Вы находитесь на странице: 1из 13

Strategies for increasing production of

light cycle oil


Refineries should consider several strategies, including recycle and enhanced
catalyst activity, to increase LCO output from the FCCU

David Hunt, Ruizhong Hu, Hongbo Ma, Larry Langan and Wu-Cheng Cheng Grace Davison

M
aximising the FCC light cycle oil (LCO) Introduction
yield to take advantage of high diesel The Energy Information Admini-stration (EIA)
prices relative to gasoline requires re- expects higher diesel prices relative to gasoline
optimisation of the product cut point, operating to continue through 2010. An average diesel
conditions and catalyst technology. It is well price of $2.27/gallon was expected in 2009,
known that the LCO-to-gasoline ratio can be increasing to an average $2.54/gallon in 2010.
increased through lowering conversion by adjust- Gasoline was expected to be an average $1.87/
ing FCCU operating conditions and decreasing gallon in 2009, reaching an average $2.18/gallon
catalyst activity.1,11 The drawback of this approach in 2010.2
is the increase in bottoms yield. Recycle is often New Corporate Average Fuel Economy (CAFE)
required to fully maximise LCO, while maintain- standards are also expected to increase diesel
ing the bottoms yield consistent with a traditional demand relative to gasoline in the coming years.
maximum gasoline operation. This new standard requires car manufacturers to
This article is a general discussion of strategies boost fuel mileage to 35 mpg by 2020 and
for maximum production of LCO in the FCCU. applies to all passenger vehicles, including light
We will present laboratory results that quantify trucks. To meet this challenging, new standard,
the effects of various recycle
streams. A resid feedstock was
cracked over a low zeolite/ 
matrix (Z/M) Midas catalyst in 
Low Z/M
Grace Davison’s circulating 

riser (DCR) pilot plant. The 
product bottoms was distilled 

to five recycle fractions (650–
"OTTOMS WT


750°F, 650–800°F, 650–850°F, 
,#/ WT

650°F+ and 750°F+), blended High Z/M 



back at various levels with the High Z/M
original feedstock, and cracked  
over a Midas catalyst in the  
ACE unit. Laboratory testing Low Z/M
results were used to model a  
commercial operation to  
demonstrate the yield advan-          
tage of selecting the appropriate #ONVERSION WT #ONVERSION WT
recycle stream, recycle ratio and
catalyst technology. Figure 1 LCO and bottoms vs conversion

www.digitalrefining.com/article/1000142 PTQ Q4 2009 1


• Feedstock:
DCR runs to generate recycle feedstock
Removal of diesel-range material from the

FCC feedstock
Rx exit temp,ºF 950 950 971 970
Regenerator temp,ºF 1350 1350 1270 1270 FCC feed hydrotreating severity optimisation

Feed temp, ºF 701 574 700 299 Residual feedstock optimisation


Conversion, wt% 54 58 68 75 • Catalyst optimisation:


Cat-to-oil ratio 4.32 5.01 5.92 9.37 Increasing bottoms conversion

Dry gas, wt% 2.01 1.93 2.63 2.24


LPG, wt% 8.21 8.88 11.39 13.34 Lower Z/M surface area

Gasoline, wt% 38.38 41.99 47.97 51.89 Maintaining C3+ liquid yield and gasoline

LCO, wt% 22.15 21.71 19.18 16.73 octane.


Bottoms, wt% 23.97 20.04 12.84 8.55
Coke, wt% 5.15 5.34 5.89 7.13
Recycle streams: heavy cycle oil or bottoms
Boiling point distribution of 650ºF+ bottoms, wt% Reducing the gasoline end point is a simple and
650–700 5.31 4.75 3.45 2.47 effective way to increase the production of LCO.
700–750 4.82 4.25 2.87 1.95 The LCO flash point specification and main frac-
750–800 4.30 3.61 2.18 1.43
800–850 3.57 2.85 1.62 1.05 tionator salting often determine how low a
850–900 2.51 1.95 1.17 0.73 refiner can reduce the LCO initial boiling point
900–950 1.63 1.24 0.77 0.49 (reduce the gasoline end point). The LCO end
950–1000 1.02 0.78 0.80 0.44 point should be increased within the maximum
1000–1050 0.83 0.60 0.00 0.00
constraints of the main fractionator bottoms
650–750 10.13 9.00 6.32 4.42 temperature, slurry exchanger fouling and diesel
650–800 14.44 12.60 8.50 5.85 hydrotreater constraints.
650–850 18.01 15.46 10.12 6.90 Maximising LCO in the FCCU at reduced
650+ 24.00 20.03 12.86 8.56
750+ 13.87 11.04 6.54 4.14 conversion without produc-ing incremental
bottoms oil is the true challenge. Shifts in oper-
ating conditions to reduce conversion increase
Table 1 LCO, but also increase the bottoms oil yield, as
seen in Figure 1. In order to produce increased
more efficient vehicles powered by hybrid and LCO without producing incremental bot-toms,
diesel engines are expected. refiners will often reformulate their FCC catalyst
Refiners are increasing the production of LCO to a lower activity with lower zeolite to matrix
from their FCCUs to take advantage of the signif- for improved bottoms conversion and to mini-
icantly higher value of diesel relative to gasoline. mise LCO over-cracking.
Figure 1 shows how LCO and bottoms shift Recycle is eventually required to minimise
versus conversion for a high and low Z/M ratio bottoms production as the refinery reduces
catalyst. LCO, such as gasoline, is an intermedi- conversion to reach an optimal LCO yield, as
ate product increasing with conversion at very suggested by Figure 1. Almost all FCCUs oper-
low conversion levels, eventually reaching an ated with large recycle streams prior to the
over-cracking point. Past this point, the LCO introduction of zeolite catalyst in the 1960s and
yield declines with increasing conversion. This 1970s. However, since the 1970s, recycle was
high conversion regime represents the traditional generally removed from FCCUs as the catalyst
FCCU operating point. A low Z/M catalyst gener- and equipment technology improved and the
ally produces higher LCO at the expense of FCCU operated at high conversion and feed rates
bottoms for a given conversion level, as suggested to produce gasoline, C4s and C3s. As a result, our
by Figure 1. industry has limited recent experience with these
Refiners tend to focus on the following strate- recycle streams and their effect on FCC products
gies for maximum FCC LCO production: and coke.
• Reduced gasoline end point The better to understand these recycle effects
• Operating conditions: with the maximum bottoms cracking catalyst
Lower reactor temperature
■ Midas, Grace has conducted a series of riser pilot
Higher feed temperature
■ plant (DCR), ACE and computer simulation
Lower equilibrium catalyst activity
■ studies.

