Impact of Feed Properties on the FCC Unit – Key Issues in a Diesel Market

Abstract

Nine heavy vacuum gas oils having different hydrocarbon compositions and different distillation
characteristics were cracked in a fluid catalytic cracking (FCC) micro activity unit with the aim to
determine which feedstock characteristics impact on diesel (LCO) yield maximization. While it was
shown that maximum gasoline yield and the conversion at which gasoline is at a maximum depend
on FCC feed hydrocarbon composition, the maximum LCO yield and conversion at which LCO is at
a maximum was found to be independent of feed hydrocarbon composition. The FCC feedstock
parameter that showed an impact on the maximum LCO yield and the conversion at which LCO is
at a maximum was the diesel fraction content in the FCC feed. The lower the diesel fraction
content in the feed the higher the net diesel yield and the higher conversion at which LCO yield is
at a maximum was observed. Hydrotreatment of the FCC feed was found to lower the LCO yield.
Increasing the FCC feed conversion showed a decrease in the LCO cetane index. The increase of
conversion leads to lowering the low value slurry yield but also deteriorates the LCO quality. This
implies that hydrotreatment is always necessary before blending this fraction into the final diesel
pool.

Introduction

The global diesel demand is already much larger than gasoline. In the middle of 2009 the
worldwide diesel demand accounted for about 24 million b/d, while that of gasoline was about
21.5 million b/d [1]. The demand for diesel in 2020 is expected to reach 29.4 million b/d compared
with 23 million b/d [1] for gasoline. The continual growth of diesel demand, especially in Western
Europe, Asia and Latin America is driven by a number of factors: the superior economy and
efficiency of diesel engines; advances in combustion systems and improvements in emission
devices including filters for the removal of micro-particulates and NOx reduction systems. The fluid
catalytic cracking unit (FCC) is still a main conversion unit in many refineries. It is able to process
very large amounts of heavy oil fractions and it is flexible enough to convert them preferentially to
propylene, butenes, gasoline or diesel, with minor modifications of the unit or the operating
conditions [2]. Historically, the fluid catalytic cracking unit (FCCU) has taken the role of the
refinery’s top gasoline producer. Its role has expanded from a gasoline machine to an olefins
maker, a sulfur remover, a residue upgrader, and possibly a significant contributor to the ultra low
sulfur diesel (ULSD) pool. These FCC technology advances are drivers of the perpetual refinery
profit cycle. For many refiners it would be advantageous to achieve a level of flexibility in the
FCCU, producing varying amounts of gasoline and light cycle oil (LCO) as the market changes. The
FCCU contributes about 30% of the US’s LCO supply, but this percentage is much higher in Europe,
Asia, and Latin America, where middle distillates are in higher demand. Furthermore, Europe,
China, and Brazil are known to be short of diesel [3]. FCCU LCO production can be increased by
modifying feedstock composition, introducing improved catalysts and additives, and modifying
operating conditions (i. e., recycle ratio, temperature, cat-to-oil ratio, etc.). The addition of an
active alumina matrix is a common feature to help refiners increase LCO production when cracking
heavy feeds. There is some overlap in catalysts tailored for LCO production and those tailored for
resid feeds. Consequently, refiners looking to produce more diesel may want to consider heavier
feeds to their FCCU. Process/ hardware technologies to improve FCCU LCO yield include proper
feed injection systems and risers/reaction zone designs as claimed by Petrobras, Shell, and Sinopec
in their latest commercial processes [3]. Finally, the use of biofeeds in the FCCU to produce a large
volume of high-quality LCO has also received some attention. These feeds, namely animal fats and
vegetable oils, not only increase LCO yield, but also provide high-quality products in terms of
cetane number [4]. The FCC feedstock quality is known as the main factor that influences
conversion, gasoline, gas, and coke yields [5–8]. However, little information has been published
about the effect of FCC feedstock quality on LCO yield maximization. For that reason we
investigated the effect of feed properties on maximizing diesel yield in the fluid catalytic cracking
process. The aim of this work is to discuss the obtained results.

Experimental

In order to find out the impact of hydrocarbon composition and distillation characteristics of
vacuum gas oil (VGO) on FCC LCO yield two vacuum gas oils (SIHGO – Sour Imported Heavy Gas Oil,
and HTREB – Hydrotreated heavy vacuum gas oil from Russian Export Blend crude) having
different distillation characteristics were treated with H2SO4 according to the procedure described
in [9] with the aim to obtain vacuum gas oils that have the same distillation characteristics and
different hydrocarbon compositions. Properties of the two base vacuum gas oils and the model
vacuum gas oils obtained from them are presented in Table 1. All investigated vacuum gas oils
were cracked using the commercial equilibrium catalyst described in Table 2 after calcination for 3
h in flowing air at 540 °C. Catalytic activities and selectivities were determined by the micro activity
test ASTM D3907-92. The ratio between catalyst and oil was varied from 0.5 to 6 to obtain yield
curves as a function of conversion. Experiments were carried out at a reaction temperature of 527
°C and a catalyst time-onstream duration of 30 s. The details of the experimental procedure and
the modelling of selectivity data obtained from microactivity testing of FCC catalysts are described
in ref. [10]. Conversion is defined as 100 – (LCO+HCO). Light cycle oil (LCO) and heavy cycle oil
(HCO) are the yield fractions in the cracking products as wt.% of the feed with cut-points of 216 °C
< LCO < 343o C < HCO.

