PTQ Q4 2024 Issue

simulation model with ideal stages, the 8.8% higher vapour flow rate from the equivalent model using the actual num - ber of trays with point tray efficiency could lead to overde - sign of the column diameter or unnecessary replacement of existing trays in a revamp column. The vapour-to-tray mass flow rate discrepancy from the ideal stage model to actual stage model can be reduced or eliminated by using the OTE feature, which is available in most commercial programs. Figure 3 shows the vapour-to-tray flow rates in ft3/sec generated from two similar simulation models in PetroSim Version 7.2 with the default PR package. These two mod - els, including the column specifications, are identical, except that one utilises 29 ideal stages and the other actual num - ber of trays/stages with 70% OTE, instead of the point tray efficiency. As shown, the two models result in essentially the same maximum vapour-to-tray flow rate in each of the high-traffic tray sections, with less than 1% deviation. This example shows that two methods – ideal stages and actual trays with OTE – can generate consistent vapour-to-tray flow rate data. The example crude column includes pump-around loops and side strippers, which are common in columns such as crude towers, fluid catalytic cracking (FCC) main fractiona - tors, and others. In this example, the PetroSim model, based on the actual trays and OTE, results in liquids from draw trays that are reasonably close to the bubble or saturation temperatures. The heavier draw liquids, such as light and heavy atmospheric gasoils, differ from the saturation tem - peratures by less than 1ºF, while the kerosene liquid draw differs by 2.7ºF. In a design or installation case where the pump-around pump suction pressure is limited, checking the simulated draw liquid saturation temperature and the associated vapour pressure can be worthwhile. C₃ splitter and extractive distillation The crude column example discussed so far involves pseu - do-components derived mainly from crude assays. Two additional column model examples with pure components, a C₃ splitter and an extractive distillation, are included to assess the tray loading data resulting from an ideal stage model and the alternate actual trays with OTE. The C₃ split - ter involving propane and propene is simulated with Hysys Version 12, PR package. One model uses 185 theoretical stages (plus the con - denser and the reboiler) and another 215 actual trays at 85% OTE. The models are otherwise identical. Comparing the simulated results in this example, the differences in the maximum vapour-to-tray and liquid from tray volumetric flow rates are respectively less than 0.03% and 0.1% in the rectifying section and 0.5% and 0.13% in the stripping section. These indicate the ideal stage and the actual trays with OTE options give essentially the same results. The extractive distillation example involving benzene, cyclohexane, and acetone is simulated with the Uniquac package in UniSim Version R492. Benzene and cyclohex - ane form an azeotrope, and acetone is used as the extrac - tive solvent and fed to the sixth stage (from top) of the 28 total ideal stages (excluding two stages for the condenser

Section 1 (Top) Max.882 ( i deal stages), 887 ( a ctual trays ft/sec

1.20E+03

1.00E+03

8.00E+02

Section 3 Max. 1108 ( i deal stages), 1118 ( a ctual trays ft/sec

Section 2 Max. 1040 ( i deal stages), 1049 ( a ctual trays ft/sec

6.00E+02

4.00E+02

PR Actual trays PR Ideal stage

2.00E+02

0.00E+00

29 i deal stages or 42 a ctual trays

Figure 3 Vapour flow rates based on actual trays – PR

agreement with the field data. However, as discussed ear - lier, the overall tray efficiency feature may not be directly available in all commercial programs. As stated in Ref. 3, using the point efficiency could lead to inconsistencies that need to be considered before the tray loading data is finalised. For the same crude column example with 41 actual trays (plus one bottom draw stage), the ProII column simulation model can be based on the actual number of trays, and the feed and draw trays can be specified in the model to exactly match the actual column. If 29 ideal stages are estimated in the crude column example, the corresponding OTE would be about 0.7 (or 29/41). As this OTE feature is not directly available, the user could likely specify one of the available point tray efficiency features. Figure 2 shows the vapour-to-tray flow rates calculated from the ProII model based on the default GS with the actual trays at Murphree efficiency of 0.7, which is assumed to be the same for all trays. The model is identical to the models for generating GS Ideal Stage data in Figure 1, except that actual trays with Murphree tray efficiency are used instead of the ideal stages. As shown in Figures 1 and 2, the differences in the maximum vapour-to-tray (or stages) flow rates are 1.6% and 0.4%, respectively, for the top section and the sec - tion below. However, this difference increases to 8.8% for trays 25 to 36. As column diameter and trays are generally designed mainly based on the maximum vapour-to-tray flow rates, determining the per cent jet flood and others, a vapour traffic data discrepancy of 8.8% is excessive for proper design of the column or hydraulic evaluation of the selected trays. Generally, new columns are designed to stay below 80-85% jet flood limit, and existing columns, varying from one case to another, may be considered acceptable up to 90-95% jet flood before opting for new, high-capacity tray or packing replacements, or even a new column. Compared to the maximum vapour-to-tray flow rate from the

PTQ Q4 2024