I.A.

Karimi and Rajagopalan Srinivasan (Editors), Proceedings of the 11th International

Symposium on Process Systems Engineering, 15-19 July 2012, Singapore.
2012 Elsevier B.V. All rights reserved.

Simulation and Optimization of Saponification

Process in Propylene Oxide Plant
Li Xia a, Xiaoyan Sun a, you Li b, Shuguang Xiang a
a
The Hi-Tech Institute for Petroleum and Chemical IndustryQingdao University of
Science and TechnologyQingdao266042ShandongChina
b
China Petroleum Engineering&Construction corporation (CPECC ) East-China
Design Branch Qingdao266071ShandongChina

Abstract
By combining Electrolyte NRTL and the NRTL activity coefficient model, and using
the reaction kinetics of propylene chlorohydrins (PC) saponification and propylene
oxide (PO) hydrolysis, the simulation of the saponification process in propylene oxide
plant is accomplished. In order to reduce the Chemical Oxygen Demand (COD) of
wastewater and improve the yield for propylene oxide, the effects of several conditions
for the COD of wastewater are studied. The simulation results showed that the yield for
propylene oxide has increased 0.47% and the COD of wastewater has reduced 371 mg/l
in the case of the optimum condition. Moreover, the unit operation models and
thermodynamic models used here had been proved accurately and reliably to simulate
the saponification system. So the optimized operating conditions could provide the
theoretical basis to the operation of propylene oxide plant.
Keywords: Propylene oxide, Chlorohydrination, Saponification reaction, Simulation

1. Introduction
Propylene oxides are one of the important intermediates in the petroleum and chemical
industry. Today, it is produced industrially using multistep manufacturing processes,
namely, the antiquated chlorohydrin process and the peroxidation process (Berndt, et al.,
2003).In the chlorohydrin process, a large volume of wastewater which contains organic
compounds is generated (Steltenpohl & Graczov, 2008). There is an enhanced
environmental stress caused by these organic compounds. The COD of wastewater is
very important guide posts. It not only directly reflects the status of propylene
chlorohydrins saponification and propylene oxide hydrolysis, but also affects the
wastewater treatment system. Adopting the suitable operating conditions and reducing
the COD of wastewater is crucial.
In this study, a COD-based optimization of the saponification process is presented that
develops a detailed simulation model and obtains the optimum operating conditions and
equipment configuration.

2. Process Description
The saponification process is shown schematically in Fig. 1. The saponification column
from the feed stage is a packed bed and the lower half of the column is a multistage
column. The type of multistage part is a sieve tray, and the packed bed is packed
irregularly with 4 meters height of ceramic pall rings.
 126 Li Xia et al.

Fig. 1. Process diagram of Saponification Process in Propylene Oxide Plant

2.1. Reaction Kinetics

Choosing the optimal operating conditions for saponification process requires the
kinetic parameters for the above reactions involved in the process. In a paper (Carr, et
al., 1979), a kinetic of the reaction involved in the manufacturing of propylene oxide
from chlorohydrin have been given in table 1.
Table1. Kinetic parameters for the saponification and hydrolysis reaction
Reaction Adopted kinetic model Ai Ei
(kcal/mole)
 E1 / RT
saponification rPC k1[ PC ] , k1 A1 u e 3.02h10 s9 -1
15.86
1  E2 / RT
hydrolysis rPO k 2 [ PC ] [OH ] , k1 A2 u e 8
2.68h10 mol s -1 -1
16.253
note: Ai :preexponential factors; Ei :activation energy;[OH-1]:OH-1 concentration; r:reaction rates;
[PC]:propylene chlorohydrins concentration; [PO]:propylene oxide concentration.

2.2. Exchangers, pumps, separators, reactive distillation column models

The saponification process, shown in Figure 1, is modeled using ASPEN PLUS.
Standard equipment modules are used for modeling exchangers, pumps, separators,
reactive distillation column, etc., given in table 2. Every model is constructed and
solved by the sequential modular approach.
2.3. Simulation results
Finally, the material and energy balance for a saponification process is computed in the
model. The simulation results are analyzed and compared with actual plant operation
data, given in table 3. The results showed that the simulation proposed in this paper is
suitable and can be applied to process optimization.

3. Process Optimization
Since the hydrolysis reaction lowers the yields, propylene oxide must be removed as
quickly as possible from the reaction environment. For these reasons, the production of
propylene oxide is mainly performed in saponification column, and propylene oxide is
flashed out with the stream to shorten the contact time and to prevent hydrolysis.
Simulation and Optimization of Saponification Process in Propylene Oxide Plant 127

Table2. Modeling in the saponification process

equipment model in configuration(mm) thermodynamic reaction kinetics
ASPEN PLUS diameterhheight model

V-101 RPLug 8001000 Electrolyte NRTL S. CARRA et al.5

pipelines(from RPLug 30013800 Electrolyte NRTL S. CARRA et al.5
V-101 to C-101)
V-102 Flash2 - Electrolyte NRTL -
C-101 RadFrac 2200/360025292 Electrolyte NRTL S. CARRA et al.5
E-101/102/103 Heater+Flash2 - NRTL -
P-101 Mixer - NRTL -
P-102 Pump - Electrolyte NRTL -
Table3. Comparison of the Simulation Results and the Actual Operating Data
conditions simulation results actual operating data
The dilute PC solution flow rate, kmol/h 20837.7 20837.7
Ca(OH)2 solution flow rate , kmol/h 4097.2 4097.2
Top temperature of saponification column, 90.0 92.7
bottom temperature of saponification column, 99.8 100.5
product flow rate (78wt% propylene oxide), kmol/h 274.4 -
The yield for propylene oxide, % 95.87 95.80
The COD of wastewater, mg/l 1218.0 1300.0

