.0 respectively. For two-layer bed, it is found that the combination of n, = n, = n, = 5 can fit all the runs. The value of the model parameters P, w and y, ranging from 0.5 to 0.83, 0.2 to 1.0 and 0.52 to 1.0 respectively.