PTQ Q3 2023 Issue

1 requires nine new columns. The PPD design shown in Figure 2 requires only six new columns. The net reduction in new columns through the use of the PPD scheme pre- sented is three columns. The use of a DWC in the C4 to C10 recovery section accounts for a net reduction of one column, and the application of PPD to the C12+ recovery section accounts for a net reduction of two columns. Energy efficiency The energy usage of the C12 to C20-C24 recovery col- umns is shown in Tables 3 and 4. Energy usage of the C4 to C10 recovery columns is not presented because, in a large throughput increase, there is a negligible difference between the base case and the PPD case (a small energy reduction amounting to 0.7 MM Btu/hr is obtained by sub- stituting a DWC for the DeC6 and DeC8 columns). In the base case, the total reboiler duty in the C12+ recovery sec- tion for the sum of the existing distillation train (350k met- ric t/y) and for adding a new parallel distillation train (239k metric t/y) is 17.58 MM Btu/hr. The total reboiler duty for the PPD case (589k metric t/y) is 18.54 MM Btu/hr. At first glance, it appears there is a small energy penalty associated with PPD designs vs conventional single-prod- uct distillation of LAO products. However, the total reboiler duty of the PPD case assumes that 100% of the product pair volume separated in the prefractionation columns is sent to the main columns for separation into individual LAO products. It is unnecessary to separate the entire produc- tion of product pairs into individual LAO products because a portion of the product pairs can be diverted to product storage for sale as LAO product blends (such as C12/C14 and C14/C16). A simple calculation shows that if 15% of the total volume of product pairs is sent directly to product storage, the total reboiler duty for the PPD case drops from 18.54 MM Btu/hr to 17.46 MM Btu/hr. This leads to the conclusion that, in most cases, the energy impact of using PPD designs instead of conventional single-product distil- lation is not likely to be significant. The use of complex main column designs can further improve the energy efficiency of PPD designs. A new sim - ulation was developed to determine how efficient a single five-product main column design would be in comparison to the three-product main column design shown in Figure 2 for separating C12/C14, C14/C16, C16/C18, and C18/C20- 24 intermediate products pairs. A remarkable reduction of 54% in reboiler energy usage (3.3 MMBtu/hr reboiler duty for a single main column design vs 7.2 MMBtu/hr combined reboiler duty for a design with two main columns) was obtained using a single main column design. The remarkable energy benefit of using a single main col - umn design must be weighed against the drawbacks of a single main column design. A single five-product main col - umn is larger and more costly than a three-product column because the five-product main column requires significantly more theoretical stages than each of the two three-prod- uct columns shown in Figure 2. A single five-product main column also presents a significantly more complex process control challenge than the three-product main column con- figuration shown in Figure 2.

Comparison of product yields

Base case yield

PPD case yield

589k metric t/y, lb/hr

C4 C6 C8

20,588 21,798 20,158 17,627 14,880 12,082

20,588 21,803 20,153 17,627 14,933 12,037

C10 C12 C14 C16 C18

9,688 7,453

9,723 7,398

C20-C24

14,489

14,461

C26+ Total

9,432

9,472

148,196

Table 5

are observed in the DeC20-C24 column, and the bottoms temperature for the DeC20-24 column is identical for the base case and PPD case. It seems unlikely that the slight increase in column bottoms temperatures observed in the prefractionation columns in the PPD case will have a major impact on the overall rate of thermal degradation in the dis- tillation train after conversion to PPD. A comparison was also made for the final LAO product yields. The results in Table 5 show that very similar LAO prod - uct yields are obtained for the base case and the PPD case. Figure 2 provides an overview of the major scope required to implement a PPD design for a 68% throughput increase. Columns in black represent existing columns reused as-is (former DeC14, DeC16, and DeC18 columns). The column shown in dashed lines represents the existing DeC6-C10 column converted to DeC12/14 service that requires mod- ification to major associated equipment (column vacuum system). Columns in red represent new equipment. A comparison of the scope required to achieve a 68% throughput increase by the conventional method of install- ing a new parallel distillation train to duplicate the distil- lation train vs implementing the PPD design developed in this case study can be made by comparing Figures 1 and 2. Replacing the entire distillation train shown in Figure

Benzene product

Toluene product

MC1

Xylene product

B/T

T/X

Combined xylene product

Xylene product

BTX feed to PPD conguration

C9+ product

Existing x ylene column

DeT/X (formerly t oluene column)

DeB/T (formerly benzene column)

Figure 4 PPD distillation scheme for BTX separation

PTQ Q3 2023