2 PTQ Q4 2009 www.digitalrefining.com/article/1000142


Experiments
DCR pilot plant runs and

#ONRADSONCARBON WT

preparation of recycle streams 


!0)GRAVITY
A commercially available 

Midas catalyst was deacti-

vated, without Ni or V, at 

1465°F (800°C) for 20  

D
&
 ²&
²&
²&
²&
²&
²&
hours, using the Advanced

ED
²&
²&
²&
²&
²&
²&
²&

²
FEE




FE

n
n
n




n 
n 
n 




Cyclic Propylene Steam

SH
SH







&RE
&RE
protocol described by &EED)$
&EED)$
Wallenstein. After deacti-
3

vation, the catalyst had a  


94 m2/g zeolite surface

(YDROGEN WT

 VOL&
area, 83 m /g matrix
2
 
surface area and a unit cell 

size of 24.30 Å. The deac-
 
tivated catalyst was
D
&
 ²&
²&
²&
²&
²&
²&

ED
²&

 ²&
²&
²&
²&
²&
²&
²

charged in our DCR pilot


FEE


FE
n
n
n

n
n
n









SH

SH
plant,4 where cracking of a





&RE

&RE
&EED)$ &EED)$
resid feedstock was
conducted. Reaction sever-
ity was varied by adjusting Figure 2 Properties of recycle feedstocks obtained from DCR run at 54% conversion
the temperature set points
of the riser top, regenera-
tor and feed preheat. We API gravity Conradson carbon, wt%
 
obtained four balanced

runs with conversion levels 
$ATA


of 54, 58, 68, and 75 wt%. 
$ATA


The DCR conditions and 

product yields are listed in  
ED

 ²&

# ²&

 ²&

 ²&

Table 1.
D

C ²&

 ²&

C ²&

 ²&
ON

ON

ON

ON

ON

ON

ON

ON
FEE
FE

n
 #

#

#

n

n
 C

C
The C4 products were
SH

SH
n

n
 





 









&RE




&RE

analysed by gas chromato-


&EED)$ &EED)$
graph, while C5+ liquid
products (syncrude) were
Hydrogen, wt% 50 vol% F
analysed by simulated  
distillation and expressed 

as gasoline, LCO and
$ATA


$ATA

 
bottoms, as shown in Table 

1. The detailed boiling  
point distribution of the
ED

 ²&

 ²&

 ²&

# ²&
D

# ²&

 ²&

# ²&

 ²&

ON

ON

ON

ON
ON

ON

ON

ON
FEE

FE

n 
n

n

#

#

#

bottoms fraction is also


#

#

SH
SH

n
n



 

 















&RE
&RE

shown in Table 1. These


results provide the amount &EED)$ &EED)$

of hydrocarbon in a given
boiling range when an Figure 3 Properties of 650–750°F (340–400°C) recycle fraction obtained from DCR
ideal distillation is runs
achieved. These results
were used as a basis to determine the maximum modified Hempel still (ASTM D295) to obtain
quantity of each recycle stream. the 650°F+ (340°C) fraction. Each 650°F+ frac-
C5+ liquid products from each DCR run were tion was further separated by vacuum distillation
first separated by atmospheric distillation on a (ASTM 1160) to obtain the desired boiling

