Parameterization with respect to the aqueous reflux ratio has been carried out for the steady-state analysis of a tower to dehydrate secbutanol using disecondary butyl ether as the entrainer. With an equilibrium-stage model, singularities are observed as the second liquid phase is introduced on the trays and regions of steady-state multiplicity are demonstrated. Furthermore, a critical aqueous reflux ratio, in the vicinity of which two liquid phases are introduced on most of the trays, varies only slightly with the number of trays in the tower. The model predicts that the product purities and recoveries remain high using one-third of the trays in the industrial SBA-II tower. Bifurcation analysis for a single stage does not explain these observations, and the effect of recycle is suspected. IntroductionMany entrainers used in azeotropic distillation systems are partially miscible with one of the species in the mixture. This selection of entrainer gives a better column performance, but results in an overhead stream that condenses into two liquid phases. Normally, the heavy phase is decanted (withdrawn as product), and the lighter entrainer phase is returned to the tower as reflux. Small amounts of the heavy phase, often water, are also added to the reflux; and under certain operating conditions (e.g., at high aqueous reflux ratios), two liquid phases may extend into the rectifying and even the stripping sections (Kovach and Seider, 1987a).The behavior of a distillation system under these conditions has not been well studied, although potential problems have been suspected for some time. A few experimental studies report that the second liquid phase has a negative effect on the column efficiency (Goodliffe, 1934), while others report no adverse effects (Schoenborn, 1941). In addition, the possibility of unusual tray hydraulics and problems in control has led chemical engineers to overdesign these columns and settle for suboptimal designs.In this paper, new results are presented for an equilibriumstage model that simulates an industrial heterogeneous azeo- tropic distillation tower to dehydrate secondary butyl alcohol (SBA) with disecondary butyl ether (DSBE) as the entrainer. First, a discussion of prior work is presented with emphasis on the computational aspects and the inconclusive experimental results concerning the impact of the second liquid phase on the column efficiency and tray hydraulics. Prior Work Experimental and theoretical resultsThe dehydration of SBA was studied both experimentally and theoretically by Kovach and Seider (1987a), with emphasis on the ARC0 SBA-I1 tower. The SBA-I1 tower, which is shown schematically in Figure 1, has 40 valve trays and a 1.98 1-m ID, and is 25.1-m-tall. The tower has a large decanter and a thermosiphon reboiler. In operation, the tower is controlled manually by adjusting the aqueous reflux rate to maintain a temperature difference of 7 K between trays 31 and 36. Under normal operation, the tower has a high recovery of SBA in purities exceeding 98 wt. %. Over a two-week...

A model and algorithm are presented for the separation of mixtures where the phase distribution on the trays is extremely uncertain, as it occurs when the mixtures have two partially-miscible binary pairs and a minimum-boiling ternary azeotrope. Included are an algorithm for the consistent initialization of index-I differential/ algebraic equations, a novel algorithm for branch switching when the phase distribution changes at a real bifurcation point, and a reliable algorithm for phase stability analysis. Open-loop responses are presented for the dehydration of secbutanol with disecbutylether in single-stage, I2-tray, and 33-tray separators. These simulation results for the 33-tray tower are in qualitative agreement with experimental measurements for the ARCO SBA-II tower.

ABSTRACT

No abstract

Free-radical copolymerizations were studied under nonisothermal conditions with emphasis on their thermal runaway and ignition behavior. Computational models are presented in generalized form and compared with experiments on the system styrene-acrylonitrile. A new, useful method is proposed for the evaluation of runaway parameters from scant kinetic data. EXPERIMENTALBoth comonomers were vacuum distilled to remove commercially added inhibitors, then stored at 273K. Initiator azo-bis-isobutyronitrile (AIBN) was twice recrystallized from chloroform with methanol, dried in vacuo, and stored at 273K.The thermal ignition point apparatus (TIPA) has been described elsewhere (4). Some modifications were made