103
103
103
This issue of the Catalagram once again finds us discovering the synergies between Grace
Davison and ART products and technologies and applying them to the challenges refiners face.
Our lead article, “New Opportunities for Co-Processing Renewable Feeds in Refinery Process”,
presented at the 2008 NPRA Annual Meeting, is a collaboration between our Refining
Technologies, ART, and Biofuels Technologies groups. The article explores the ramifications of
co-processing renewable fuels in a conventional FCC or DHT unit with traditional straight run
diesel or cat feed. Extensive pilot plant testing over a wide range of feed blends and conditions
resulted in valuable data that can aid refiners in designing the optimum configuration to maxi-
mize their profitability.
More exciting news about the commercial performance of our innovative catalysts continues to
come in from the field. We have introduced two new high activity GENESIS™ catalyst compo-
nents and have made improvements to our MIDAS® catalyst and PINNACLE® catalyst families,
which will deliver higher activity and improved gasoline selectivity at equivalent bottoms and
coke yields. ART’s 420DX™ catalyst, the newest member of the ultra high activity DXTM catalyst
series, is expected to exceed refinery expectations in its ability to tolerate difficult feed blends in
demanding ULSD applications.
Please join us in welcoming our new Vice President and General Manager for Refining
Technologies, Shawn Abrams. Shawn will run the global FCC business, while Bob Bullard con-
tinues as Managing Director of ART and focuses his Davison responsibilities on joint ventures
across the businesses. This added depth to our team is yet another way that Grace Davison
and Advanced Refining Technologies demonstrate our strong commitment to the future of the
refining industry.
Joanne Deady
Vice President
Marketing/R&D
Grace Davison Refining Technologies
IN THIS ISSUE
NU M B E R 1 0 3 Spring 2008
New Opportunities for Co-Processing Renewable Feeds in Refinery Processes
By Brian Watkins, Supervisor, Laboratory Technology, Advanced Refining Technologies; Charles
1
Olsen, Worldwide Technical Services Manager, Advanced Refining Technologies; Kevin Sutovich,
Senior R&D Chemist, Grace Davison Refining Technologies; Natalie Petti, Vandelay Management
In the ever growing market of transportation fuels, we demonstrate the effectiveness of
an ART hydroprocessing catalyst and state-of-the-art FCC catalysts in co-processing
renewable oils (biofeeds) as possible new feedstock opportunities for hydrotreaters and
FCC units.
Managing Editor: Effect of Hydrocarbon Partial Pressure on Propylene Production in the FCC
By Ruizhong Hu, Sr. Principal Scientist; Gordon Weatherbee, Principal Engineer; Hongbo Ma, 22
Joanne Deady Research Engineer; Terry Roberie, Director FCC Evaluations; Wu-Cheng Cheng, Director R&D,
Grace Davison Refining Technologies
Contributors: We have been able simulate the effect of varying hydrocarbon partial pressure in the
FCC unit in the Grace Davison DCR Unit. The results indicate that increasing the
Mike Beshara
hydrocarbon partial pressure increases the hydrogen transfer activity. The result is a
Alan Birch
decrease in the olefinicity and olefins yield of LPG and gasoline, and a decrease in the
Wu-Cheng Cheng
concentration of gasoline sulfur species.
Ruizhong Hu
Garry Jacobs
Meet Clean Fuels Challenges with Advanced Refining Technologies
Al Jordan
Newest ULSD Catalysts – ART 420DXTM 33
Adam Kasle
By Brian Watkins, Supervisor, Laboratory Technology; Charles Olsen, Worldwide Technical
Doc Kirchgessner Services Manager; Dave Krenzke, Technical Services Manager, Advanced Refining Technologies
Ernst Köhler ART has developed a new technology catalyst that will provide the refiner enhanced
Dave Krenzke sulfur removal activity with greater flexibility in meeting their HDS activity requirements
Hongbo Ma while minimizing hydrogen consumption.