3.1. Optimization conditions

On initial analysis, it quickly became apparent that the COD of wastewater were heavily
dependent on hydrolysis reaction and separation of the saponification column. To
reduce the COD of wastewater, several factors, shown in Table 4, are taken into account
separately.
Table4. Optimization decisions regarding saponification
actual condition Boundary
Saponification column pressure, kPaA 97.0 67.0~107.0
The number of trays for saponification column 9 9~13
The height of weir, mm 80.0 70.0~110.0
3
The fresh steam injection rate, 10 kg/h 23.0 20.0~24.0
-1
The mole ratio of OH to the sum of PC and HCl 1.3 1.1~1.5
The saponification reaction temperature, . 349.6 344.6~354.6

3.2. Optimization Results and Discussion

In Figure 2 and Figure 3, the effects of the conditions, shown in Table 4, for the COD of
wastewater are studied ordinal. The optimal conditions are determined via a COD-based
optimization calculation that minimized the COD of wastewater.
128 Li Xia et al.

100
1200 COD
yield 98

1000
96

The COD of wastewater, mg/l 800 94

The yields for PO,%

60 70 80 90 100 110 120
Saponification column pressure,kPaA
100
1200 COD
1000 yield 98
800
96
600
400 94
8 10 12 14
The number of trays
1000 100
COD
800 yield 98
600 96

400 94
70 80 90 100 110 120
The height of weir,mm
Fig. 2. Top:Trend of saponification column pressure vs the COD and yields. Middle:Trend of the
number of trays vs the COD and yields. Bottom: Trend of the height of weir vs the COD and
yields.
1000 98
COD
yield 97
800
96
The COD of wastewater, mg/l

The yields for PO,%

600 95
20 21 22 23 24
3
The fresh steam injection rate ,10 kg/h
1000 98
COD
yield 97
800
96

600 95
1.0 1.1 1.2 1.3 1.4 1.5 1.6
-1
The mole ratio of OH to the sum of PC and the chlorine hydride
1000 98
COD
yield
97
800

96
344 346 348 350 352 354 356
The saponification reaction temperature,K

Fig. 3. Top: Trend of the fresh steam injection rate vs the COD and yields. Middle: Trend of the
mole ratio of OH-1 to the sum of propylene chlorohydrins and chlorine hydride vs the COD and
yields. Bottom: Trend of the saponification reaction temperature vs the COD and yields.
Simulation and Optimization of Saponification Process in Propylene Oxide Plant 129

Optimization is accomplished which directs appropriate changes in the independent

variables to the executive system (H. Martin Rodriguez, et al., 2010). Boundary
constraints on the independent variables, shown in Table 5, are checked before
performing the computations in the ASPEN PLUS.
Process optimization requires consideration of many trade-offs as a whole. In addition
to minimizing the COD of wastewater, the actual situation of the propylene oxide plant
must be taken into account: Because of restriction of its pumped vacuum system, the
lowest of saponification column pressure is 77 kPa(A); Although increasing the
steam consumption can lower the COD of wastewater, fresh steam is 21000kg/h from
the economic point of view.
Finally, the optimum conditions are obtained as follows: (1) saponification column
pressure is 77kPa(A); (2) the number of trays is 13; (3) the height of weir is 80mm; (4)
fresh steam is 21000kg/h; (5) mole ratio of OH-1 to the sum of propylene chlorohydrins
and chlorine hydride is 1.1; (6) the saponification reaction temperature is 354.6K.

4. Conclusions
During the course of the operation optimization described in this article, the yield for
propylene oxide has increased 0.47% and the COD of wastewater has reduced 371 mg/l.
In addition to delivering an optimal operation, the project proved that:
x Equipment configuration decisions, such as the height of weir of trays and the
volume of the saponification mixer, should be taken into account.
x Models are sufficiently accurate to be able to reflect the effects of small changes in
operating condition, for example, saponification column pressure.
x The methods used in this optimization are suitably general and can be applied to any
process plant. They are available via commercially-available simulation and
modeling tools.

Acknowledgment
Financial support of National Natural Science Foundation, China (Project 21176127),
for this work is gratefully acknowledged.

References
Berndt, T., O. Bge, et al., 2003, From Atmospheric Research to an Industrial
Process:The Formation of Propylene Oxide, Ind. Eng. Chem. Res.,42,12, 2870-2873.
Steltenpohl, P. and E. Graczov, 2008, VaporLiquid Equilibria of Selected
Components in Propylene Oxide Production, J. Chem. Eng. Data, 53,7, 1579-1582.
Carr, S., E. Santacesaria, et al., 1979, Synthesis of propylene oxide from propylene
chlorohydrins-I: kinetic aspects of the process, Chem. Eng. Sci., 34,9, 1123-1132.
Carr, S., M. Morbidelli, et al., 1979, Synthesis of propylene oxide from propylene
chlorohydrins-II: Modeling of the distillation with chemical reaction unit, Chem. Eng.
Sci.,34,9, 1133-1140.
H. Martin Rodriguez, A. Cano, et al., december 2010, Improve engineering via whole
plant design optimization, Hydrocarbon Process., Int. Ed., 43-49.