Distillation sequencing indirect sequence

FIG. 13-4 Distillation sequences for the separation of three components, a) Direct sequence, (h) Indirect sequence. 

S. Hernandez, S. Pereira-Pech, and V. Rico-Ramirez. Energy efficiency of an indirect thermally coupled distillation sequence. Can. J. Chem,. Eng., 81 (5) 1087-1091, 2003. 

When multicomponent mixtures are to be separated into three or more products, sequences of simple distillation columns of the type shown in Fig. 13-1 are commonly used. For example, if aternaiy mixture is to be separated into three relatively pure products, either of the two sequences in Fig. 13-4 can be used. In the direct sequence, shown in Fig. 13-4, all products but the heaviest are removed one by one as distillates. The reverse is true for the indirect sequence, shown in Fig. 13-4 7. The number of possible sequences of simple distillation columns increases rapidly with the number of products. Thus, although only the 2 sequences shown in Fig. 13-4 are possible for a mixture separated into 3 products, 14 different sequences, one of which is shown in Fig. 13-5, can be synthesized when 5 products are to be obtained. 

Figure 11.1 The direct and indirect sequences of simple distillation columns for a three-product separation (From Smith R and Linnhoff B, 1998, Trans IChemE ChERD, 66, 195, reproduced by permission of the Institution of Chemical Engineers.)...

Figure 11.16 The thermally coupled indirect sequence for crude oil distillation.

The majority of atmospheric crude oil distillation columns follow the configuration shown in Figure 11.17, which is basically the partially thermally coupled indirect sequence. 

Indirect reduction, 14 501 Indirect sequence heuristic, for simple distillation, 22 299... 

The six sequencing heuristics are formulated to reduce the separation load on downstream columns, favoring easier separations early and difficult separations in the absence of nonkey components. If only two products are to be derived from a mixture and all of the components in one product are more volatile than all of the components in the other product, then the next split should divide the mixture into the two products. The presence of hazardous or corrosive materials can gready increase costs, and such components should be removed as early as possible. The most plentiful product in a mixture should be removed (if it can be) with one separation and if the relative volatility is favorable. Direct sequences, ie, removing a light product as distillate, generally are favored over indirect sequences, ie, removing a heavy product as bottoms. If no product dominates the feed composition, then separations that yield approximately equimolar splits are favored. Only if no other heuristic applies should the easiest separation be performed next. 

A neural network consists of many processing elements joined together. A typical network consists of a sequence of layers with full or random connections between successive layers. A minimum of two layers is required the input buffer where data is presented and the output layer where the results are held. However, most networks also include intermediate layers called hidden layers. An example of such an ANN network is one used for the indirect determination of the Reid vapor pressure (RVP) and the distillate boiling point (BP) on the basis of 9 operating variables and the past history of their relationships to the variables of interest (Figure 2.56). 

Figure 3.6 Direct and indirect sequences of simple distillation columns.

The above method cannot be applied for highly nonideal mixtures involving azeotropes and distillation boundaries. In this case reducing the separation to handling of ternary mixtures is recommended, for which two or three columns are normally sufficient, either by direct or indirect sequence. Since the entrainer plays an important role in economics, the sequences with the entrainer recycle as bottoms are preferred. 

For example, the separations d,h, and d2h2 are sloppy splits with different amounts of A in the distillate, and accordingly with different recoveries. The separation of pure A at the top is represented by the split d h, which corresponds to a direct sequence . Accordingly, the separation of B/C in a second column is represented by the edge BC, on which h is the feed. Similarly, the first split in an indirect sequence , in which C is separated in bottoms and A/B at the top, is shown by the segment d"b". The locus of all splits between the above limit cases allows the regions of attainable products to be defined. 

Figure 3.13 presents suitable RCMs [5]. A and B but not both must be saddles, except in extractive distillation. Two columns are sufficient, either as a direct or as indirect sequence. Entrainer and mixture can be merged in the feed, except the extractive distillation, where the entrainer goes on the top. As an example we cite the separation of acetone from its azeotrope with heptane by using benzene. Contrary to expectations, the indirect sequence has better indices of investment and energy consumption. 

The separation section receives liquid streams from both reactors. For assessment the residue curve map in Figure 5.7 is of help. The first separation step is the removal of lights. This operation can take place in a distillation column operated under vacuum (200mmHg) with a partial condenser. Next, the separation of the ternary mixture cyclohexanone/cyclohexanol/phenol follows. Two columns are necessary. In a direct sequence (Figure 5.15) both cyclohexanone and cyclohexanol are separated as top products. The azeotrope phenol/cyclohexanol to be recycled is the bottoms from the second split In an indirect sequence (Figure 5.16) the azeotropic phenol mixture is a bottom product already from the first split. Then, in the second split cyclohexanone is obtained as the top distillate, while cyclohexanol is taken off as the bottom product The final column separates the phenol from the heavies. 

Table 5.14 Optimal sizing of the distillation columns in indirect sequence.

The above separation sequence, although the most used in industry, is not unique. Another possibility would consist of adopting the indirect sequence . In this case, EDC separates in the first split as bottoms, followed by VCM/HC1 distillation, as pictured in Figure 7.7 (left-hand). This alternative is penalized by excessive bottom temperature in the first split at pressures above 5 bar. In addition, in the second step an intermediate compression would be necessary for an efficient separation HC1/VCM. 

Two-column case. If the relative volatility of the product C is intermediate between the two reactants, a two-column distillation system is typically used. Either the light-out-first (LOFj, direct separation sequence, or the heavy-out-first (HOF), indirect separation sequence, can be used. The former is more common because the lightest component only has to be taken overhead once (in the first column) and not twice (as would be the case in the HOF configuration). However there are processes in which the HOF is preferred because it sometimes has the advantage of reducing the exposure of temperature-sensitive components to high base temperatures. 

This heuristic favots the direct sequence because it removes the products one-by-one as distillates and therefore minimizes vapor Row in the column. Lockhart6 points out, however, that the direct sequence is not optimal when the least-volatile component is the primary constituent of the feed stream, in this case, the indirect sequence is preferred. 

A separation is considered dilute when distillate (D) or bottom product (B) is less than 5% weight of the feed (F). The separation of a dilute mixture by distillation (simple, extractive or azeotropic) is not economical, and other methods, as liquid-liquid extraction, stripping, crystallisation, adsorption, or membrane permeation, should be applied (Bamicki and Fair, 1990). It is obvious that the decision depends on the mixture composition and the nature of components. However, even in the case of dilute mixtures, distillation may be interesting, for example as indirect sequence (see later in this section). 

---