Charles Olsen
Natalie Petti
Ben Prins Successful Implementation of State-of-the-Art ULSD/Dewaxing Technology
Terry Roberie
Greg Rosinski
at Irving Oil, Saint John, NB
By Mike Beshara, Project Manager, Irving Oil; Greg Rosinski, Technical Services Engineer,
36
Rosann K. Schiller Advanced Refining Technologies; Charles Olsen, Worldwide Technical Services Manager,
Kelly Stafford Advanced Refining Technologies; Ben Prins, Senior Process Engineer, Fluor; Garry Jacobs,
Kevin Sutovich Technical Director, Fluor; Alan Birch, Account Manager, Süd Chemie; Ernst Köhler, Global Product
Brian Watkins Manager-Zeolites, Süd Chemie
Gordon Weatherbee Irving Oil decided to convert the existing VGO Hydrocracker/LCO Desulfurizer at the
Saint John refinery in New Brunswick, Canada to an LCO /heavy diesel ULSD unit. An
integrated approach between Grace, Süd-Chemie and Fluor met Irving Oil’s request for
Please address
a process to produce 7 ppm sulfur in diesel within cloud point specifications while
your comments to
allowing them to “turn off” the dewaxing function during the summer months.
betsy.mettee@grace.com
Grace Davison Multi-Loader System
By Al Jordan, Director, Sales Operations; Adam Kasle, Technical Sales Manager, Grace Davison 43
Refining Technologies
Grace Davison’s Multi-Loader System provides an effective reliable means of adding
FCC fresh catalyst and additives on a continuous basis, whether you are adding one or
several materials to the FCC.
www.e-catalysts.com
©2008 The information presented herein is derived from our testing and experience. It is offered, free of charge, for your considera-
tion, investigation and verification. Since operating conditions vary significantly, and since they are not under our control, we
W. R. Grace & Co.-Conn.
disclaim any and all warranties on the results which might be obtained from the use of our products. You should make no
assumption that all safety or environmental protection measures are indicated or that other measures may not be required.
New Opportunities for
Co-Processing Renewable
Feeds in Refinery
Processes
T
processed renewable diesel. Some
Supervisor, based sources of feed to pro- common sources of renewable
Laboratory Technology, duce fuels is becoming more feeds are those produced for food
Advanced Refining Technologies widely employed as a means of grade oils such as soybean, rape-
decreasing dependence on non- seed and other vegetable oils. The
renewable fossil fuel sources. There traditional process for introducing
Charles Olsen are typically three common produc- these sources into the diesel pool is
Worldwide Technical Services tion routes for biodiesel. Fuel which is to use the transesterification reac-
Manager, Advanced Refining produced by the FAME (Fatty Acid tion for breaking the glycerol from
Technologies Methyl Ester) process to meet a fuel the fatty acid chains. This reaction
specification of ASTM D6751 is con- requires the use of an alcohol (such
sidered biodiesel. Fuels produced as methanol) and a catalyst (such
Kevin Sutovich from biological material using thermal as sodium or potassium hydroxide,
Senior R&D Chemist, depolymerization to meet ASTM D975 NaOH or KOH) in order to break the
Grace Davison Refining or ASTM D396 are considered renew- long chained fatty acids apart from
Technologies able diesel. Fuels that are produced the glycerin molecule. (Figure 1)
when vegetable oils or animal fats are
processed in traditional refining
Natalie Petti processes are considered co-
Vandelay Management
Table I
Composition of Various Oils and Fats1,2
Fats and Oils Butyric Caproic Caprylic Capric Lauric Myristic Palmitic Palmitoleic Stearic Oleic Linoleic Linolenic Behenic & <C16:1 others
Molecular wt. 88 116 144 172 200 228 256 254 284 282 280 278 326 226 324
Tallow, wt.% 3 27 2 24.1 40.7 2 0.7 0.3
Lard, wt.% 1 26 2 13 45.2 10.3 2.5
Butter, wt.% 3.5 1.5 25 3 11 30 3.5 12 26 3 1.65 1.5 0.85
Coconut, wt.% 8 8 48 16 8.5 2.5 6.5 2 0.5
Palmkernel, wt.% 3 5 48.5 17 7.5 0.5 2 14 1 1.5
Palm, wt.% 3.5 395 3.5 46 7.5
Safflower, wt.% 52 2.2 76.4 16.2
Peanut, wt.% 0.5 7 1.5 4.5 52 27 7.5
Cottonseed, wt.% 1.5 19 2 31 44 2.5
Maize, wt.% 1 9 1.5 2.5 40 45 1
Olive, wt.% 1 13 2 2 68 12 0.5 1
Sunfower, wt.% 6 4.2 18.7 69.4 0.3 1.4
Soy, wt.% 0.3 7.8 0.4 2.5 26 51 5 7
Rapeseed, wt.% 3.5 0.2 2 13.5 17 7.5 0.9 56.3
Mustard, wt.% 4 3 1.5 39.5 12 8 36