Results and Discussions In many studies the cracking of gas oil is presented following the reaction
scheme given below, which represents the main fractions in the fluid catalytic cracking process
[11]:

The three main cracking products are presented as primary products, with selectivities derived
from cracking rate constants that are usually in the range [11]: LCO: 20–40% Gasoline: 40–70%
Gas: 10–30% These selectivities can vary greatly depending on feedstock composition [5–7].
Vacuum gas oils consist of a complex mixture of molecules that can be classified in three main
groups: paraffins, naphthenes and aromatics. Paraffinic feeds (high KUOP) allow high conversions,
with high selectivities to gas and gasoline products, while LCO and HCO yields are low. Aromatic
feeds (lower KUOP) give lower conversion, with higher selectivities to LCO. Moreover, LCO quality
is poorer from aromatic feeds than from paraffinic feeds, but both need to be improved to reach
current diesel standards. Naphthenic feeds present an intermediate behaviour. Residual
feedstocks are similar in nature to VGOs, with a higher boiling point and an increased content of
multi-ring core aromatics. Thus, cracking patterns from VGOs also apply to residual feeds. Fisher-
Tropsch waxes are also a potential FCC feedstock, but differ from the feeds already described, as
they consist of an unusually high proportion of linear paraffins. As a result, high conversion and
high gas yields were observed, together with low LCO yield [12]. Corma discussed the conversion
of vacuum gas oil from the point of view of the relative cracking rates of VGO and LCO for the
different classes of components present in the feed [11]. He found out that although the cracking
of paraffinic feeds has the potential to yield high quality diesel, they are not well suited for the
operation since they give a low yield of LCO with common FCC catalysts. Feeds containing one and
two-ring molecules may be the best option for acceptable diesel yield and quality if the unit is run
at moderate VGO conversion, using low temperature and recycle operation. Feeds containing
molecules with three and more rings should be avoided since the concentration of aromatic cores
in the cracked diesel can hardly be avoided, leading to the poorest light cycle oil (LCO) quality with
regard to cetane index and particulate emissions [11]. The conclusions made by Corma [11] on the
base of the discussion about the LCO yield that could be obtained in the FCC process by cracking of
individual paraffins, one-ring naphthenes and aromatics, two and more fused rings boiling in the
VGO range may be different when cracking VGO that consists of a mixture of paraffins,
naphthenes and aromatics with a different distribution of one, two and more fused rings. The
impact of VGO distillation characteristics on FCC LCO yield has also not been discussed. The data in
Table 1 show that the four model vacuum gas oils obtained from SIHGO have the same distillation
characteristics as SIHGO, but the model gas oils differ in their group hydrocarbon composition. The
Model 1-SIHGO base and Model 1-HTREB base have hydrocarbon compositions which are very
close to the original SIHGO and HTREB respectively and distillation characteristics equivalent to
those of the base gas oils. However, the Model 1-SIHGO base and Model 1-HTREB base do not
contain basic nitrogen compounds. Therefore, a difference between the LCO yield obtained from
the base vacuum gas oils and the Model 1 feeds should be attributed to the different
content of basic nitrogen in the original and model feeds. Figure 1 shows the dependence of LCO
yield of SIHGO, HTREB and Model 1-SIHGO base and Model 1-HTREB base on conversion. It is
evident from these data that as a primary unstable product the LCO yield from the four vacuum
gas oils peaks by increasing conversion and then decreases. These data indicate that removing
basic nitrogen compounds from SIHGO and HTREB leads to an increase in LCO yield. The data also
indicate that the HTREB and Model 1-HTREB base achieve a maximum yield of LCO at a higher
conversion (55%) than that at which a maximum yield of LCO is attained during cracking of SIHGO
and Model 1-SIHGO base – 40%. The HTREB and Model 1-HTREB base contain a 5% fraction boiling
up to 343 °C while SIHGO and Model 1-SIHGO base contain 21% of this diesel fraction. These
results suggest that the distillation characteristics of the vacuum gas oil have an impact on the
level of conversion at which the maximum LCO yield is obtained during catalytic cracking. Reducing
the diesel fraction content in the FCC feed may shift the conversion at which LCO yield is at
maximum to higher levels. Figure 2 presents a graph of LCO yield against conversion for the model
vacuum gas oils obtained from SIHGO: Model 2-SIHGO base, Model 3-SIHGO base, and Model 4-
SIHGO base. These data do not indicate a significant difference in the LCO yield obtained from the
three feeds. Therefore, the difference in the hydrocarbon composition of the model vacuum gas
oils does not affect the LCO yield in the investigated range. In contrast to the LCO yield, the
gasoline yield is strongly influenced by the hydrocarbon composition (Fig. 3). With an increase in
the content of saturates and light aromatics in the VGO (so called gasoline precursors) an increase
of gasoline yield and conversion at which gasoline yield is at a maximum is observed. The
conclusion could be made that the level of saturates and light aromatics in the FCC feed has a
strong impact on gasoline yield and maximum conversion level. However, the gasoline precursor
level in the FCC feed did not show any significant effect on the LCO yield and the conversion at
which LCO is at a maximum. Table 3 summarizes the results obtained from the catalytic cracking of
SIHGO, HTREB and the Model feeds for the maximum yield of LCO and the conversion at which
LCO yield is at a maximum. It is evident from these data that the strongest effects on the net LCO
yield (the LCO yield minus content of the fraction boiling up to 343o C in the FCC feed) result from
the VGO distillation characteristics and the basic nitrogen content. For example the net LCO yield
of SIHGO (21% diesel) was 5.2% while that of HTREB (5.6% diesel) was 10.9%. The net LCO yield of
Model 1-SIHGO base (21% diesel) was 7.7% while that of Model 1-HTREB base (4.5% diesel) was
15.5%. The maximum conversion at which LCO is at a maximum for SIHGO and Model 1-SIHGO
base was about 40% while that of HTREB and Model 1-HTREB base was 55%. Therefore, reducing
the diesel content in the FCC feed increases not only the level of maximum conversion at which
LCO is at a maximum but also the net LCO yield. This has a strong positive effect on the refinery
economics increasing the high value diesel and decreasing the low value fuel oil production. In
order to assess the influence of FCC feed hydrotreatment on LCO yield, heavy vacuum gas oil from
Urals crude (HVGO) and hydrotreated heavy vacuum gas oil from Urals crude (HTHVGO) were
cracked in the MAT FCC unit. Properties of these vacuum gas oils are presented in Table 4. These
data indicate that HVGO hydrotreatment leads to
saturation of the three and two fused aromatic rings to one-ring aromatics without changing the
total aromatics content. There is no difference in the distillation characteristics of both vacuum gas
oils. Figure 4 presents a graph of LCO yield against conversion for both HVGO and HTHVGO. It is
evident from these data that HVGO yields more LCO than HTHVGO during catalytic cracking and
the conversion at which the LCO yield is at a maximum for both feeds is about 60%. The higher
LCO yield obtained by cracking the unhydrotreated VGO suggests that the higher content of two
and three fused aromatic rings in it contributes to a higher LCO yield. This finding is in line with the
results reported in earlier investigations [13]. The drawback of maximizing LCO is the in

