Sinopec Zhenhai Refinery and Chemical Co. (ZRCC)
started up a 450,000-tpy paraxylene (PX) complex in August
2003. The complex was originally designed to use a locally
produced xylene isomerization catalyst. The performance of
xylene isomerization catalyst during the first cycle began to
deteriorate rapidly after the regeneration in March 2007. The
ethylbenzene (EB) conversion rate dropped from 25 wt% before
regeneration to 22 wt%, and the PX/xylene ratio in the product
also dropped from 22.7% to 22% and directly led to a 0.4 wt%
reduction of PX concentration in the PX adsorption unit feed.
As a result, the PX production dropped by 50 tpd of PX; the
C8 aromatics loss was higher than 4%, which is very
high. Temperature and pressure increases in the reaction system
were increasing significantly faster than before regeneration,
with the rate of temperature increase jumping from 0.5°C
/month to 2°C/month. As the isomerization catalyst
performance deteriorated rapidly, ZRCC had planned to replace
the xylene catalyst during a scheduled maintenance turnaround in 2008.
After discussions with a number of catalyst technology owners (foreign and
domestic) and conducting a catalyst technical evaluation among
them, ZRCC selected the latest generation of a xylene
isomerization catalyst. The new first-generation xylene
isomerization catalyst was initially commercialized in 2001,
and it has now been applied in more than 11 units outside of
China. In June 2006, the Sinopec Yangzi branch applied the
first-generation catalyst successfully in its aromatics plant
with outstanding results. The newer generation catalyst is
developed based on the concept of the first-generation catalyst
with an improved manufacturing process. Processing benefits
from the new catalyst system included high activity, high EB
conversion rate, high PX/xylene ratio in the product, low
C8 aromatics losses and long cycle life. In
addition, this catalyst is very robust and can perform well in
different operating conditions. It is in operation at three
operating units in Taiwan and outside of China.
This catalyst is jointly developed and uses a proprietary
carrier. Upon delivery, ZRCC sampled and analyzed the catalyst.
The results showed that the catalyst had a loss of ignition of
0.77 wt% at 420°C and a specific surface area of 267
m2/ g, with no particles smaller than 30 mesh.
Xylene isomerization process.
The xylene isomerization reaction of the EB reforming type
catalyst is designed to isomerize aromatics present with PX in
an amount often less than 1% in the reactor feed into four
xylene isomersPX, metaxylene (MX), orthoxylene (OX) and
EBclose to equilibrium, at a defined temperature and
pressure with the presence of a catalyst. The objective is to
reduce the EB content and to increase xylene concentration of
the feed for the PX adsorption unit. The higher xylene content
to the adsorption unit increases the PX product yield and
minimizes recycling and energy consumption.
For the isomerization reaction, higher EB conversion rate
and PX concentration in the product will bring the
C8 aromatics closer to equilibrium. At the same
time, the C8 aromatics loss will be higher. This
shows that, within a certain range, the activity and
selectivity are in an inverse relationship. Therefore, careful
consideration should be given to the activity and selectivity
while operating with this catalyst. Table 1 lists the
guaranteed values of the catalyst performance parameters.
INDUSTRIAL APPLICATION OF ZRCC PX COMPLEX
ZRCCs PX isomerization unit has a designed throughput
of 267 tons/hr. In the design, an extra C8
naphthenes recycle column was added downstream of the
deheptanizer to reduce the circulation path of the
C8 naphthenes. Fig. 1 is a simplified flow diagram
of the ZRCCs isomerization unit.
Fig. 1. Simplified
flow diagram of
ZRCCs PX isomerization unit.
ZRCCs isomerization reactor is a radial-flow reactor.
Based on the calculation, centerpipe modification in the
original reactor was required to fully optimize the catalyst
performance. This was done by removing the original seal and
slump-catalyst layer and covering the top of the catalyst with
a proprietary material that could withstand high temperatures.
This top cap was followed with a layer of ceramic balls.
Pretreatment of the reactor. The spent
catalyst did not undergo carbon burning before unloading.
