Abu-Eishah & Luyben 1985 _Design and Control of a Two-Column Azeotropic Distillation System
9
132 Ind. Eng. Chem. Process Des. Dev. 1985, 2 4 , 132-140 variation o f the column capacity. Nevertheless the SR R diminishes from 43% to 37%. If these results are compared with those obtained from the ethylbenzene/styrene separation (example2) in which the column rates are 60% of example 1, we come to the following conclusions. 1 . The parameters having the most influence on the determination of the profitability of the vapor recom- pression heat pump applied to an existing distillation column are: (i) th e tempe ratu re d iffer ence bet ween top and bottom, (ii) th e column capacity, and (iii) th e need of preheating. 2. The smaller the temperature difference between top and bottom is, the more prof itable the heat pump appli- cation will be, for th e follo wing reasons: (i) small tem- perature differences indicat e difficult separations with high reflux ratios, and consequently, high steam and cooling water consumptions, and (ii) the pressure ratio and the compression power which are required increase with t he increase of temperature difference. 3. The larger t he column is, the more favorable will be the cost reduction of the main equipment referre d to th e proportional increase in energy saving. 4 . The smaller the preheating is, the more interestin g th e heat pump applica tion wil l b e, beca use: (i) there is a reduction in t he size and cost of E-1, (ii) the compression power decreases a s the inlet temperature of the compressor decrea es, and (iii ) the steam consumption in E 6 decreases if the condensate is allowed to give more heat in the re- boiler-condenser E-2 by subcooling. Concluding Remarks Calculation programs of simulation and optimization have been developed which allow us to analyze the eco- nomical viability o f substituting the convention al rebo iler an d condenser o f a distil lation column by a vapor re- compresion heat pump. For the two cases presented th e PO T is lo we r tha n 2. 5 years. Th e compariso n among th e above presented cases and other similar ones which have been studied by us (Flores, 1983), indicates th at the parameters show ing th e position o f t he optimum for the same system are relatively stabl e in front of the variation of the column capacity, the ratio o f energy prices/equipment costs, and t he relation bet wee n coolin g water a nd electricity cost. However, th e value of th e POT in th e optimum is ver y sensitive to th e variation of these factors, and therefore from this point of view it is necessary to take into th e account the possibl e re-usin g of the conventional equipment to be substituted, with which the required additional investment would be re- duced. Literature Cited Barnwell. J. ; Morris, C. P. Hy&ocerbon Process. 1982, 67(7), 117-199. Chauvel, A.; Leprince, P.; Barthel, Y.; Raknbault, C. ; A r b , J. P . “Manual of New York, Danziger, R . Chem. Eng. Prog. 1979, 75(9), 58-64. Flnelt, S. Hydrocarbon Process. 1979, 58(2), 95-98. Flores, J . Hydrocarbon Process. 1984, 63(7), 59-62. Kuester, J. L.; Mlre, J. H. “Optimlzatlon Tech nique s wlth F ortran”; McG raw- Menzies, M. A. ; J ohnson , A. I . Can. J. Chem. Eng. 1971, 49 , 407-411. Null, H. R . Chem. fng. frog. 1978, 72(7), 56-64. Quedrl, G. P. Hydrocarbon Process. 198la, 60(2), 119-126. Quadri, G. P. Hydrocarbon Process. 1981b, 60(3). 147-151. Economic Analysis of Chemical Processes”; McGraw-Hill: 1961. Hill: New York, 1973. Received f or review July 6, 1983 Accepted March 2 6 , 1984 This work was financially supported by Do w Chemical Iberica, S.A. Design a n d Control of a Two-Column Azeotropic Distillation System Samlr . Abu-Elshah and Wllllam L. Luyben’ Department of Chemical Engineering, L ehlgh University, Bethlehem, Pennsylvania 1801 5 Th e steady-state desi gn of a two-column azeotropic distillati on system operating at two different pressures was studied with th e objective of reduclng energy consumption. The minimum-boiling, homogeneous binary azeotropic system tetrahydrofuran-water was used as a speci fic example. Energy consumptionwas reduced by a factor of 2 from conventional designs by using heat integration and feed preheat and by optimizing column pressures and overhead purities. The dynamics and control of the two-column, heat-integrated system was also explored. A control system was develop ed that effectively handled a variety of disturbances withou t any severe interactions between columns. Introduction Th e design o f azeotropic distill ation systems has been the subject of many papers and books (Hoffman, 1964). Th e use o f two columns, operating a t two different pres- sures, is one o f the simplest a nd most economical tech- niques for separating binary azeotropes (Van Winkle, 1967 ), provided that a substanti al shift in th e composition of the azeotrope occurs when pressure is changed. For a minimum-boiling homogen eous binary azeotrope the distillate from the low-pressu re column is fed t o the high-pres sure column (see Figure 1). The distillate from the high-pressure column is recycled back to the low- pressure column. These distillates have compositions th at are close to th e azeotr opic co mpos itions a t their respective pressures. More or less pure products are removed from the bottom of each column. In the tetr ahydrof uran (THF)-wate r system, water is produced from the base of the low- pres- sure column. TH F, the low-boiling component, is pro- duced from the bottom of the high-pressure column. While the steady-state design of these systems has bee n discussed, at least qualitatively, in th e literature, there ha s been very little reported concerning the dynamics and control of th e two-column system. Shinskey (1977) gav e 0196-4305/85/1124-0132$01.50/0 0 1984 Amerlcan Chemical Society