Codliver Oil, wt.% 0.2 10 14.5 0.5 28 1 42
Linseed, wt.% 6 5 17.3 16 55 0.5
Tung, wt.% 8 12 80
2 www.e-catalysts.com
as normal paraffin components in Figure 2
the 500-650˚F boiling range. These Simulated Distillation (D2887) of Soybean oil
n-paraffins can be of significant
value for ULSD as they have typi-
cal cetane numbers ranging from Chromatogram: Boiling point (°F)
Signal
In the unbroken, unprocessed form,
1.000e+006
the triglyceride molecules are signif-
icantly outside the diesel pool range 8.000e+005
Start Time
End Time
2.000e+005
analysis of soybean oil is shown in
FBP
IBP
Figure 2 and indicates that these 0.000e+000
0 5 10 15 20 25 30 35 40
materials have a fairly narrow distil- Retention time (min)
lation showing up in the C50-C60
range. Note that simulated distilla-
tion of these compounds is based
on the carbon content and molecu-
lar weight of the materials and this
can sometimes skew the estimated
boiling points. Biofeed sources typ- Table II
ically have a true boiling point that is Analysis of Different Biofeed Sources
much lower than that reported by
simulated distillation equipment due Soybean Rapeseed *Palm
to molecular weight interference. In Oil Oil Oil
the unconverted state these triglyc- API (°) 21.58 21.98 22.98
eride molecules cannot be blended Specific Gravity (g/cc) 21.6 22.0 23.0
into the diesel pool at the levels Sulfur, ppm 0 3 1
required to meet renewable fuel Oxygen, wt.% 10.5 10.62 11.33
standards. Nitrogen, ppm 3.9 16 1.6
D2887 Distillation, °F
Another concern is that these IBP 702 710 625
renewable feed sources can 5% 1059 1065 941
include various contaminants. An 10% 1069 1077 1026
analysis of several different bio- 30% 1090 1095 1062
feed sources has indicated the 50% 1102 1106 1079
presence of contaminants such as 70% 1111 1115 1090
sodium, calcium and phosphorus. 90% 1183 1188 1146
Table II shows the measured con- 95% 1232 1238 1197
taminant levels of the soybean, FBP 1301 1311 1302
rapeseed oil and palm oils used in Metals Contamination,
this work. The palm oil shows no ppm
trace impurities, which indicates Na 2.0 4.7 0.0
that it has been previously Ca 3.0 13.8 0.0
processed while the soybean and Mg 0.9 0.3 0.0
rapeseed oils have not. In the fore- P 6.5 4.0 0.0
seeable future it is unlikely that the Zn 0.1 0.6 0.0
use of these renewable sources Al 0.1 0.2 0.0
Mn 0.0 0.1 0.0
* Oil was pre-processed to remove impurities
=
O
+ 6H2
capacity guard materials and Grace O
=
H C -O-C
Davison specialty catalysts are O
H C -O-C
=
capable of protecting the down- H
HEAT
=
O
Since these renewable feeds are O
H H
=
H C -O-C
derived from a biological source, O H H
H C -O-C H
=
they also contain a high concentra- H H
H
+ 3 H2O H H H
tion of oxygen. For the materials
listed in Table I the oxygen content
ranges from 10 to 15%, and is H
H C -H H
H
entirely dependent on the length H C -H H HEAT
H
and degree of saturation of the fatty H C -H +
H
H H 4.5H2
acid chains. + 6H2
+ 3 O2
4 www.e-catalysts.com
Figure 5
Chromatogram of Sulfur, Carbon and sources of feedstock at the target-
Oxygen in a Co-Processed Product ed 10% level. Figure 6 summarizes
some of the results of the testing.
The testing showed that soybean
and rapeseed oils behave similarly
when co-processed in a SR diesel.
The feed blends required essential-
ly the same temperature for 10 ppm
product sulfur, and the apparent
activation energy (temperature
response) for the two feed blends is
similar to that of the SR feedstock
alone. The palm based oil, which
had been previously processed,
was apparently easier to treat to low
sulfur diesel levels, but for 10 ppm
product sulfur the temperature was
only slightly lower than that for the
Analytical techniques using GC- sulfur, nitrogen and aromatic contents SR feed. The apparent activation
AED have shown that in normal and decrease the API gravity. energy for this feed blend was lower
ULSD operation, no oxygen is than the SR component indicating
detected in the products at levels ART then conducted testing on blends the temperature response in the unit
below 500 ppm sulfur. This can be containing the various renewable was lower. Comparing the feeds at
seen in Figure 5 which shows the
analysis of one of the co-processed
products which has a total sulfur of Table III
31 ppm and less than 1 ppm nitro-
Straight-Run (SR) and Bio-Blend Analyses
gen.
