|
Scale-Up of a Three-Phase Reactor for Catalytic Glucose Oxidation
Scale-Up of a Three-Phase Reactor for Catalytic Glucose Oxidation
Purpose
Three-phase or gas-liquid-solid (GLS) reactors are widely used for oxidation reactions. However,
oxidation reactions and their reactors have not been greatly examined industrially. Excess oxidation of the suspended
catalyst in the slurry medium leads to the problem of rapid catalyst deactivation. The reaction that we have been
examining is the oxidation of glucose to gluconic acid (and sodium gluconate upon addition of sodium hydroxide):
|
|
C6H12O7 + ½ O2 +
NaOH → C6H11NaO7
+ H2O |
(1) |
From literature, the most suitable and utilized catalysts are palladium (Pd) and platinum (Pt) on a carbon support
that is promoted with bismuth (Bi). For our purposes, we have been using 1 wt% palladium supported over alumina (Pd/Al2O3)
catalyst in its powder form. There is present information on different types of multiphase reactors available in literature. However,
reaction kinetics, mass transfer, and physical phenomena of fluid dynamics concerning multiphase reactor design have rarely been
analyzed. The overall objectives of our work are the development of a suitable GLS slurry reactor and the advancement of expertise of
reaction kinetics and transport processes for oxidation reactions. Our work has been focused on a model system of
Pd/Al2O3 catalytic oxidation of aqueous glucose with oxygen gas.
Design Summary
The scale-up of our lab-scale reactor to an industrial-size reactor has been determined for sodium gluconate
production. We assumed a conversion rate of 90% from glucose to sodium gluconate, which would produce 122 gallons per day of sodium
gluconate. The sodium gluconate plant would need 551.5 kilograms of C6H12O6, 45.1 kilograms of
Pd/Al2O3 catalyst, 10,522 liters of H2O, and 243 liters of NaOH on average per day. This plant would
have 36 cycles per day with a cycle time of 40 minutes. The total reactor volume with a void fraction of 50% would be approximately
16,148 liters. However, instead of constructing one batch reactor with this large volume, the plant would consist of four batch
reactors in series each with a volume of 112.1 liters per cycle, height of 17.9 meters, diameter of 11.8 meters, and impeller length of
4.5 meters. Another significant factor for scale-up would be the impeller speed. Since mass transfer heavily relies on intraparticle
diffusion of the catalyst, an agitation speed of 2,000 revolutions per minute would be needed.
Design Calculations
Sodium Gluconate Production
Taking an average of the world-wide market for sodium gluconate production to be approximately 1,001,200 lb/yr between the
1987-1995, we will be scaling up our GLS reactor to produce 25% of this (250,300 lb/yr).
Conversion is the degree to which a reactant is consumed in a reaction, where the limiting reactant is chosen as the basis. We
define the conversion of species A, with notation XA, as the number of moles of A that have reacted per mole of A fed into
the reactor:
For batch systems, we are concerned with the time required to achieve the desired amount of product. The difference between the
starting amount of moles of A, NA0, and the moles of A remaining in the reactor at the end of the reaction, NA is
the moles of A reacted (moles of A reacted = NA0 - NA). Dividing this by the moles of A entering the reactor
gives conversion:
Reactor Sizing
At 90% conversion, we need NAo moles of glucose per day to produce
121.574 gallons of sodium gluconate per day. Solving for NAo , determines that
1215.74 gal/day of glucose will be needed.
Semi-Batch Reactor
Reaction Rate Calculation
We can calculate the number of moles of A remaining in the reactor at the end of the reaction in terms of the conversion:
 
From F. Bang et al. (1999),
For the following batch design equation, we were able to determine the differential conversion (data from 05/02/02):
Figure 1: (Figure 3 above) Variation of reaction rate versus oxygen concentration (W. Bang et al 1999)
From Figure 1, we can see that with similar reacting and operating conditions near total conversion can be achieved in about 2
hours.
Then the molar flow rate of A leaving the system is the amount entering less the amount reacted:
Knowing feed conditions (NA0 or FA0) and reaction rate as a function of conversion will allow us to evaluate
the design equation to find the processing time or reactor volume. If the reaction rate as a function of conversion is not known, it
must be determined from either the literature or from laboratory experiment. Once a rate law is determined, use of stoichiometry allows
calculation of reaction rate as a function of conversion. The design equation is then evaluated as before.
For constant volume, CA = CA0(1–X)
For variable volume,
 
Utilizing first order kinetics, we can determine the time required to achieve a conversion, X, in a variable volume batch
reactor:
From Figure 2 (below), we found that for 90% conversion our cycle time is 19.5 minutes per cycle. As the change in the number of
moles increases, the length of reactor required for a given conversion increases. For Î
<1, the system is essentially controlled by kinetics and the mass transfer limitation is negligible. Î
is the change in total number of moles for complete conversion divided by the total number of moles fed to our reactor.
Figure 2: Conversion as a function of distance down the reactor (Fogler, 1999)
Scale-Up
24 hour reaction time: (without decant time)
|
