Case Study for Methyl Acetate Production

We shall illustrate some of the tray hardware design issues by considering a case study of the production of methyl acetate in which the reaction is catalyzed by H2SO4, added to the liquid phase [1]. In the RD process for methyl acetate, invented by the Eastman Chemical Company [3, 12], the entire process is carried out in a single column, as shown in Fig. 7.6. In this single column, high-purity methyl acetate is made with no additional purification steps and with no unconverted reactant streams to be recovered. By flashing off the methyl acetate from the reaction mixture, conversion is increased without using excess reactants. The reactive column has stoichiometrically balanced feeds and is designed so that the lighter reactant MeOH is fed at the bottom and the heavier acetic acid is fed at the top. The column consists of three sections. The reaction takes place predominantly in the middle section. The bottom section serves to strip off the MeOH from water and return it to the reaction zone. The vapors leaving the reactive section consists of the MeOAc-MeOH azeotrope, which is 'broken' in the rectifying section by addition of AcOH, which acts as entrainer. The RD column represents an entire chemical plant, costs one-fifth of the capital investment of the conventional process, and consumes only one-fifth of the energy. Following the design study of Doherty and Malone [1], we take each of the molar feed flows of MeOH and AcOH to be 280 kmol/h, on the stages 5 and 40. The reflux ratio is set at 1.9. The column operates at a pressure of 1 atm. We employed a partial reboiler and a total condenser. The non-reactive section contains 10 theoretical stages and 33 catalytically (homogeneous) active theoretical stages in the reactive section. Simulations of the liquid composition profiles, using the EQ stage model are shown in Fig. 7.6.

We set the column specifications, and the desired conversion and purities of MeOAc and water, to a value of 98.5 %. These specifications are only fulfilled for a fairly small range of reboil ratios: see calculations in Fig. 7.7a. For low reboil ratios, the reactants cannot be recycled efficiently and high water purities in the

Fig. 7.6 The Eastman RD process for methyl acetate manufacture. Adapted from Agreda et al. [4], The column composition profiles sketched on the right were obtained with an EQ model

bottom stream cannot be reached. On the other hand, if the reboil ratio is too high then the residence time in the reactive section decreases, reducing conversion. The sensitivity of the AcOH conversion to volumetric hold-up at each stage, Vstg, is shown in Fig. 7.7b.

If the value of Vstg is less than about 2 m3, it is impossible to achieve the target AcOH conversion of 98.5 %. For a reboil ratio of, say, 1.4 it is necessary to ensure a Vstg of at least 2.75 m3. Let us examine the hardware design factors that would meet this hold-up requirement. Fig. 7.8a presents the calculations for the clear liquid height and the volumetric liquid hold-up Vstg for a 5.7 m diameter column, operating at a range of superficial vapor velocities. We see from Fig. 7.8a that to achieve a value Vstg = 2.75 m3, we need to operate at a superficial vapor velocity as low as 0.27 m/s, keeping the weir height at a high value of 140 mm. The operation of the reactive section is in the bubbly flow regime. For the non-reactive section at the top of the RD column, there is no corresponding requirement of liquid hold-up and a much higher superficial vapor velocity of 1.27 m/s can be chosen, along with a smaller column diameter of 2.8 m and a lower weir height of 80 mm: see Fig. 7.8b. The operation of the column is close to the transition between froth and spray regimes.

Fig. 7.7 a) Acetic acid conversion and purities of MeOAc and water in leaving product streams as a function of reboil ratio. The volumetric liquid hold-up on each reactive stage Vstg is set at 3 m3. b) Acetic acid conversion as a function of the reboil ratio for various liquid hold-ups on each reactive stage

(a) Conversion and purity constraints 100 r

0 0

Post a comment