www.digitalrefining.com/article/1000142 PTQ Q4 2009 3


The ACE runs5 were conducted using the same
Combined feeds used in ACE study
laboratory-deactivated Midas catalyst as above
and the same commercial FCC resid feed as the
Conversion, Recycle Blend ratio Original feed API°
wt% stream, ºF , wt% , wt%
base feedstock. All ACE runs were conducted at
– – – 100.0 20.60 a reactor temperature of 950°F (510°C) using the
54 650–750 8.3 91.7 20.42 same 1.5g of feed and a constant feed delivery
54 650–750 6.3 93.7 20.39 rate of 3.0g per minute. In order to achieve the
54 650–800 11.7 88.3 20.37
54 650–800 9.7 90.3 20.38
desired conversion, the catalyst-to-oil ratio was
54 650–850 13.4 86.6 20.29 varied by changing the amount of catalyst
54 650–850 11.4 88.6 20.30 charged in the reactor in each run. As in the
54 650+ 14.7 85.3 19.83 DCR study, gas and liquid products were anal-
54 750+ 7.1 92.9 19.87
58 650–750 8.3 91.7 20.29
ysed by gas chromatography and simulated
58 650–750 6.3 93.7 20.29 distillation. Coke on catalyst was measured using
68 650–750 7.3 92.7 19.95 a Leco analyser.
68 650–750 5.3 94.7 20.03
75 650–750 5.4 94.6 19.72
75 650–750 3.4 96.6 19.93
Data analysis
In steady operation with HCO recycle, it is
conceivable that some hydro-carbon molecules
Table 2 could go through the riser a number of times. In
our DCR-ACE experiment, we are approximating
fractions. The properties of the various boiling steady-state yields with yields from two-pass
fractions are shown in Figures 2 and 3. cracking. We will use the element-tracking
approach to discuss the validity of this approxi-
ACE of recycle blends mation. Consider 100g of oil, which is fed into
To simulate HCO and bottoms recycle, we the FCCU and cracked into various products, of
prepared feed samples by blending various boil- which the bottoms are partially recycled. For
ing-range fractions back into the starting resid example, 10g of bottoms are recycled and fed
feedstock. These feed blends, listed in Table 2, into the unit again to crack further. Additional
can be separated into two groups. One group products are obtained, and some of the resulting
consists of recycle fractions with various boiling bottoms, (eg, 1g) are recycled and so on. By
ranges obtained at 54 wt% conversion, while the collecting the products along the route of this
other group consists of recycles with one boiling 100g crude oil, we can get the product yields on
range, 650–750°F (340–400°C), but obtained at a fresh feed basis. The process is shown in Figure
various conversion levels from the DCR runs. 4. RR is the recycle ratio, defined as the fraction
The percentage of recycle in each blend was of the recycled HCO stream in the total feed into
selected based on the simulated distillation listed the unit. RR is equal to 0.1 in the following
in Table 1 and the strategy to keep the recycle example.
fraction low enough to use two-pass cracking to Using the element-tracking approach, we can
simulate steady-state operation. The steady-state calculate the LCO yield as follows:
approximation will be discussed further. The
HCO streams were blended at two recycle ratios LCO = LCO1 + RR x LCO2 + RR2 x LCO3 + • • • + RRi-1 x LCOi (1)
to demonstrate the sensitivity and reproducibil-
ity of yield changes due to recycle. where LCOi is the LCO yield of i-th pass cracking
of the recycled portion of (i-1)-th
pass bottoms. Since the quality of
the feed worsens when recycling
further, LCOi < LCOi-1. Obviously,
if the recycling ratio RR is small,
FCC

FCC

FCC

&EED 22
FEED 22
FEED
the second order and above term
of RR can be safely ignored. In
this work, the maximum RR is
Figure 4 Schematic of element-tracking approach 0.15; so, the third term on the

4 PTQ Q4 2009 www.digitalrefining.com/article/1000142


right-hand-side of Equation 1 is
only about 2.25% of the first LCO, wt% Bottoms, wt%
term.  

Therefore, if we can get LCO2,  


we will have a reasonable esti-  
mate for LCO. Looking at the
 
second-pass cracking, the total
 
feed in this pass consists of (1-          
RR) fresh feed and RR recycled 2ECYCLERATIO 2ECYCLERATIO
HCO from the first-pass cracking.
Coke, wt% C5 + gasoline, wt%
This second pass corresponds to
the ACE study in our DCR+ACE 


experiment. Denoting the LCO



yield in the ACE study as LCO’, 
LCO1 as the LCO yield in the 
cracking of the base feed (which 
    
corresponds to the first pass), and 2ECYCLERATIO
    
2ECYCLERATIO
noting that the portion of the
n& 
fresh feed in the combined feed n& 
has the LCO yield of LCO1, we n&

have:
Figure 5 Plots of interpolated yields at 55% conversion vs recycle ratio
LCO’ = (1-RR) x LCO1 + RR x LCO2 (2)

which can be rearranged to: Cat-to-oil ratio C5+ gasoline, wt%




LCO2 = LCO1 + (LCO’ - LCO1)/RR (3) 
$ATA

$ATA

 

The yields on a fresh feed basis 
 
are determined as follows: SE & & &
 & &
SE
"A   
&

& &
 & &

"A  
n n
n  n n    n   
     
LCOFF = LCO’/(1 - RR) (4) &EEDNAME &EEDNAME


BotFF = (Bot’ - RR)/(1 - RR) (5)
LCO, wt% Coke, wt%
 
The element-tracking approach  
$ATA
$ATA


predicts that at low (<15%) recy- 


cling ratio, two-pass cracking is 


very close to the steady-state SE


"A 
&

& &
 & &

"A
S E

& & & & &
 
n n n n n  
operation because the higher- 
 n



 
  
order term in the yield expression &EEDNAME &EEDNAME

is negligible.
Figure 6 Theoretical yields of second-pass cracking at 55% conversion of
Results and discussion recycle streams
Effect of recycle streams
Table 3 shows the interpolated yields of the orig- of the recycle ratio, are plotted in Figure 5. With
inal feed at 70% and 55% conversion, as well as the exception of the 750°F+ (400°C+) recycle
the yields of the combined feeds at 55% conver- feed, all recycle-containing feeds made higher
sion. The yields are expressed as wt% of the total LCO and lower bottoms than the original feed.
feed (fresh + recycle). To better illustrate the With the exception of the 650–750°F (340–
contribution of each recycle stream, the yields of 400°C) recycle feed, all recycle-containing feeds
LCO, bottoms, coke and gasoline, as a function made higher coke and lower gasoline than the