SR 10% 10% 10% 40% 80%
Pilot Plant Testing of Renewable Oil Soybean Palm Rapeseed Soybean Soybean
Oils
API 34.44 33.03 33.50 33.29 29.24 24.38
In order to understand the process Specific 0.852 0.859 0.857 0.858 0.879 0.907
for co-treating renewable fuel com- Gravity, g/cc
ponents in a hydrotreater, Advanced Sulfur, wt.% 1.123 1.083 1.092 1.042 0.670 0.210
Refining Technologies completed a Ni trogen, 130 82 75 67 47 16
number of pilot plant studies. A ppm
wide range of ULSD operating con- Oxygen, 0.0 1.2 1.2 1.1 4.7 8.9
ditions were investigated to deter- wt.%
mine if there is an optimal operating Aromatics,
window for processing these types wt.%
of feeds. The conditions included Mono 17.76 15.85 15.87 15.86 10.32 3.34
hydrogen pressures from 450 to Di 7.39 6.60 6.60 6.60 4.29 1.39
1100 psia and hydrogen to oil ratios Poly 2.1 1.87 1.88 1.87 1.22 0.39
of 1000 to 3000 SCFB. Total 27.25 24.32 24.34 24.33 15.84 5.12
D2887
The three different renewable Distillation,
sources of oil were blended in sepa- ˚F
rately with a typical straight run (SR) IBP 222 209 209 210 239 329
diesel feedstock. The renewable 10% 477 465 459 465 498 571
component level was varied from
30% 579 559 557 559 579 1009
10% to 80% and hydrotreated over
50% 613 595 592 594 631 1119
the range of processing conditions
70% 643 632 628 630 1108 1130
listed above. The SR component
properties are listed in Table III, along 90% 681 720 688 715 1130 1135
with 5 different blends of the bio FBP 740 1127 1121 1127 1134 1139
components. As can be seen in the Cloud 19.9 21.7 24.1 22.8 22.0 19.1
table, the effects of blending in the Point, ˚F
renewable source are to dilute the Cetane Index 53.8 50.6 51.2 50.9 46.4 NA
61
shown in Figure 8 for a 10% renew-
able feed blend. Not surprisingly,
60
lower pressure operation results in a
59
lower cetane index for the SR feed,
58 but the addition of the renewable oil
57 again provides a consistent two
56 number increase in cetane index.
55
570 580 590 600 610 620 630 This is a good indicator that the
WABT, ˚F large fatty acid molecules are being
broken down into the three individ-
SR 10% Soy 10% palm 10% rapeseed ual fatty acid chains via the break-
ing of the C-O bonds. Figure 9
Figure 8 compares the D-2887 distillation
chromatograms of the SR products
Cetane Boost when Co-Processing Bio-Feeds
at 10 ppm sulfur to that of the co-
at Low Pressure processed products, and it is evi-
dent that there is an increase in the
60 concentration of the n-paraffins
between 500°F and 600°F boiling
Product Cetane Index
6 www.e-catalysts.com
Due to the addition of unsaturated
Figure 9
chains from the bio component,
there is expected to be an increase Boiling Point Comparison Between SR and 10% Bio Blends
in hydrogen consumption to satu-
rate these C=C bonds. With this
additional hydrogen usage, it is Chromatogram: Boiling point (°F) Chromatogram: Boiling point (°F)
7.000e+005
Signal
Signal
5.000e+005
IBP Time
bio-blended feeds. The total aro-
Start Time
End Time
End Time
1.000e+005 2.000e+005
Start
FBP
FBP
IBP
matics are consistently two num- 0.000e+000
0 5 10 15 20 25 30 35 40
0.000e+000
0 5 10 15 20 25 30 35 40
Straight-run product
Retention time (min)
Co-processed product
is the same as the actual difference @ 10 ppm sulfur @ 10 ppm sulfur
in the total aromatic content of the
two feeds. The lower aromatic con-
tent of bio-blended feeds allows the Figure 10
refiner to achieve lower product aro- Comparison of Total Aromatics of SR Oil with
matic content, which may be valu-
10% Renewable Oil at High Pressure
able as future regulations may
require a lower total aromatic limit
on diesel fuel. 30
Total Product Aromatics, wt.%
8 www.e-catalysts.com
Figure 13
Conversion vs. Coke potentially allowing for downstream
hydrotreating benefits (less gaso-
Biofeed Pilot Plant Study line octane loss, extended catalyst
81.0
80.0
run length) or for a lower cost base
79.0 FCC feed at constant product sulfur.
78.0
Conversion, wt.%
Propylene
Total C3's
Total C4='s
Total C4's
C5+ Gasoline
RON
MON
Gaso. Isoparaffins
Gaso. Aromatics
Gaso. Napthenes
Gaso. Olefins
LCO
Bottoms
Coke
Conversion
can be evaluated by comparing the
propane yield for the feeds tested.
