Process Optimization of the TMPDE factory at

Perstorp Specialty Chemicals AB

Eric Borg
2004-11-25

Introduction
Allyl Ether is widely used in unsaturated polyesters for direct gloss
coatings, e.g. for furniture and putties of different types. The major
advantage of the functionality in the allyl is the ability of inhibiting
oxygen from giving surface tackiness in radical curing resins, i.e. no
wax is required.
Allyl Ether is a colorless liquid with a characteristic odour. The compound has very low solubility in water (4 g/dm3 at 25C). The boilingpoint is 175C and the meltingpoint -40C. Allyl ether is divided into
allyl alcohol and acrolein compounds when the temperature is above
320C.
The production of Allyl Ether is made via a reaction between polyol X,
allyl chloride and sodium hydroxid. In addition to the product,
impurities are produced, mainly other chlorinated propanes or
propenes, such as 2-chloropropane, 2-chloropropene etc. The impurities affect the exchange at the synthesis and the purification of the
product. The chlorine compounds cause major corrosion in pipes and
equipment in the factory.
The purpose of this project was the following.

raise the exchange of Allyl Chloride.

lower the amount of Allyl Alcohol in the factory.

lower the amount of chlorinated compounds (impurities), which

normally are sent to external combustion at high costs.

The Project
The project was divided into two parts; First, the recycling of allyl
chloride and then the optimization of the reactor reflux.

1. Allyl chloride recycling

In the allyl chloride recycling, waste from all of the steps in the process
is gathered in a tank prior to separation via batch distillation. In the
destillation, four fractions are separated. The first fraction contains
mainly allyl chloride, which is recycled to the reactor. The breaking
point of this fraction is when the temperature in the top of the
distillation column raises above 48C. The second and third fractions
contain a lot of chlorinated compounds, which are sent to the external
combustion. The breaking points of these fractions are 60 and 83C.
The fourth fraction of the distillation is supposedly free from chlorides.
The distillation stops when the temperature reaches 92C, which
makes it possible to take care of it internally at the Perstorp site.
Prior this project, the concentration of chlorinated compounds was so
high it had to be sent to the external incineration together with the
chlorine fractions. This is not necessary anymore.
2. Reactor Reflux
Most of the compounds in the synthesis have a lower boiling point
than the reaction temperature. Hence large amount of allyl chloride
will be driven off without having reacted with other compounds. The
purpose of the reflux part is to minimize the waste of raw material. The
excess water will give an increase of the energy needed to perform the
synthesis.
In the reflux part, the compounds will condense and are then
separated into two phases. The organic phase is pumped back to the
reactor, and the water phase is taken to the allyl chloride recycling
step. However, the separation was not satisfying. Too much water was
recycled back to the reactor and too much of the allyl chloride was
driven to the allyl chloride recycling step.

Methods
The first part of this project was to find another way of performance of
the allyl chloride distillation. This way of performance should lower the
rate of chlorinated compounds in the end fraction to a value of 0.1 %
(wt/wt). This is also the limit for the internal handling and combustion
at Perstorp site.
The second part was to find a way of performing the distillation in the
reactor that would result in a fraction, free from chlorides, and a
temperature of the first of the condensors that would contribute to a
better separation between allyl alcohol and allyl chloride. In this way
more water and allyl alcohol would be separated and the energy cost
would be reduced.

Performance
To optimize the distillation, the simulation program, CHEMCAD was
used to simulate the process. The Wilson model was used as the
activity factor model in the simulations. A literature study was
performed to complete and verify the data in CHEMCAD. The results
from the simulations were confirmed by taking samples of the
distillate, the still, and the tank from the end fraction.
In order to find the optimal temperature of the cooling water, measurement samples were taken during two syntheses. The cooling temperature in the first synthesis was 72C, and 93C in the second synthesis.

Results
The chloride rate at the end fraction was lowered from a value of 4.0%
(wt/wt) to a value of 1.2%. This result was not completely satisfactory.
The tank with the end fraction can not be totally emptied, thus the
content does not fully represent the distillations after a change in
performance. However a trend where the rate of the chlorinated compounds is decreasing could be interpreted.
The separation in the first synthesis was not satisfactory. The organic
phase contained too much water, up to 34%. During the second
synthesis, the separation was more satisfying and after two and a half
hours, the water phase contained almost only water and allyl alcohol.
Therefore, it is not possible to send to waste without recycling of the
allyl chloride step. The rate of allyl alcohol decreased by 40%
compared to the first synthesis. The corresponding decrease for the
rate of diallyl ether was 10% and for the chlorinated compounds 18%.