www.digitalrefining.com/article/1000142 PTQ Q4 2009 5


Interpolated yields of base and combined feeds at 70% and 55% conversion

Conversion, wt% 70 55 55 55 55 55 55 55 55 55
Recycle boiling range – – 650–750°F 650–750°F 650–800°F 650–800°F 650–850°F 650–850°F 650°F+ 750°F+
Recycle ratio, wt% 0.0 0.0 8.3 6.3 11.7 9.7 13.4 11.4 14.7 7.1
Cat-to-oil ratio 6.05 3.43 3.52 3.54 3.56 3.60 3.74 3.57 3.71 3.59
Hydrogen, wt% 0.11 0.09 0.08 0.09 0.09 0.09 0.10 0.09 0.10 0.09
Total C1s & C2s, wt% 1.37 0.99 1.02 0.98 1.08 1.07 1.08 1.11 1.04 1.06
Propylene, wt% 3.34 2.08 2.16 2.10 2.17 2.19 2.24 2.21 2.10 2.01
Total C3s, wt% 3.88 2.38 2.48 2.41 2.49 2.51 2.57 2.55 2.44 2.36
Total C4=s, wt% 5.11 3.95 3.91 4.17 3.82 3.90 3.96 3.97 3.87 3.81
Total C4s, wt% 8.50 5.59 5.74 6.03 5.61 5.71 5.87 5.80 5.81 5.51
C5+ gasoline, wt% 49.44 40.63 40.82 40.38 40.15 40.12 39.84 39.80 39.88 40.03
RON 89.61 89.48 89.32 89.55 89.50 89.48 89.45 89.48 89.50 89.26
MON 78.66 77.55 77.65 77.72 77.73 77.74 77.79 77.72 77.81 77.65
LCO, wt% 20.49 24.72 25.75 25.47 25.64 25.45 25.13 24.99 25.01 24.67
Bottoms, wt% 9.51 20.28 19.25 19.53 19.36 19.55 19.87 20.01 19.99 20.33
Coke, wt% 6.73 5.59 5.60 5.57 5.69 5.62 5.70 5.67 6.07 6.09
Relative feed rate
Coke burn limited 1.00 1.20 1.20 1.21 1.18 1.20 1.18 1.19 1.11 1.10
Wet gas limited 1.00 1.53 1.49 1.46 1.49 1.48 1.44 1.45 1.48 1.54
Catalyst circulation limited 1.00 1.76 1.72 1.71 1.70 1.68 1.62 1.69 1.63 1.68
Fresh feed rate
Coke burn limited 1.00 1.20 1.10 1.13 1.04 1.08 1.02 1.05 0.95 1.03
Wet gas limited 1.00 1.53 1.36 1.37 1.32 1.33 1.25 1.29 1.26 1.43
Catalyst circulation limited 1.00 1.76 1.58 1.60 1.50 1.52 1.40 1.50 1.39 1.56

Table 3

original feed. The data quality confirms that the catalyst-to-oil ratios required to achieve the
ACE testing has the sensitivity to measure the same conversion. This is expected, as the easy-
yield contribution of the recycle streams at the to-crack material of the recycle streams has been
desired range of recycle ratios. cracked in the first pass. The crackability of the
Using the element-tracking approach described recycle streams increases with the API gravity
earlier, we can calculate the theoretical yields (see Figure 7). As expected, the 650–750°F
derived from the second-pass cracking of each of (340–400°C) stream made the most LCO and
the recycle streams. This is shown in Figure 6. gasoline and the lowest coke for a given conver-
The recycled streams are less crackable than the sion, when comparing yields among the recycle
base feed, as indicated by the much higher streams. The trends in LCO and gasoline yields
from the lightest stream (650–750°F,
340–400°C) to the heaviest stream
Cat-to-oil ratio vs Coke vs (750°F+, 400°C+) appear to be contin-
API gravity Conradson carbon
  uous and consistent with the trend in
"ASEFEED the API gravity (see Figure 2). However,
 2ECYCLESTREAMS
 the increase of coke appears to be very

gradual up to the 650–850°F (340–

450°C) stream and becomes stepwise
#AT TO OILRATIO


#OKE

 higher for the 650°F+ (340°C+) and



 750°F+ (400°C+) streams. The coke
  yield trends very closely with the
 Conradson Carbon (Figure 7), which is


concentrated in the 850°F+ (450°C+)
          range (see Figure 2). These results
!0)GRAVITY #ONRADSONCARBON
suggest that during first-pass cracking,
coke precursors in the boiling range of
Figure 7 Effect of API gravity and Conradson carbon on catalyst-to- 850°F+ (450°C+) are formed. These
oil ratio and coke yield at 55% conversion molecules are responsible for coke

