Conference Dates

May 22-27, 2016


The present work proposes the determination of solids circulation rate in a freely circulating fluidized bed cold model by using a method that performs quantitative detection of fluid-dynamically similar ferromagnetic tracer particles (1). The method has already been proven for determination of particles residence time distribution with encouraging results (2). The method is based on inductance changes of a coil by effect of changes in the concentration of ferromagnetic particles in its core (Fig. 1). There is a direct correspondence between measured inductance and proportion of ferromagnetic material in the coil’s core. The tracer particles are fluid-dynamically consistent, the injection and sensing methods are non-invasive, the inductance signal has a high resolution and the measurement is performed continuously.

A cold flow model of a dual circulating fluidized bed (3) has been used to test the idea. The precise location of injection and sensing devices is presented in Fig. 2; this is the configuration proposed for the determination of the global circulation rate. A tracer injection pulse is assumed to be a perfect Dirac-delta function, and the response signal is measured at the outlet of the loop seal. These two signals are processed to determine the normalized residence time distribution curve, E(t), of the particles when flowing thought the loop seal (Eq. 1). Considering that the density of the bulk of particles is constant, for the mean residence time, Eq. 2 is valid. Eq. 3. indicates the variance of E(t), it is, the degree of dispersion around the mean.

Under the assumption that the loop seal is maintained at conditions of minimum fluidization, regardless of the circulation rate and the conditions inside the reactors, the pressure difference is given by Eq. 4, which in turn allows the correlation between global circulation rate and mean residence time as shown in Eq. 5. At higher circulation rates and/or high accumulation of solids in the downcomer, the fluidization conditions might diverge from minimum fluidization, and the pressure drop per length unit might vary. This variation is proportional to the change in the variance of the residence time distribution curve, and can be quantified.

Please click Additional Files below to see the full abstract.