The right hand side of equation (6.1) should be minimised for a selected value of x2 to find T, and x,, where T* is evaluated at each incremental value of x in the adiabatic reactor. If T2
is relatively close to T,, equation (6.8b) is a good approximation to equation (6.8a) and can be used to calculate T* more easily. The relationship between a, and x, will vary and should be calculated at each incremental value of x using the UNIFAC equations or a similar activity coefficient model.
The results of the above optimisation will vary with the selected value of x2 (the final conversion) and the composition of the hydrocarbon feed stream to be used. Other factors, such as the relative cost of the tubular, water-cooled reactor and the adiabatic packed-bed reactor, are not incorporated in the above equations but should be considered in order to reflect local economics. The optimised value of T, can be used to size both the isothermal and the adiabatic reactor, via equation (6.1). The results of a semi-rigorous optimisation using equations (6.1) to (6.8) suggest that an isothermal reactor temperature of approximately 80°C is optimal to reach a final isobutene conversion of 90% at the outlet of the second reactor.
Ideally, the process reactants should always be supplied in a stoichiometric ratio to minimise their concentration in the reaction product. However, feeding an excess of ethanol to the reactors has two significant advantages for ETBE synthesis: firstly, it increases the conversion of isobutene by Le Chatalier's principle; and secondly, it minimises the dimerisation side-reaction by effectively maintaining the catalyst surface covered with ethanol (Kitchaiya and Datta, 1996). The ETBE purity decreases rapidly as the ethanol excess increases so that an excess of around 5% is preferred for most conditions.
The reaction product contains small quantities of unreacted ethanol and isobutene and up to 80% inert C4 hydrocarbons (mostly n-butenes, isobutane, n-butane and/or butadiene, depending on the source of the hydrocarbon). The Reid vapour pressure (RVP) of these components is generally too high to be blended directly into the gasoline pool. Even small quantities (less than 1%) are sufficient to substantially suppress the boiling point and the flash point of the mixture and, therefore, need to be removed from the reaction product before gasoline blending. Distillation is the simplest and most effective means of separation, although the presence of azeotropes restricts the range of feasible distillation product compositions. Most significantly, the azeotrope between ethanol and ETBE prevents an ethanol-free ETBE product from being distilled without losing a considerable quantity of ETBE in the distillate. At higher pressures, azeotropes also form between ethanol and various C4 components. These azeotropes result in a small concentration of ethanol always being present in the distillate.
The design of the purification column is dependent on the specification of both the ether and distillate products. A low C4 concentration in the bottoms product (less than 0.1%) is required while the loss of ETBE in the distillation should be minimised. The design of the purification column can be optimised using conventional techniques for distillation design (e.g. see Kister, 1994), although the combination of a very high number of stages and a low reflux ratio could result in an impractical design (i.e. too tall and too thin) due to the relatively low feed rates which are often used for etherification units. Typically, around 30 ideal stages is sufficient to meet the design requirements with a practical reflux ratio. A higher number of stages (or a higher reflux ratio) will reduce the concentrations of C4 in the ether and ethanol in the distillate, and might be desirable for some refinery configurations. Similarly, if the ether product specification is less stringent, fewer stages and/or a lower reflux ratio might be adequate.
MTBE processes have been developed to include a recovery system that removes and recycles methanol from the distillate. This is required because of the azeotrope between methanol and isobutene (and also other C4s) which contains around 10% methanol. As a consequence of the azeotrope, if the stoichiometric excess of methanol is below a critical value (around 40% for isobutene-rich feeds), methanol is preferentially recovered in the distillate product rather than the bottoms product. Some means of recovering this methanol must, therefore, be installed to avoid the loss of valuable raw material and remove the risk of downstream contamination or catalyst deactivation. The simplest method for achieving this is a water wash (which is highly selective in separating the hydrophilic methanol from the hydrophobic hydrocarbons) followed by a methanol-water distillation.
The same considerations do not always apply to ETBE units. Most significantly, the ethanol concentration in the distillate product is usually substantially lower (less than 1%). Therefore, the incentive to provide any form of recovery is significantly reduced. Secondly, the azeotrope between ethanol and water results in at least some water always being recycled to the reaction stage with the recovered ethanol. The water introduced to the system immediately reacts to form isobutanol and, thereby, decreases the yield of ETBE and the purity of the ether product. Since the ratio of fresh ethanol to recovered ethanol is around 100, the ingress of water is relatively minor and the same equipment (water wash plus ethanol-water distillation) can be used with only a slight effect on the ether purity (usually only around 1%). In practice, the need for ethanol recovery equipment will be determined primarily by downstream processing requirements and restrictions but the justification for additional equipment is much weaker than for MTBE production.
Was this article helpful?