Thus, the reactor needed a pre-treatment step to remove the
residual hydrocarbons. In the pre-treatment process, the carbon dioxide (CO2)
level was monitored every hour. When the CO2 level
in the reaction system was reduced to less than 0.2% and the
water content to less than 1,800 ppm, the carbon-burning
treatment was considered complete.
Reactor catalyst loading. After the carbon-burning treatment of the
reactor, the catalyst was densely loaded using a proprietary
dense-loading technology. The total actual
catalyst loading was 84.96 tons.
Pretreatment of catalyst before feed.
Industrially produced catalysts will absorb moisture during
manufacturing, transportation and loading. To ensure the
activity of fresh catalyst, a drying process is done before
introducing feed to the catalyst bed. The drying process
requires heating the catalyst using nitrogen with an oxygen
content of 1 mol% to 3 mol% at 10 barg. As new insulation
materials were installed in the furnace during the downtime, to
dry the insulation materials, the heating rate was very slow at
first, and temperature was kept constant. When the temperature
reached 200°C, after which the reactor inlet temperature
was increased to 400°C at a rate of 25°C/hr and kept
constant for two hours. Water was detected at 230°C and, at
260°C, the maximum water content was about 1,800 ppm, but
no free water was detected at the low points of the system.
Catalyst reduction and passivation. After
drying the catalyst was completed, the oxygen in the system was
fully displaced by nitrogen. The pressure was then gradually
raised with hydrogen to reduce the catalyst. During the
reduction, four hours after the inlet temperature reached
420°C, water began discharging at low points of the system.
The temperature was kept constant until the amount discharged
started to decline. In total, 150.2 kg of water was collected
during the reduction process; it is equivalent to 0.18 wt% of
the water content of the fresh catalyst.
Catalyst sulfiding. Platinum, in a reduced
state of the fresh catalyst, has a too high activity, which can
lead to side reactions such as extensive cracking and/or
temperature runaway in the case of direct feed to the reactor.
All side reactions impact the catalysts long-term
performance and service life. For that reason, catalyst
pre-sulfiding is done before introducing feed in the catalyst
A total of over 100 kg of DMDS sulfiding agent was injected
into the system in two steps through a specific device onsite.
It was hard to detect hydrogen sulfide (H2S) due to
monitoring a large range of tubes, and no H2S was
detected at the outlet of the reactor after the sulfur
injection. After the sulfiding, the reactor inlet temperature
was maintained at 335°C for the final preparation for feed
Catalyst feed introduction.
Prior to introducing feed to the catalyst bed/reactor, the
composition of the liquid feed (raffinates and makeup hydrogen)
was analyzed, and the feed met the required specification. The
liquid feed was introduced into the reactor under these
conditions: inlet temperature of 335°C, reactor pressure of
11 barg, recycle-gas hydrogen purity greater than 80% and
makeup-hydrogen purity greater than 97 mol%. The weight hourly
space velocity was 3.0 hr1. After feed
introduction, the reaction pressure decreased rapidly, and
temperature rose sharply, during which the reactor outlet
temperature rose to a maximum of 418°C, and the maximum
reactor ∆ temperature was 50°C. After the first round
of heat waves passed, the reactor ∆ temperature dropped
to 35°C. This signified a successful catalyst feeding. Fig.
2 shows the reaction temperature increase and pressure changes
during feed introduction.
Fig. 2. Reaction
temperature increase and
pressure changes during feed
During the initial operating period after feeding, there were
obvious increases in hydrogen consumption and in gas generation
due to the formation of C8 naphthenes within the
first few hours. After 24 hours, the difference between the
temperatures at the reactor inlet and outlet was reduced to
about 21°C, and the reactor outlet pressure was reduced
from 10 barg to 6.3 barg, with the hydrogen purity up to 86%.
The EB-ate and PX-ate were maintained at a relatively high
level. With the catalyst further stabilized, the C8
aromatics loss was significantly reduced, thus indicating that
the C8 naphthenes balance was established. After 48
hours, the PX-ate and EB-ate remained unchanged, with the
C8 aromatics losses further reduced to the expected
values. Thus, the feed introduction was successful.