In cracking a biofeed, the fatty acid
molecules that were liberated by the
Figure 15 initial thermal cracking step will sub-
Hydrogen Yield vs. Conversion sequently crack along the pathways
Biofeed Pilot Plant Study defined for either paraffin or olefin
0.55
molecules. If the fatty acid mixture
0.50
is more olefinic, it will be very reac-
tive and will easily crack to produce
0.45
smaller gasoline range olefins6. A
0.40
Hydrogen, wt.%
CARBENIUM ION
ß OLEFIN
reduction in propane and hydrogen
SCISSION
+A
(Product)
is potentially due to reduced pro-
LK
AN
E ALKANE PROPAGATION tolytic cracking of the fatty acids
(Product)
ORM OLEFIN
compared to a typical FCC feed.
RE F
- H+ B RO
NST
ED (Product)
TERMINATION
This also indicates that the double
CARBENIUM ION
+ H+
REF
ORM
ALKANE bonds on each of the free fatty acid
LEW (Product)
IS
molecules, which are more reactive,
are the initiation sites for the crack-
Catalytic Cracking Catalysts, Chemistry and Kinetics, Chemical
Industries 25, reprinted by permission ing reactions.
10 www.e-catalysts.com
Figure 17
The fatty acid molecules are crack- Paraffin Reaction Pathway
ing into gasoline and propylene, as
can be seen in Figures 19 and 20.
0.6
The influence of the bio-based 71.0 72.0 73.0 74.0 75.0 76.0 77.0 78.0 79.0 80.0 81.0
materials on the gasoline properties Conversion, wt.%
50.5
same technique as in Figure 4 and
5. The only peaks that showed up 50.0
Propylene, wt.%
6.0
of RON is fairly small. (Figure 21)
MON is consistently lower with the 5.5
biofeeds, and this is due to the
lower aromatics content of the 5.0
Gasoline MON
cules are liberated from the glycerin
Gasoline RON
83
backbone via thermal cracking, 92.5
they follow a typical FCC rule of 92 82
thumb, which is that the longer the
chain (ie. the higher the carbon 91.5
81
number in the molecule), the more 91
broad the distribution of product 90.5
80
olefins that will result from catalytic
cracking of the molecule. Thus, the 90 79
palm oil produces the largest 71 73 75 77 79 81
amount of C4 olefins, as seen in
Figure 22. Base Feed Palm Soy Rapeseed
7.2
While the properties of the specific
hydrowax feed are not known, in 7.0
general these feeds are highly
6.8
paraffinic and the addition of veg- 71.0 72.0 73.0 74.0 75.0 76.0 77.0 78.0 79.0 80.0 81.0
etable oils to that type of feed would Conversion, wt.%
in fact degrade the overall proper-
12 www.e-catalysts.com
Base Feed 15% Palm 15% Soy 15% Rapeseed
Figure 23
Quantitative Yields for Different
Hydrowax/Rapeseed Oil Blends References
30 LCO
3. Wallenstein, D., Roberie, T., and
Bruhin, T., Catalysis Today 127, 2007, pp. 54-
20 69.
HCO
Our technologies ensure that fuels that are environmentally friendly are
manufactured in processes that are economically sound.
Make biofuels with Grace, fueling innovation for sustainable energy worldwide.
www.GraceBiofuels.com
Biofuels@Grace.com
GRACE® is a trademark, registered in the United States and/or other countries, of W. R. Grace & Co.-Conn.
©2008 W. R. Grace & Co.-Conn. All Rights Reserved.
GENESISTM Catalyst
Commercial Update
he marketplace has shown great lyst families which will deliver higher
Rosann K. Schiller
Product Manager,
Grace Davison Refining
Technologies
T enthusiasm for the GENESIS™
catalyst approach to catalyst for-
mulations1. The number of applications
activity and improved gasoline
selectivity at equivalent bottoms
and coke yields.
Doc Kirchgessner
is growing and we expect this trend to
continue in 2008 and beyond. GENESISTM Catalyst Provides
Operating Flexibilty to US
Technical Sales Manager, GENESIS™ catalyst performance has Refiner
Grace Davison Refining met or exceeded expectations in
Technologies applications versus competitive tech- A major US refiner processing
nologies as well as over existing hydrotreated feed was using a
and Grace Davison catalyst technology. Grace Davison LIBRA® catalyst.