6 PTQ Q4 2009 www.digitalrefining.com/article/1000142


production during second-pass GC/MS analysis of bottoms fractions from resid cracking in the DCR
cracking.
While the 750°F+ (400°C+) 54 wt% conversion 68 wt% conversion
stream is not a practical recycle 650–750°F 750°F +650–750°F 750°F+
stream, it does provide valuable Saturates AVE wt% AVE wt% AVE wt% AVE wt%
C(N)H(2N+2) paraffins 5.2 4.3 3.4 1.2
insight into the negative impact C(N)H(2N) monocycloparaffins 7.2 8.7 3.6 3.1
of recycling heavy bottoms. This C(N)H(2N-2) dicycloparaffins 4.9 5.4 2.4 2.7
stream made more than double C(N)H(2N-4) tricycloparaffins 3.1 3.9 1.5 2.0
the coke yield of the base feed. A C(N)H(2N-6) tetracycloparaffins 0.0 0.2 0.0 0.1
Total saturates 20.4 22.4 11.0 9.1
close examination of the hydro- Mono-aromatics
carbon compounds by gas C(N)H(2N-6) alkylbenzenes 4.3 6.1 2.6 2.6
chromatography mass spectrom- C(N)H(2N-8) benzocycloparaffins 2.4 2.3 0.6 0.6
etry (GC/MS) (Table 4) shows C(N)H(2N-10) benzodicycloparaffins 1.6 1.1 0.3 0.1
Di-aromatics
that the 750°F+ (400°C+) frac- C(N)H(2N-12) naphthalenes 4.3 2.9 3.4 1.8
tion contains higher aromatic C(N)H(2N-14) 14.5 3.2 15.6 1.9
compounds, and in particular C(N)H(2N-16) 21.2 7.7 24.4 6.4
tetra-aromatic compounds, than Tri-aromatics
C(N)H(2N-18) 13.9 9.3 18.9 9.8
the lighter 650–750°F+ (340– C(N)H(2N-22) 3.3 25.0 4.23 8.5
400°C+) fraction. It is likely that Tetra-aromatics
the coke precursors formed C(N)H(2N-24) 0.0 8.7 0.0 13.6
during first-pass cracking are C(N)H(2N-28) 0.0 1.1 0.0 2.2
Total aromatics 65.3 67.4 70.1 77.5
indeed the tetra-aromatic
compounds. We noticed that Ye
and Wang6 reported slightly less Table 4
coke formation (0.6%) with recy-
cling of highly aromatics bottoms in the FCCU. selected recycle ratios. The following examples
However, their recycling ratio was much lower at demonstrate how to use this data to determine
only 1.5%. the recycle stream and recycle ratio to optimise
LCO production. We will examine a maximum
Modelling overall yields recycle case and a constant bottoms case. A
Table 3 also lists the interpolated yields for modelling of a commercial unit (heat balanced)
maximum gasoline operation at 70 wt% conver- will be presented later.
sion of the base feed. Compared to the yields at
70% conversion, the LCO yield at 55% conver- Maximum recycle
sion is higher, while the yields of wet gas and The goal of the calculation was to maximise the
coke are much lower and the catalyst-to-oil ratio recycle ratio of each recycle stream until the coke
is lower. If the unit changes from maximum yield of the base feed at 70% conversion was
gasoline (70% conversion) to maximum LCO reached. The hydrocarbon yields, on the fresh
(55% conversion) operation, one should be able feed basis, calculated using the element-tracking
to increase the total feed rate until the unit approach, are shown in Table 5. In the cases of
reaches coke burn, wet gas compressor or cata- the 650–750°F (340–400°C) and 650–800°F
lyst circulation constraint, assuming there are no (340–425°C) streams, the maximum available
other limitations. The results of Table 3 suggest recycle levels, based on SIMDIST (Table 1), were
that the coke burn constraint will be reached reached before the coke yield limit was reached;
much sooner than the wet gas or catalyst circula- therefore, the maximum recycle ratio was used.
tion constraints. (Catalyst circulation could be a The highest LCO yield, 30.2%, was achieved
limit at reduced catalyst activity.) At a coke burn with maximum recycle (14.4%) of the 650–800°F
limitation, the combined feed rate of the maxi- (340–425°C) HCO stream. The next highest LCO
mum LCO operation is 10–20% higher than with yield of 29.9% was achieved with 15.6% recycle
the maximum gasoline operation. of the 650–850°F (340–450°C) HCO stream.
The data analyses so far have been confined to Even though the 650–750°F (340–400°C)
yields on the combined feed basis with the stream had the best yields by the element-

www.digitalrefining.com/article/1000142 PTQ Q4 2009 7


Case 1 (max recycle): yields on fresh feed basis (constant coke yield relative to max gasoline case)

Max gasoline base Base no recycle 650–750°F 650–800°F 650–850°F 650°F+


Conversion 70.0 55.0 61.2 64.2 65.2 64.7
Recycle ratio 0 0 0.10 0.14 0.16 0.15
Maximum recycle available – – 0.10 0.14 0.18 0.24
Cat-to-oil ratio 6.05 3.43 3.48 3.56 3.59 3.60
Hydrogen, wt% 0.11 0.09 0.10 0.11 0.12 0.12
Total C1s & C2s, wt% 1.4 1.0 1.1 1.3 1.4 1.4
Propylene, wt% 3.3 2.1 2.4 2.6 2.7 2.7
Total C3s, wt% 3.9 2.4 2.7 2.9 3.1 3.1
Total C4=s, wt% 5.1 3.9 4.5 4.5 4.7 4.8
Total C4s, wt% 8.5 5.6 6.6 6.6 6.9 7.0
C5+ gasoline, wt% 49.4 40.5 44.6 46.8 47.0 46.4
RON 89.6 89.2 89.4 89.5 89.5 89.7
MON 78.6 77.3 77.7 77.8 77.7 77.9
LCO, wt% 20.5 24.7 28.9 30.2 29.9 29.3
Bottoms 9.5 20.2 9.9 5.6 5.0 6.0
Coke, wt% 6.7 5.6 6.1 6.5 6.7 6.7
Relative combined feed rate const coke 1.00 1.20 1.23 1.21 1.18 1.18
Relative fresh feed rate 1.00 1.20 1.10 1.04 1.00 1.00
Relative coke production rate 6.7 6.7 6.7 6.7 6.7 6.7

Table 5

tracking approach, because it was limited to a lower LCO throughput and higher bottoms
maximum recycle ratio of 10.1%, the combined throughput than the operation with recycling
feed with 650–750°F (340–400°C) stream made HCO streams of 650–800°F (340–425°C) and
only 28.9% LCO and much higher bottoms. In 650–850°F (340–455°C). Thus, it is advanta-
the case of the 650°F+ (340°C+) bottoms stream, geous to recycle HCO rather than bottoms.
due to coke limitation, only 15% out of the avail- Gasoline yields on fresh feed base for all the
able 24% recycle stream could be recycled. The recycling streams are about 4–6% higher than
operation recycling 650°F (340°C) bottoms made that of the case without recycling, which corrob-
orates the results reported by
Fernandez et al.7
Max. recycle ratio C5+ gasoline, wt%
 Constant bottoms

n The goal of this calculation was to

n adjust the recycle ratio of each recy-
 n cle stream until the bottoms yield of
n the base feed at 70% conversion was

     reached. The hydrocarbon yields, on
    
#ONVERSION #ONVERSION the fresh feed basis, are shown in
Table 6. In this case, all the
combined feeds with HCO recycle
LCO, wt% Coke, wt%
had higher LCO selectivity than
 
bottoms (650°F+, 340°C+) recycle.
 
