Coal Liquification


liquid fuels


petroleum fractions

desulferized fractions




Bio-Chemical Processes

Bio-Chemical Processes operation are possible for bubble column reactors. These bubble column reactors are also extensively used for gas-liquid-solid processes. Bubble column reactors provide excellent heat and mass transfer characteristics. Some of the important industrial applications are listed in Table 11.1. Similar to any reactor type, the engineering of bubble column reactors begins with the analysis of process requirements and evolving a preliminary configuration for the reactor. Before relating the process requirements and design of bubble column reactors, it will be useful to give a brief overview of the complex fluid dynamics of bubble columns. This will facilitate an appreciation of various design issues and the role of rigorous flow modeling in the design and scale-up of bubble column reactors.

In a bubble column reactor, gas is sparged at the bottom of the liquid pool contained by the column. The net liquid flow may be co-current or counter-current to the gas flow direction or may be zero. Large varieties of spargers or gas distributors are used in industrial practice to introduce gas in bubble columns. Sparger design controls bubble size distribution in the bottom portion of the bubble columns. Spargers, like porous plates, generate uniform size bubbles and distribute the gas uniformly at the bottom of the liquid pool. For such spargers, when gas superficial velocity is small (less than 2 cm s-1), all the bubbles formed at the sparger rise almost vertically. Larger (>0.2 cm diameter) bubbles may rise with inherent oscillations. This flow regime is called homogeneous. In this flow regime, macroscopic internal liquid circulation does not exist. The presence of bubbles may generate turbulence and affect the transport characteristics as described by Ranade and Joshi (1987). However, such an operating regime is unstable and even small perturbations can cause transition to a heterogeneous regime.

In a heterogeneous regime, significant bubble-bubble interactions occur and coalescence sets in to generate a wider bubble size distribution. Long-time averaging indicates that gas bubbles move towards the column center while rising upwards. Experimental data reported by Yao et al. (1991), measured using an ultrasonic Doppler technique, clearly show radially inward motion of gas bubbles (in a time-averaged sense). Such an inward motion and strong interphase coupling results in non-uniform gas hold-up profile, with maximum at the column center. This leads to strong macroscopic internal liquid circulation in the column with upflow in the central region and downflow in the near-wall region. Such internal re-circulation results in increased backmixing, which is one of the major drawbacks of bubble column reactors. Several internal designs like draft tubes, radial baffles etc. have been proposed to control the degree of backmixing in bubble column reactors. In recent years, inherently unsteady characteristics of gas-liquid flows in bubble columns have been studied (Chen et al., 1994; Delnoij, 1999). These studies indicate that capturing the unsteady flow structures may be essential for accurate description of mixing in bubble columns. Various factors such as type of sparger, column diameter, height to diameter ratio, physico-chemical properties, solid volume fraction (and other properties such as size and settling velocity) and operating conditions (pressure, temperature, and superficial velocities) affect the unsteady flow and mixing in bubble column reactors (Ranade and Utikar, 1999; Ranade and Tayalia, 2001). The presence of solid particles may further complicate the fluid dynamics of bubble columns. The fluid dynamic influence of solid particles depends on mean particle size, size distribution, particle density and solids volume fraction. The superficial gas velocity, resulting internal circulation, and solid particles interact in a complex way. In most design applications, empirical correlations and pilot-scale experiments on two or more scales are used to establish the relationship between these adjustable design and operating parameters and self-adjusting fluid dynamics and mixing in various phases.

A general procedure for the design and scale-up of reactors is discussed in Chapter 1. As discussed there, preliminary configurations are evolved on the basis of laboratory study and reactor models, which assume idealized fluid dynamics and mixing. Using idealized reactor models, various configurations and modes of operation are evaluated. In most industrial cases, this step itself may involve several iterations. The process of evolving a preliminary configuration helps to firm up performance targets for the reactor. Transformation of such a preliminary reactor configuration to an industrial reactor proceeds through several steps. Some of the relevant reactor engineering issues for bubble column reactors are summarized in Fig. 11.2. Some of these issues are discussed later, in Section 11.3. As discussed in Chapter 1, basic reaction engineering models based on approximations of the underlying fluid mechanics allow estimates of reactor size and reactor performance for different operating modes. Such studies are often used to select the type of reactor configuration and mode of operation (co-current/counter-current, upflow/downflow and so on). The influence of operating conditions on conversion and selectivity of the desired products can be examined using these reaction engineering models. These models are used to select feasible operating windows, and also to understand the sensitivity of reactor performance with a degree of backmixing, mass transfer coefficient, interfacial area, heat

FIGURE 11.2 Reactor engineering of bubble column reactors: relevant issues.

transfer, operating regime and so on. Such studies are useful to quantify key desired fluid dynamic characteristics of the reactor. These models are, however, unable to predict the influence of actual hardware details, as all the fluid dynamic information used in these models is usually based on empirical correlations.

Studies from reaction engineering models and basic process economics set the 'wish list' of the reactor, which may read something like the following: reactor hardware should (1) operate in a churn-turbulent regime over the specific gas flow rate range; (2) provide a certain minimum volumetric mass transfer coefficient and certain minimum heat transfer coefficient; (3) provide a certain minimum heat transfer area; (4) provide radially and axially uniform gas distribution; (5) provide adequate mixing to ensure that mixing time is less than the specified time; (6) entrainment of liquid droplets with the escaping gas phase should be less than the specified mass flow rate; and so on. In order to address these issues, it is necessary to develop comprehensive fluid dynamic models. If it turns out that some of the demands cannot be met with realistic reactor hardware, new iterations of studies with reaction engineering models are carried out, based on the information provided by the detailed fluid dynamics model of realistic reactor hardware. In some cases, it is possible to include detailed reaction engineering models within the CFD framework to develop a combined model. In some cases, for example, those involving very fast reactions, such combined models are mandatory for realistic simulations. However, even in other cases where it is usually more efficient to keep reaction engineering models and detailed fluid dynamic models separate, significant exchange of information takes place between these two types of model (see examples discussed in Chapter 9).

CFD-based models allow identification and quantification of the extent of non-idealities (such as bypass and channeling). These models also allow reliable extrapolation of results obtained on experimental and pilot scales. Apart from providing inputs to reaction engineering models, detailed CFD models establish relationships between hardware configurations such as sparger, internal baffles or draft tubes and resulting fluid dynamics and, therefore, with reactor performance. More often than not, the development of reactor technologies relies on prior experience. New reactor concepts are often sidelined due to lack of resources (experimental facilities, time, funding etc) to test them. Experimental studies have obvious limitations regarding the extent of parameter space that can be studied and regarding extrapolation beyond the studied parameter space. Computational flow models, which allow a priori predictions of the flow generated in a bubble column reactor of any configuration (different mode, spargers, internals) with just a knowledge of geometry and operating parameters, can make valuable contributions to developing new reactor technologies.

The fluid dynamics of bubble column reactors is very complex and several different CFD models may have to be used to address critical reactor engineering issues. The application of various approaches to modeling dispersed multiphase flows, namely, Eulerian-Eulerian, Eulerian-Lagrangian and VOF approaches to simulate flow in a loop reactor, is discussed in Chapter 9 (Section 9.4). In this chapter, some examples of the application of these three approaches to simulating gas-liquid flow bubble columns are discussed. Before that, basic equations and boundary conditions used to simulate flow in bubble columns are briefly discussed.

0 0

Post a comment