Simple Reactive Distillation Design

REACTIVE DISTILLATION DESIGN USING COLUMN PROFILE MAPS 

As with the CPMs, permitting the analysis of negative compositions aids in understanding the overall process. Due to the nature of the reaction rate model (Equation 8.1), additional nodes have been introduced to the composition space. For this particular example, there are in fact five different stationary points, but only four can be identified here due to the scope of our view of the composition space. As with azeotropic systems which also contain additional stationary points, both inside and outside composition space, the presence of additional nodes can give one more freedom or options in finding a suitable design. 

Example 8.1 In the reaction system lA + 1C - 3B and 15, we would like to determine whether it is possible to design a batch reactive distillation system that will produce acomposition of x esired = [0.02,0.93]. It has been predetermined that the feed to the system should be xf = [0.4,0.2],the relative volatilities to the systemaiea= [a, c = [6,1,3], and the reaction rate is elementary. If the stem is indeed possible, produce the relevant R-RCM-M and comment on the system s behavior. 

Solution 8.1 In the problem statement all data has been provided, excq t for a reactor size variable (k) and vapor boil-c f rate variable (fi). Since we are interested in producing a particular product with a composition of Xdesiicd = [0.02,0.93], we can calculate through the steady state version of Equation 8.10 d /d( = 0) what values of K and P will result in a pinch point at Xogsmd- vapor in equilibrium with x esired 

---

We now consider the design of industrial DCRs. First, we consider a simple design without going into details of the role of distillation boundaries for both straight distillation and reactive distillation. This will be followed by a mathematical analysis of these factors. 

Despite strong nonlinearity (Figs. 13.3 and 13.4), workable temperature control of these reactive distillation systems can be obtained using a systematic design procedure with rather simple control structures. It should be emphasized that the selections of controlled and manipulated variables (control structure design) play a cmcial role for these highly nonlinear processes and that decentralized control provides a better stmcture to cope with steady-state gain variations than model-inversion-based control stmctures. 

---