Kelly Stafford
We will share some of the results with New operating economics dictated
you here. a change in catalyst formulation,
specifically to maximize LCO with a
Technical Sales Coordinator,
We are also pleased to introduce two further goal of maximum gasoline +
Grace Davison Refining
new high activity GENESIS™ compo- LCO yield (G+D). Since the refinery
Technologies
nents. We have made improvements was constrained by main air blower
to our MIDAS® and PINNACLE® cata- (MAB) and wet gas compressor
2.5
10
2.0
5
Delta Coke
1.5
Delta RgT, ˚F Delta RxT, ˚F Shift, % Rel
0
C/O Shift % Rel 1.0
-5
0.5
-20
-1.0
(WGC), the desired shifts should be hydrotreater operation, as favorable caused by circulation instability. In
achieved without increasing LPG, refinery margins shift between gaso- the last example, Refiner E realized
coke, or dry gas. The main opera- line, alkylate, and ULSD production. a dramatic 15% improvement in flow
tions strategy was to lower conver- characteristics with GENESIS™ cat-
sion without sacrificing bottoms GENESISTM Catalysts Improve alyst, and this improvement has
cracking. Laboratory testing pre- Fluidization Properties of been sustained for over one year.
dicted that GENESIS™ catalyst Circulating Inventory
would provide an incremental 0.8 How do GENESIS™ catalyst sys-
wt.% G+D yield, lower bottoms, and GENESIS™ catalysts systems have tems improve fluidization? The
reduced H2 yield at constant coke demonstrated superior yield perform- Umb/Umf is a function of unit oper-
and LPG yield. Based on the posi- ance in field application. Improved ations as well as catalyst properties
tive laboratory results, a GENESIS™ fluidization characteristics have also such as particle size distribution
formulation was selected for use in been observed where GENESIS™ and pore volume. The excellent
this application. catalyst systems are being used. The fines retention of Grace Davison
Umb/Umf is a fluidization factor used alumina-sol catalysts ensure that
GENESIS™ catalyst has exceeded to determine the fluidization capabili- the optimal mix of particles is
expectations at this refinery. In ties of an equilibrium catalyst2. The retained in the circulating inventory.
addition to the selectivity benefits value of a good Umb/Umf is unit MIDAS® catalysts have very high
observed in the laboratory testing, dependent and a value of 2.0 or pore volume1 that when combined
GENESIS™ catalyst has also pro- greater represents an optimal opera- with an optimal particle size distri-
vided a significant reduction in delta tion. Figure 26 shows trends in bution creates an easily fluidized
coke and hydrogen. The improve- Umb/Umf over time in four separate inventory, increasing the Umb/Umf.
ment in delta coke allowed the refin- FCC applications. The higher this ratio, the more for-
er to increase cat-to-oil ratio (C/O) giving the fluidized catalyst is to
at lower operating severity. The The first example is Refiner B, who changes in density, and the more
increase in C/O coupled with the had good fluidization characteristics easily it will tend to circulate in an
selectivity advantages resulted in in the inventory with a Grace catalyst FCC unit3. The unique properties of
an increase in gasoline + LCO of prior to starting GENESIS™ catalyst. the individual components in GENE-
more than 2 lv.% (Figure 25). The As the unit turned over to GENESIS™ SIS™ catalyst systems result in an
reduction in dry gas alleviated the catalyst, Umb/Umf improved further equilibrium catalyst inventory that
WGC constraint, providing room to facilitating smooth circulation. has preferred flow characteristics.
optimize reactor temperature with- Refiners C, D, and E were using com-
out compromising octane. Since petitive catalysts before GENESIS™. High Activity GENESISTM
the introduction of GENESIS™ cat- Refiners C and D were operating with Catalyst Components
alyst, the unit has realized signifi- a fluidization factor below the desired
cant operating flexibility. The optimum of 2.0. The flow characteris- The industry requires high activity
reduced operating severity has tics of GENESIS™ catalysts enabled and stability of FCC catalyst.