The difference also comes from the
  coke yield differences, which allow
  the feeds with HCO recycle to be
  processed at higher feed rates than
          the feed with bottoms recycle. Again,
#ONVERSION #ONVERSION
this example shows it is advanta-
geous to recycle HCO rather than
Figure 8 Yields of second-pass cracking minus fresh feed cracking vs bottoms.
conversion

8 PTQ Q4 2009 www.digitalrefining.com/article/1000142


Yields on fresh feed basis (constant bottoms yield relative to max gasoline case)

Max gasoline base Base no recycle 650–750°F 650–800°F 650–850°F 650°F+


Conversion 70.0 55.0 61.2 62.0 62.4 62.5
Recycle ratio 0.00 0.00 0.101 0.112 0.118 0.120
Maximum recycle available – – 0.10 0.14 0.18 0.24
Cat-to-oil ratio 6.0 3.4 3.48 3.53 3.56 3.56
Hydrogen, wt% 0.1 0.1 0.10 0.10 0.11 0.11
Total C1s & C2s, wt% 1.4 1.0 1.1 1.2 1.3 1.3
Propylene, wt% 3.3 2.1 2.4 2.5 2.5 2.6
Total C3s, wt% 3.9 2.4 2.7 2.8 2.9 3.0
Total C4=s, wt% 5.1 3.9 4.5 4.4 4.5 4.6
Total C4s, wt% 8.5 5.6 6.6 6.4 6.6 6.8
C5+ gasoline, wt% 49.4 40.5 44.6 45.1 45.1 44.9
RON 89.6 89.2 89.4 89.5 89.5 89.6
MON 78.6 77.3 77.7 77.7 77.7 77.8
LCO, wt% 20.5 24.7 28.9 28.6 28.2 28.0
Bottoms, wt% 9.5 20.2 9.9 9.5 9.4 9.5
Coke, wt% 6.7 5.6 6.1 6.2 6.4 6.4
Relative combined feed rate const coke 1.00 1.20 1.23 1.21 1.19 1.19
Relative fresh feed rate 1.00 1.20 1.10 1.08 1.05 1.04
Relative coke production rate 6.7 6.7 6.7 6.7 6.7 6.7

Table 6

Effect of conversion level can be explained by examining the properties of


The objectives of this work were to determine the recycle steams in Figure 3. Generally speak-
how the composition of the HCO stream changed ing, higher cracking severity in FCCUs leads to
with conversion and how recycling HCO, more gasoline, but a much higher concentration
obtained at varying conversion levels, affected of condensed aromatics in the bottoms.8
the LCO yield. As described earlier, DCR Although the 50 vol% boiling points are about
syncrude samples obtained at 54, 58, 68 and 75 the same for each stream, the API gravity and
wt% conversion were distilled, and the 650– hydrogen content decrease with increasing
750°F (340–400°C) fraction of each syncrude conversion. This is consistent with the mass
was collected and analysed (see Figure 3). The spectrometry data in Table 4, which shows the
650–750°F (340–400°C) fractions were blended tri-aromatics of the 650–750°F (340–400°C)
with the original feed and tested in the ACE. stream, obtained at 68% conversion, is much
The theoretical yields of the second-pass crack- higher than that at 54% conversion.
ing on a unit recycle feed basis are calculated Figure 9 shows the yields of gasoline, LCO,
using the element-tracking approach. The differ- bottoms and coke as a function of conversion for
ence in the yields of gasoline, LCO and coke cracking of only the base feed (first cracking).
between the second-pass cracking of the recycle The same figure also shows the corresponding
stream and cracking of the fresh feed is shown yields, normalised to the fresh feed basis, for
in Figure 8. The maximum recycle ratio at each cracking of the combined feed (base feed + maxi-
conversion, calculated based on simulated distil- mum recycle of the 650–750°F (340–400°C)
lation, is also plotted in Figure 8. At lower stream at each conversion level). At a given
conversion, there is more 650–750°F (340– conversion, by recycling the 650–750°F (340–
400°C) fraction available for recycle. The 400°C) fraction, one can lower bottoms and
low-conversion recycle stream made much increase LCO without sacrificing gasoline and
higher LCO than the fresh feed, while making with only a minor penalty in coke. This is
about the same gasoline and coke. However, at achieved for two reasons. First, recycling lowers
higher conversion, there is less 650–750°F bottoms because part of the bottoms is cracked
(340–400°C) stream available. The high-conver- twice. Second, the second-pass cracking of the
sion recycle stream made much lower gasoline, 650–750°F (340–400°C) stream is very selective
similar LCO and much higher coke. These results toward producing LCO and, while coke increases,

www.digitalrefining.com/article/1000142 PTQ Q4 2009 9


yield shift by lowering conversion on
the combined feed basis and select-
Bottoms, wt% C5+ gasoline, wt%
ing the proper recycle stream.
 

  Midas catalyst technology


 Application of the correct catalyst

technology is critical to ensure a

 high LCO yield and minimal bottoms
          and coke yield. A balanced approach
#ONVERSION #ONVERSION
is required to achieve maximum
bottoms upgrading to LCO and other
LCO, wt% Coke, wt% valuable products. Grace Davison

 .ORECYCLE Midas catalysts have been proven to
2ECYCLEn

2ECYCLEALLn& minimise bottoms production.9 A

 Midas catalyst was used for the
experiments described above.
 