After smooth operations for more than three months, ZRCC and
the catalyst provider conducted a 72-hour catalyst performance
test run. Tables 2 and 3 summarize the catalyst performance
test run conditions and results, respectively. During the
catalyst performance test run, all performance indicators were
met, with PX-ate and EB-ate, and the C8 aromatics
loss was better than the guaranteed values.
Performance comparison between catalysts.
From the catalyst performance test run results, the new
catalyst shows a great advantage in performance as compared to
the original isomerization catalyst. The new catalyst system
operates better in each of the indicators, especially the
PX/xylene ratio, EB conversion rate and C8 aromatics
loss, under less severe operating conditions (at a similar
temperature but at a lower pressure of 0.7 barg). The
comparison of specific performance indicators is summarized in
From the comparison of the reactor inlet and outlet
compositions, the PX concentration in the product increased by
2.2 wt% and EB concentration was reduced by 3.1 wt% using the
new catalyst. Daily PX production increased by about 110 tpd,
and ZRCC can achieve the operations target of increasing PX
production while significantly reducing the C8
aromatics losses without any modifications to the process or to
the configuration in the complex.
In 2009 and 2010, after the new catalyst was installed, ZRCC
produced 500,000 tons of PX annually, in spite of some
performance decline in the PX adsorption unit observed at the
end of the period. The PX production was maintained at a
similar level before replacing the catalyst, with the
throughput remaining unchanged, and the C8 aromatics
losses were reduced by 30,000 tons and 20,000 tons,
respectively. The excellent range of performance offered by the
new xylene isomerization catalyst allowed to compensate
declining PX adsorbent separation performance by increasing
operational severity and thus helped to maintain the
competitiveness of the complex.
Longer operation cycle for the catalyst.
By January 2011, the xylene isomerization unit has been in
operation for over two years, and the catalyst still
demonstrated good isomerization activity and conversion rates
while maintaining a low C8 aromatics loss. The
reaction inlet temperature is now 374°C, and the rate of
temperature increase is less than 0.5°C/month on average,
particularly in 2010 when the temperature was only raised by
4°C for the entire year. These conditions indicate that the
catalyst has excellent stability.
ZRCCs PX plant has been in continuous operation for
nearly eight years. In the second half of 2010, a decline in PX
separation efficiency was observed. In order to maintain PX
production throughput, ZRCC decided to reduce the EB
concentration in the adsorption unit feed and to optimize the
adsorption unit efficiency. ZRCC increased the operating
pressure of the isomerization unit to maintain a relatively
high EB conversion rate. In this case, even with a relatively
high EB conversion rate and high PX/xylene ratio, the
C8 aromatics loss was kept at a relatively low
level, indicating that the catalyst is a very robust catalyst
and can operate well in a wide range of conditions.
The new generation xylene isomerization catalyst showed a
very good performance after the startup. The performance test
run data confirmed that indicators such as PX-ate, EB-ate and
C8 aromatics loss are superior to the guaranteed
values. The application at ZRCC is considered as a success.
With better PX/xylene ratio in the product and a higher EB
conversion rate, the catalyst helped to optimize the PX
adsorption unit feed while increasing PX production and
lowering C8 aromatics loss. This led to a reduction
in the C8 aromatics feed requirements and increased
the total competitiveness of the PX complex, without any
modification to the process and/or configuration of the
complex. In addition, data from the long cycle operation
indicate that the catalyst operated with good performance even
under severe feed conditions, and it is a very robust catalyst
with a good response to different operating conditions.
Yingbin, Q., The progress and application of catalysts
for isomerization of C8 aromatics,
Engineering Science, Vol. 1, No. 1, 1999.
Zhanggui, H., Industrial application of the SKI-400C
catalysts for isomerization of C8 aromatics,
Petroleum and Petrochemical Today, Vol. 35,
No. 4, 2005.
Yang, J. and D. Shi, Industrial application of the
SKI-400-type catalysts for isomerization of C8 aromatics,
Petroleum Refinery Engineering, Vol. 1,
No. 1, 1999.
1 This catalyst is jointly developed by Zeolyst
and Axens, using a proprietary carrier.
Guo Shouquan joined ZRCC in 2001 in
the aromatics production and technical management
Jenson Chua joined Zeolyst
International in 2006 as a technical consultant.