allowed the refiner the ability to opti- these units to achieve optimal Stability is needed to combat cata-
mize FCC operations with feed Umb/Umf levels and minimize upsets lyst deactivation at high metals
16 www.e-catalysts.com
Figure 26
GENESIS™ Catalyst Improves Equilibrium Umb/Umf Trends
2.30
2.00
Base Grace Davison Competitor 1
GENESISTM GENESISTM
2.25
1.95
UMB/UMF
UMB/UMF
2.20 1.90
2.15 1.85
Refiner C
2.10 1.80
Refiner B
Sep Nov Jan Mar
Nov Jan Mar May
2.1 2.4
Competitor 1 Competitor 2
GENESISTM GENESISTM
2.3
2.0
UMB/UMF
UMB/UMF
2.2
1.9
2.1
2.0
1.8
Refiner D Refiner E
Nov Dec Feb Apr Jul Jan Jul Jan
loadings when processing resid. lyst stability without a gas penalty4. Its version, without affecting particle
High activity and optimal fluidization open particle morphology makes it integrity, at equivalent coke and gas
characteristics can overcome a cir- easy for MIDAS® catalyst to selectively selectivity of MIDAS®-100. Catalyst
culation limit. Stable activity pro- crack the heaviest feed components properties are shown in Table VI.
vides the necessary delta coke for without compromising coke or gas
an ultra-clean hydrotreated feed. selectivity. The balance between As shown in Figure 27, MIDAS®-238
Grace Davison is pleased to intro- properly sized mesopores and opti- catalyst increases conversion by 2.5
duce two new catalyst technologies mized matrix acid strength delivers points over the base MIDAS®-138
that address these specific opera- the selectivity advantages over com- catalyst at constant C/O in Davison
tional challenges. The MIDAS®-200 petitive catalyst systems. Pore volume Circulation Riser (DCR) Pilot Plant
and PINNACLE® series of catalysts in the 100-600 Å range is critical for testing. The increased hydrogen
deliver increased activity relative to the destruction of heavy hydrocar- transfer activity of the zeolite
IMPACT® and MIDAS®-100 catalysts. bons and MIDAS® catalysts contain improves gasoline selectivity and
Both catalysts are excellent singular the highest amount of mesoporosity reduces wet gas yields without
catalyst solutions and exhibit syner- available in the industry today. compromising coke selectivity
gistic benefits when utilized in GEN- (Table VII).
ESIS™ catalyst systems. To answer the industry call for higher
activity, we have commercialized the Commercial testing demonstrates
New High Activity MIDAS® MIDAS®-200 series of catalysts. Each the activity advantage of the
Catalyst Technology is FCC catalyst technology is limited by MIDAS®-200 catalyst series of cata-
Commercialized the amount of active ingredients that lyst. Refiner F switched from GENE-
can be incorporated before attrition SIS™-1 catalyst, containing
MIDAS® catalysts are designed for resistance is compromised. The MIDAS®-138 catalyst, to GENE-
refiners who are interested in maxi- MIDAS®-200 family of catalysts SIS™-2 catalyst with an equivalent
mum bottoms upgrading and cata- achieves our goal of increased con- amount of MIDAS®-238 catalyst
Catalagram 103 Spring 2008 17
Table VI replacing the MIDAS®-138 catalyst
Properties of MIDAS®-200 Catalysts in the formulation. As shown in
Figure 28, GENESIS™-2 catalyst
increased equilibrium catalyst
Fresh Catalyst Properties MIDAS-238 MIDAS-138 (Ecat) MAT by 1# at equivalent cat-
alyst additions. Commercial data is
Al2O3 wt.% 49.8 50.8
Re2O3 wt.% 2.35 1.97
shown in Figure 29 and confirms the
gasoline selectivity advantages
Surface Area m2/g 280 290 seen in pilot testing.
ZSA m2/g 170 175
MSA m2/g 110 115 MIDAS®-200 series catalysts are
ABD g/cc 0.73 0.75 currently in use at four FCC units
DI 12 12
around the world. The ability to
adjust MIDAS® catalyst activity over
Deactivated Properties, with 3,000 ppm Ni and 3,000 ppm V
a broad range provides additional
Surface Area m2/g 155 160 flexibility to GENESIS™ catalyst
ZSA m2/g 90 85 users who may require higher MAT
MSA m2/g 65 75 but cannot sacrifice coke or bot-
Unit Cell Size Å 24.32 24.30 toms selectivity.
8.0
7.5
lent coke yield. The new formula-
tions are appropriate for refiners
7.0
interested in maximizing total fuels
6.5
production and reducing the load
6.0 MIDAS-238 on their wet gas compressor
5.5
MIDAS-138
(WGC).