Commercialised in 2008, the
          Midas-300 catalyst series is designed
#ONVERSION #ONVERSION for a distillate-driven market. Midas
catalyst has been used in 52 units
Figure 9 Yields vs conversion of base feed and combined feed since its first introduction. The
normalised to fresh feed basis Midas-100 series of catalyst was
introduced in 2005, and in 2007 the
so does conversion; consequently, coke selectiv- Midas-200 series was commercialised.
ity is not much higher than that of the first-pass Midas-300 catalysts feature a high-activity
cracking of the base feed. matrix surface area, balanced with optimised
The results from the maximum recycle case zeolite content. The majority of matrix porosity
(see Table 5) of 650–800°F (340–425°C) and in Midas technology is found in the 100–600 Å
650–850°F (340–455°C) recycle are also plotted pore diameter range for high LCO selectivity.
in Figure 8. Compared to recycling with the LCO selectivity is maximised via the three-step
650–750°F (340–400°C) stream, the LCO gain bottoms cracking mechanism originally described
from recycling these heavier streams is higher; by Zhao10 (Figure 10). Each of these bottoms
however, the coke penalty is greater. These cracking mechanisms is particularly important in
results suggest that one can achieve the desirable lower conversion operations to ensure a high
LCO yield and low bottoms.
It is also important that feed
vapourisation is maintained when
#ATALYTIC
the FCCU operates at a reduced
reactor temperature to maximise
Type II Type II
LCO. At a low reactor temperature,
2 2
optimisation of Type I cracking
becomes more important because of
the reduction in the riser mix zone
Type I Coke
-IDASCATALYSTISTHE temperature. Catalyst design plays
Feed
MOSTEFFECTIVECATALYST an important role in maintaining the
4HERMALCATALYTIC FOR4YPE)AND)))
right conditions. Since resid feeds
contain a high percentage of mole-
4YPE) 0RECRACKINGANDFEEDVAPOURISATION cules boiling above the mix zone
4YPE)) $EALKYLATIONOFALKYLAROMATICS
4YPE)) #ONVERSIONOFNAPHTHEOAROMATICS
temperature, pre-cracking is neces-
sary to achieve complete
vapourisation. Porosity in the 100–
Figure 10 Bottoms cracking fundamentals 600 Å range is essential for the

10 PTQ Q4 2009 www.digitalrefining.com/article/1000142


pre-cracking reactions Commercial maximum gasoline and maximum conversion modelling results
that enable vapourisation.
Midas-300 catalysts
ensure feed is properly Case Base 1 2
Operating mode Maximum gasoline Maximum LCO Fully optimised maximum
vapourised even at low Catalyst/additive Midas-100 Midas-100 Midas-300 and OlefinsUltra
severity. Recycle % FF, vol% 0 11 11
Most of the LPG and Ecat activity 68 64 64
gasoline produced in an Feedstock props
API 21.6 21.6 21.6
FCCU comes from the Conradson carbon, wt% 3.0 3.0 3.0
dealkylation of aromatics 1050°F+, vol% 20 20 20
or Type II cracking. Operating conditions
Zeolite is much more Reactor temperature, °F 995 950 950
Feed temperature, °F 400 405 405
effective than matrix in Regenerator temperature,°F 1350 1290 1290
cracking long-chain alkyl Cat-to-oil 7.2 7.8 7.8
aromatics. Type II crack- Air blower Base Base Base
ing is important to reduce Wet gas comp. rate Base 0.75 Base Base
Product yields and props.
the molecular size and Dry gas, scfb 331 235 235
promote eventual conver- LPG, vol% FF 23.9 19.3 30.0
sion of bottoms; however, Gasoline, vol% FF 56.7 51.9 44.0
we must prevent any LCO RON/MON 92.6/80.6 90.0/79.5 92.9/80.7
LCO, vol% FF 22.9 32.0 33.4
that is produced from Bottoms, vol% FF 6.8 6.0 5.0
being over-converted to Coke, wt% FF 5.2 5.2 5.2
lighter components. The C3+, vol% 110.3 109.2 112.4
zeolite level in Midas-300 Incremental product value, $/B Base +0.10 +1.40
has been optimised to
provide sufficient dealkyl- Table 7
ation activity, yet
maintain the product yield as LCO rather than Recycle and Midas technology
LPG and gasoline. A commercial operation was modelled using
Finally, Type III cracking destroys naphthene results from the work described above together
rings in naphthenoaromatic compounds. The with Midas-300 technology. Table 7 shows a
typical size of naphthenoaromatic molecules is maximum gasoline/conversion FCC operation
too large to fit easily into the zeolite. The crack- with a residual feedstock. Two cases are modelled
ing of these molecules will occur on the matrix from this base operation. Case 1 is a reduced
sites or on the external surface of the zeolite. The conversion operation together with a recycle of
selective cracking of this type of molecule 11 vol% HCO 650–800°F (340–425°C). Case 2
requires proper design of matrix activity and the represents a fully optimised maximum LCO
interaction of matrix and zeolite. The high meso- operation. Cases 1 and 2 were modelled assum-
porosity of Midas-300 catalysts improves LCO ing the base maximum gasoline/conversion
selectivity by converting coke precursors into operation was operating at an air blower and wet
liquid product. gas compressor constraint.
Grace Davison can also deliver enhanced LCO January 2009 FCC product values were used
selectivity in an additive form. BX-450 is Grace’s to assess the relative product values for Cases 1
newest catalytic additive offering and is the first and 2. LCO is valued at $8.00/bbl greater than
of its kind, designed specifically for maximum gasoline.
distillate yield. BX-450 is based on Midas-300 LCO is increased in Case 1 by reducing conver-
technology and offers a high-activity matrix sion via lower reactor temperature, lower
surface area balanced with an optimised zeolite equilibrium catalyst (Ecat) activity, higher feed
level to maximise LCO selectivity. The optimised temperature and the incorporation of 11 vol%
zeolite level in BX-450 provides sufficient cata- recycle. Coke yield is the same as the base opera-
lytic activity, enabling 1:1 replacement of fresh tion. Lower reactor temperature and higher
catalyst. feedstock temperature reduce the unit coke

