A View of Reactive Distillation Process Design

We close with some comments on the use of these models in RD process design. There are many different types of model now available in the literature for screening, analysis, design, and optimization of RD columns. Below is an outline of a stepwise approach to RD process design:

1. feasibility of RD concept,

2. preliminary design,

3. choose column hardware,

4. equipment sizing,

5. check preliminary design using NEQ models,

6. process dynamics and control.

Determining the feasibility of a process (step 1) is a relatively straightforward task that can be carried out using residue curve maps [2, 3]. Equilibrium stage models are useful for the preliminary design (step 2). This involves determination of the number of theoretical stages in the reactive, stripping, and rectifying sections; the catalyst load in the reactive section(s); and some optimization of operating parameters such as the reboiler load and reflux ratio [2, 3].

The design of the equipment for the non-reactive sections should not pose too great a challenge since existing methods of estimating efficiencies or HETP that are based on key-components could be used. For the reactive section, however, the estimation of HETPS and efficiencies is more problematic. There is no easy way to identify key components. Chemical reactions influence component efficiencies in an unpredictable manner (Fig. 9.22 in [1]). Of course, like the rest of the RD community, we could behave like ostriches and bury our heads in the sand while assuming equal component efficiencies (say, 70 %) for individual species and hope for the best!

Column hardware choice can have a significant influence on the conversion and selectivity; such aspects can be properly described only by the NEQ cell model, or by a still more sophisticated model based on computational fluid mechanics (such models have yet to be developed). It is insufficiently realized in the literature that, say, for tray RD columns, the tray design can be deliberately chosen to improve conversion and selectivity. Even less appreciated is the fact that the design methodology for RD tray columns is fundamentally different from that of conventional trays. Liquid residence time and residence time distributions are more important in RD. The froth regime is to be preferred to the spray regime for RD applications; this is opposite to the design wisdom normally adopted for conventional distillation. For relatively fast reactions, it is essential to properly model intra-particle dif fusion effects. Pseudo-homogeneous reaction models may be inadequate for fast reactions. RD columns using dumped (random) packing are susceptible to maldistribution and there is a case to be made for choosing regular structured packing. For proper description of the column dynamics, it is essential to adopt the NEQ model.

Though sophisticated NEQ design models are available already, detailed information on the hydrodynamics and mass-transfer parameters for many available hardware configurations is woefully lacking in the open literature. For some internals types the information available is contradictory, for still others there is no masstransfer data at all. Paradoxically, such information has vital consequences for the conversion and selectivity of RD columns. There is a crying need for research in this area. It is perhaps worth noting here that modern tools of computational fluid dynamics could be invaluable in developing better insights into hydrodynamics and mass transfer in RD columns.

Besides more research on hydrodynamics and mass transfer, there is a need for more experimental work with the express purpose of model validation. In such process studies, parameters need to be measured along the height of RD columns. Too often measurements are confined to feed and product stream conditions. Such data cannot serve as a reliable discriminant of computer-based process models.

0 0

Post a comment