31 nitrogen leave this tower in the distillate. About 78 of the water leaves in the distillate along
with 87 of the acetic acid, while the balance leaves in the bottoms.
Tower 3 D-103 separates acrylic acid from the remaining water and acetic acid. All the water and
acetic acid entering this tower leaves in the distillate S-129. About 85 of the acrylic acid entering this tower is recovered in the bottoms S-126 and the balance leaves in the distillate.
Tower 4 D-104 separates the water and acetic acid, which leave as bottoms S-130, from the
propane, propylene, carbon dioxide, and nitrogen, which are recycled S-131. About 90 of the acrylic acid that enters this tower leaves as bottoms. All the carbon dioxide and nitrogen leave
this tower in the distillate S-131. About 80 of the propane and propylene leave in the distillate, while the remaining 20 leave in the bottoms as wastewater. About 85 of the water
and acetic acid that entered this tower leave in the bottoms and the balance leaves in the distillate.
Recovery of Acrylic Acid About 28,550 lb acrylic acid exit the reactor R-101 each hour and enter the flash drum
F-101, which is the first separation unit. The first product stream S-120 has about 23,800 lb of acrylic acid and the second product stream S-126 has about 1,450 lb acrylic acid. This gives
an overall acrylic acid recovery of about 89. About 5 exits the process in wastewater WASTE.
32
5.3 Other Assumptions
Deriving Propane to Propylene Kinetic Expression The reaction forming propylene from propane is the first step in forming all of the
observed products. It is known that the rate-limiting step is the reaction between propane and an open site on the catalytic surface, as is evidenced by the observed kinetics.
13
It is assumed that all subsequent reactions are rapid, and that the size of the pipe reactor required for the target of 10
conversion of propane can be estimated using the known kinetics of the propane to propylene reaction.
Widi et al. performed a kinetic study of propane over a diluted Mo
1
V
0.30
Te
0.23
Nb
0.125
O
x
catalyst.
14
By varying the feed composition, they were able to obtain the reaction rate for the formation of propylene as a function of propane and oxygen concentrations. They found that the
rate expression was first order in propane and zero order in oxygen, and is be approximated by 9
At 673°K, the rate of propylene formation was found to be 0.8 molh-g catalyst at a propane concentration of 0.3 molL. Assuming a catalyst loading of 1 g, this corresponds to a k value of
0.267 hr
-1
, or 7.41 x 10
-5
s
-1
. Using the Arrhenius equation and solving for the pre-exponential factor A:
10 The pre-exponential factor was found to be 4.62 per gram of catalyst, and Widi et al. computed
an activation energy value for the formation of propylene of 62.7 kJmol, allowing the estimation of a complete kinetic expression for the propane to propylene reaction.
11 Kinetics Simplification
Currently, no experimental rate law has been determined for the production of acrylic acid from propylene over a mixed metal oxide catalyst. Thus, selectivities and conversions from
patents and journal articles were used to scale up the proposed design process as seen in Figure 5.1, page 33, from Widi et al., 2009.
It is observed that selectivity of propylene decreases as propane conversion increases and that the selectivity to acrylic acid increases as propane
13
Widi, R. K. 2012. Kinetic Investigation of Carbon Dioxid, Acetic Acid, Acrylic Acid Formation on Diluted and Leached MoVTeNb Catalyst. Indonesian Journal of Chemistry, 122, 131-134.
14
Widi, R. K., Hamid, S. B. A., Schlogl, R. 2009. Kinetic investigation of propane oxidation on diluted Mo1- V0.3-Te0.23-Nb0.125-Ox mixed-oxide catalysts. Reaction Kinetics and Catalysis Letters, 98, 273-286.
33 conversion is increased. Selectivities of acetic acid and carbon dioxide remain relatively
constant.
15
Figure 5.1: Product Selectivity Profiled for Propane Oxidation Over Diluted MoVTeNb Mixed Oxide Catalyst
Aspen Plus Simulation The APSEN Plus simulations use the NRTL-RK property method. Due to the complex
interactions between acrylic acid, acetic acid, and water, three separate azeotropes form. In order to accurately model these azeotropes, 5 additional binary interaction databanks needed to be
loaded. VLE-RK, VLE-HOC, VLE-IG, LLE-Aspen, and VLE-LIT were loaded to improve the complex interaction modeling.
Catalyst Assumptions This process uses a mixed metal oxide catalyst that is not commercially produced at the
time of this report. While conversions and selectivities, which were used in modeling the reactor for scale-up, are known, there is no market information on cost. In order to estimate a cost, it was
recommended to price the catalyst as bismuth molybdate. This estimation would give a reasonable frame of reference and allow for the catalyst cost to be included in the economic
15
Widi, R. K. 2012. Kinetic Investigation of Carbon Dioxid, Acetic Acid, Acrylic Acid Formation on Diluted and Leached MoVTeNb Catalyst. Indonesian Journal of Chemistry, 122, 131-134.