www.digitalrefining.com/article/1000142 PTQ Q4 2009 11


demand, allowing the incorporation of recycle at Recycle can be employed to fully maximise
the same air blower demand as the base opera- LCO at reduced conversion, while maintaining
tion. Coke yield or air blower rate will be the bottoms equal to a traditional maximum gasoline
primary constraint. The FCC catalyst in the base operation. The crackability and LCO yield
and Case 1 operation is Midas-100 catalyst. produced by a particular recycle stream are
Case 1 has a higher catalyst-to-oil ratio due to consistent with its API gravity and hydrogen
the introduction of recycle and lower catalyst content. As a result of reduced first-pass crack-
activity, and results in increased bottoms conver- ing conversion, the potential of recycle streams
sion to LCO. The wet gas rate is lower in Case 1 to produce LCO with minimal coke production
as a result of lower LPG and dry gas production. are improved over high-conversion operations.
Despite a 9 vol% increase in LCO yield, the The 650–750°F (340–400°C) stream, when
Case 1 product value is only a modest $0.10/bbl recycled, produces the most LCO and gasoline
greater than for the base operation. Reduced C3+ and the lowest coke for a given conversion.
vol% and gasoline octane compared to the base However, it is not produced at sufficient quanti-
operation hinder the total product value. ties to fully maximise LCO. High Conradson
Case 2 represents a fully optimised operation carbon levels, consistent with higher tetra-
using Midas-300 catalyst and OlefinsUltra ZSM- aromatic and heavier compounds, limit the yield
5 additive. LCO is increased at the expense of of LCO when 650+°F or 750+°F (340°C+ or
400°C+) streams are recycled. The 650–800°F
(340–425°C) or 650–850°F (340–455°C) recycle
To fully maximise LCO from stream produces the highest LCO when recycled
against a coke burn and bottoms constraint.
the FCCU, the refinery should
consider several strategies in References
parallel 1 Ritter R E, Creighton J E, Producing light cycle oil in the cat
cracker, Catalagram, 69, 1984, 5.
2 Energy Information Administration short-term outlook Jan
2009.
bottoms due to improved bottoms cracking with
3 Wallenstein D, Harding R H, Nee J R D, Boock L T, Recent
Midas-300 catalyst. OlefinsUltra additive is used
Advances in the deactivation of FCC catalysts by cyclic propylene
to increase the gasoline octane, which dropped steaming in the presence and absence of metals, Appl.Catal. A:
sharply from the base operation due to a lower General, 204, 2000, 89.
reactor temperature. The wet gas rate increases 4 Young G W, Weatherbee G D, FCCU studies with an adiabatic
compared to the base operation due to additional circulating pilot unit, AIChE Annual Meeting, November 1989.
LPG production. Air blower and wet gas 5 Kayser J C, Versatile Fluidised Bed Reactor, US Patent 6 069
compressor rates in Case 2 are identical to the 012.
base operation. 6 Ye A, Wang W, Cracking performance improvement of FCC
The optimised Case 2 product value has feedstock by adding recycle stock or slurry, Lianyou Jishu Yu
Gongcheng, 2004, 34, 6, 5–6.
increased $1.40/bbl relative to the base opera-
7 Fernandez M L, Lacalle A, Bilbao J, Arandes J M, de la Puente G,
tion. These product values are achieved despite a
Sedran U, Recycling hydrocarbon cuts into FCCUs, Energy & Fuels,
low value of propylene, and confirm the need to
16, 2002, 615.
maintain liquid yield and gasoline octane while 8 Venugopal R, Selvavathy V, Lavanya M, Balu K, Additional
operating in a maximum LCO mode. feedstock for fluid catalytic cracking unit, Petroleum Science and
Technology, 26, 2008, 436.
Conclusions 9 Schiller R, et al, The Genesis catalyst system, Catalagram, 102,
Maximising the LCO yield is largely a bottoms Fall 2007.
management process. Operating at reduced 10 Zhao X, et al, FCC Bottoms cracking mechanisms and
conversions by shifting operating conditions and implications for catalyst design for resid applications, NPRA, AM-
reducing Ecat activity increases LCO, but also 02-53.
11 Corma A, Sauvanaud L, How can we increase the LCO yield
increases bottoms. To fully maximise LCO from
and quality in the FCC: cracking pathways analysis, Preprints
the FCCU, the refinery should consider several
— American Chemical Society, Division of Petroleum Chemistry,
strategies in parallel.
2006, 51, 2, 447–451.

12 PTQ Q4 2009 www.digitalrefining.com/article/1000142


David Hunt is Technical Service Manager with Grace Davison,
Houston, Texas. Email: david.hunt@grace.com Links
Ruizhong Hu is Manager of Research and Technical Support with
Grace Davison, Columbia, Maryland. More articles from: Grace Davison
Hongbo Ma is a Research Engineer with Grace Davison, Columbia.
More articles from the following categories:
Larry Langan is a Research Engineer with Grace Davison,
Catalysts & Additives Fluid Catalytic Cracking
Columbia.
Diesel Maximisation
Wu-Cheng Cheng is Director of R&D with Grace Davison,
Columbia.

www.digitalrefining.com/article/1000142 PTQ Q4 2009 13