crease of bottoms yield. Recycle is often required to fully maximize LCO while maintaining the
bottoms yield consistent with a traditional maximum gasoline operation. However, the recycle
quality is inferior to that of the original FCC feed. This is very well illustrated in the work of Hunt et
all [14]. They showed that the heavier the recycle fraction the poorer the properties as FCC
feedstock: lower API, lower hydrogen content and higher Conradson Carbon content. Recycle
quality also deteriorates with an increase in conversion. An investigation carried out on the Lukoil
Neftochim Bourgas FCC unit (Fig. 5) also showed that an increase of conversion results in poorer
LCO quality (lower cetane index and Kw-factor). These data indicate that the light cycle oil fraction
obtained by FCC is high in aromatics (low Kw-factor) and has a low cetane index. The increase of
conversion lowers the low value slurry yield but also deteriorates the LCO quality. This implies that
hydrotreatment is always necessary before blending this fraction into the final diesel pool.
Conclusions LCO from fluid catalytic cracking of vacuum gas is a primary unstable product whose
yield achieves a maximum by increasing conversion and then decreases. The maximum LCO yield
depends largely on three main FCC feed characteristics: 1) content of diesel fraction in the FCC
feed; 2) content of basic nitrogen compounds; and 3) content of poly-nuclear aromatics.
Reduction of the diesel fraction in the FCC feed leads to a higher net diesel (LCO) yield and a higher
conversion at which LCO is at a maximum. Removal of basic nitrogen compounds from the feed
leads to an increase in LCO yield, due to an increase in the ratio of chain cracking reaction
propagation to beta cracking. Hydrotreatment of the FCC feed decreases the concentration of the
three and two fused aromatic rings in the vacuum gas oil, which leads to a lower LCO yield.
Increasing the level of paraffins, naphthenes and light aromatics (one-ring) in the FCC feed leads to
a higher gasoline yield and higher conversion at which gasoline is at a maximum. However, it has
no discernible impact on LCO yield and the conversion at which LCO is at a maximum. The increase
of conversion reduces the low value bottoms yield. However, it also deteriorates the LCO quality.
In order to achieve the high quality of modern automotive diesel fuels FCC LCO needs to be
hydrotreated before blending.