PTQ Q3 2023 Issue

As mentioned in Part I of this article, the operating pressure and temperature of distillation columns used to separate LAO products are constrained by the thermal deg- radation of LAOs that takes place at elevated temperatures in the column bottoms and column reboilers. Minimising the risk of thermal degradation became a primary objective in selecting column operating conditions for the simulations. Very little information has been published about the ther- mal degradation temperatures of LAOs produced by ethyl- ene oligomerisation. One published patent document, US 6271434, makes references to ‘preferred operating tem- peratures’ ranging from 210°C to 280°C (410°F to 536°F) and ‘more preferred operating temperatures’ ranging from 230°C to 270°C (446°F to 518°F) for separations equip- ment associated with ethylene oligomerisation processes.² Further complicating the issue of selecting column operat- ing conditions for this study is the steady increase in molec- ular weight of the bottoms products from each column as one progresses downstream in the series of columns. The increasingly higher molecular weights of column bottoms require an increasingly deeper vacuum to be maintained in distillation columns to stay within the temperature ranges cited in US 6271434. Since the cost of achieving ultra-low vacuum levels throughout the entire series of columns is prohibitive, column operating pressures were allowed to cascade downward from column to column. Column oper- Large throughput increases using PPD schemes are more likely to make economic sense on revamp projects with strong economic drivers ating pressures for the base case were selected so that the bottoms temperatures in the series of columns started at the low end of the temperature ranges cited in US 6271434 in the upstream columns and gradually increased to the upper end of the cited temperature ranges as one progresses downstream in the distillation train. Once column operating pressures were selected for the base case, which represents the original mode of operation for the series of distillation columns, the same operating pressures were applied to the PPD simulation case. The logic for maintaining the same operating pressures for both simulation cases is that minimal changes to the ancillary equipment (heat exchangers and vacuum systems) associ- ated with the existing distillation columns would be required as part of the conversion to prefractionation service. This is the case if the same operating pressures used in the base case were determined to be suitable in the PPD case. Simulation results for large throughput increase The feasibility of using PPD to expand the capacity of an existing LAO distillation train was validated by comparing the results of the PPD simulation with those of the base case. The objective of the evaluation was to determine how much the LAO product recovery capacity could be

expanded without exceeding the hydraulic capacity limits of existing distillation columns as defined in the base case. In the early stages of the development of process simula- tions, an attempt was made to identify cost-effective PPD configurations in which all LAO product specifications were met and all of the distillation columns operated within their existing hydraulic capacity limits. Column profiles for the PPD case distillation columns were compared against the base case column profiles to ensure that individual column hydraulic capacity limits were not exceeded. Column bot- toms temperatures were also checked for PPD designs to ensure that the operating conditions did not significantly increase the risk of excessive thermal degradation of LAO products in a PPD scheme. It quickly became apparent that two different levels of debottlenecking could be obtained in the C12+ section of the product recovery system through the use of PPD. Product pair distillation schemes with two different levels of debottlenecking could be obtained because the conver- sion of existing columns to a PPD design produces une- qual throughput increases in the set of revamped series of columns. In general, the first and last columns in a PPD revamp design provide the smallest throughput increases, while the remaining columns provide the highest increases. In a scheme in which all existing columns are reused in a conversion to a PPD design, the net throughput increase is limited by the column(s) with the smallest observed increase in throughput. This column (or columns) can be considered the new bottleneck in the PPD revamp con- figuration. Simulations demonstrated that a throughput increase of 27% was obtained for this scenario, which will be referred to as a ‘small throughput increase’. Large throughput increases can be obtained in a PPD scheme by replacing the prefractionation column(s) that limit the net increase in throughput. In the case study example for a large throughput increase, a net throughput increase of 68% was obtained by replacing the first and last columns in a series of five columns in the C12+ section of the product recovery system. The distinction between small and large throughput increases is important because each option lends itself particularly well to specific categories of debottlenecking applications. PPD designs for large throughput increases are best suited to plant capacity expansions in the range of 50% to 70%. Large throughput increases obtained using PPD schemes are more likely to make economic sense on revamp projects with strong economic drivers, such as large plant capacity expansions. PPD designs for (small) throughput increases in the range of 20% to 30% are especially well suited for product yield distribution shift applications and for increased prod- uct purity applications. As discussed in Part I of this article, in many cases a section of the product recovery distillation train can be debottlenecked by converting two or more existing columns to prefractionation service and adding a single new distillation column. An example of such a case in which the first three columns in the C12+ section of the LAO product recovery system are debottlenecked is shown in Figure 3 .

PTQ Q3 2023