Ith

outlet e in II

Ozone

Water inlet

Water inlet

Ozone

100%

FIGURE 11.20 Initial (a) and modified (b) configurations of industrial ozonation reactor (from Cockx et al., 1999).

Ozone

100%

FIGURE 11.20 Initial (a) and modified (b) configurations of industrial ozonation reactor (from Cockx et al., 1999).

ozonation reactor are shown in Fig. 11.20. The proposed modifications led to better hydraulics and a significant increase in ozone transfer efficiency.

The second application deals with simulation of methanol synthesis in gas-liquid slurry bubble column reactors, discussed by Wu and Gidaspaw (2000). Most slurry bubble columns are designed using one-dimensional models with empirical correlations of gas volume fractions and other necessary fluid dynamic characteristics. Most of these one-dimensional models assume uniform solid (catalyst) volume fraction over the reactor. Wu and Gidaspaw (2000) developed a detailed hydrodynamic model by considering a two-dimensional solution domain. The hydrodynamic model was coupled with the reaction model to simulate overall performance of the methanol synthesis reactor. The comparison of their predicted results and experimental data is shown in Table 11.2. The agreement between predicted results of such a detailed two-dimensional model with experimental data was much better than that observed with predictions of a one-dimensional model. Apart from predicting the realistic values of conversion, the detailed hydrodynamic model provides valuable information on flow structures and their sensitivity with respect to design and operating parameters. Such information is essential for guiding further developments in syn-gas to liquid fuels technology.

As discussed earlier, coalescence break-up models incorporated in detailed CFD models will allow accurate simulation of interfacial area and corresponding mass and

TABLE 11.2 Methanol Synthesis Reactor: Comparison of simulations and experimental data (from Wu and Gidaspaw, 2000)

CO Gas Slurry Total Methanol,

Conversion, % hold-up, % height, in catalyst, kg gmol h-lkg-1

Simulation 14.24 26.9 215 740 16.93

heat transfer. It may, however, be necessary to calibrate the parameters of coalescence and break-up models with the help of experimental data before the combined models can be used to estimate mass transfer performance. Even in the absence of adequate experimental data for calibration purposes, detailed CFD models can be used to qualitatively evaluate different configurations and can identify the most promising sparger and internal designs. The design of feed pipe and product outlets can be evaluated as well, as illustrated by the ozonation reactor case discussed here. Computational flow models and CFD tools play a crucial role in linking actual reactor hardware to reactor performance. These models allow extrapolation of cold flow results to actual operating conditions (high temperature and pressure) and provide detailed information and insight into the reactor flow field. In general, the reactor engineer has to ensure that the computational model contains adequate basic physics, that the numerical implementation is well within the set tolerances, and that simulations capture all the relevant flow features. Judicious use of such computational flow models, (1) to help understand basic phenomena and (2) to simulate complex industrial reactors (using a hierarchy of modeling layers), will lead to better engineering of bubble column reactors.

0 0

Post a comment