5.0
66 68 70 72 74 76 78 Introduced in 2006, PINNACLE®
Conversion, wt.%
catalysts are alumina-sol catalysts
formulated with Grace Davison’s
nickel resistant matrix (TRM-400)
that provides superior coke and gas
Table VII selectivity in the presence of high
Resid Feed Constant Activity Comparison nickel feedstocks5. PINNACLE® cat-
alyst technology combines TRM-
MIDAS-238 MIDAS-138 400 with a moderate level of the
integral vanadium trap that is a sig-
nature of our IMPACT® catalyst tech-
Conversion, wt.% 74.5 72.0 nology. This combination of premi-
um metals-trapping technologies
Total Dry Gas, wt.% 2.8 3.1 minimizes deactivation and reduces
Total C3, wt.% 10.1 9.9 coke formation from contaminants,
Total C4, wt.% 13.6 13.2 resulting in enhanced zeolite stabili-
ty. And just like IMPACT® catalysts1,
Gasoline, wt.% 41.6 39.5 PINNACLE® catalyst’s unique inte-
LCO, wt.% 18.4 19.9 gral metals traps are effective at
low, moderate, and high metals
Bottoms, wt.% 7.1 8.1
Coke, wt.% 6.2 6.1
18 www.e-catalysts.com
loadings, demonstrating a coke and Figure 28
gas advantage over competitive GENESIS™ with MIDAS®-238 Catalyst Increased Ecat MAT
catalysts whether processing
hydrotreated, VGO or resid feed-
stock.
79
The proprietary zeolite modifica-
tions in PINNACLE® catalysts shift
Figure 31
Both GENESIS™ Catalyst Formulations
Have Equivalent Coke to Bottoms
10
GENESIS A
8
Bottoms, wt.%
GENESIS B
4
1.5 2 2.5 3 3.5 4
Coke (wt.% feed)
20 www.e-catalysts.com
Grace Davison:
Your proven FCCU emissions solution.
Grace Davison environmental technologies
have helped refiners reduce FCCU emissions
for over a quarter of a century.
The FCCU is often the largest point source within the refinery for SOx
and NOx emissions. The use of innovative catalytic technologies
reduces emissions from the FCCU regenerator, without the need for
capital intensive “end-of-pipe” hardware solutions.
SO
Ox x Extensive customer-driven research efforts at Grace are providing new insights into improving
Davison
Environmental
N
SOx and NOx removal. Want to find out more about our environmental technologies?
Technologies
CO
Contact us at www.e-catalysts.com or call us at (410) 531-8226.
Let Grace Davison help you meet your FCCU emission challenges.
Effect of
Hydrocarbon Partial Pressure
on Propylene Production
in the FCC
Ruizhong Hu
Introduction pressure. Studies documenting the
effect of hydrocarbon partial pres-
Senior Principal Scientist any refiners have continually sure on FCC yields are scarce.
Gordon Weatherbee
Principal Engineer
M revamped and debottle-
necked their FCC units to
increase feed throughput and improve
It is generally expected that an
increase in hydrocarbon partial
Hongbo Ma
profitability Most FCC units are run- pressure will increase the rate of all
ning at a significantly higher feed rate bimolecular reactions, including
than the original design. With higher hydrogen transfer, relative to crack-
Research Engineer throughput, in order to maintain cata- ing, which is unimolecular. An
Terry Roberie
lyst and vapor velocity in the riser and increase in the rate of hydrogen
cyclones, the unit pressure and con- transfer will result in a reduction of
sequently the hydrocarbon partial olefins in both gasoline and LPG
Director, FCC Evaluations
and an increase in gasoline range
Wu-Cheng Cheng
pressure need to be increased.
Current laboratory methods for evalu- aromatics and paraffins. The
ating FCC catalysts and additives change in the rate of hydrogen
Director, Grace Davison Refining cannot match hydrocarbon partial transfer could also affect gasoline
Technologies R&D pressures in commercial FCC units. sulfur concentration and the effec-
One reason is that available laborato- tiveness of gasoline sulfur reduction
Grace Davison Refining ry testing equipment, such as ACE catalysts and additives. Moreover,
Technologies and MAT typically operate at atmos- the effectiveness of ZSM-5 addi-
pheric pressure. The Davison tives, which are used to produce
Circulating Riser (DCR), a pilot plant- light olefins, especially propylene,
scale testing unit, is regularly operat- could be affected by hydrocarbon
ed under total pressure similar to com- partial pressure. Since ZSM-5
mercial FCC units1. However, due to works by cracking gasoline range
the small diameter of the DCR riser, a olefin molecules, changing the rate
relatively large amount of nitrogen is of hydrogen transfer could have a
needed to lift the catalyst, thus profound impact on propylene yield.
decreasing the hydrocarbon partial
22 www.e-catalysts.com
This paper will discuss the results of Schematic Diagram of Grace Davison DCR
a series of cracking experiments in
the DCR, where the hydrocarbon
partial pressure was varied by vary-
Meter
ing the total reactor pressure, the
feed rate and the amount of lift gas. Control
Condenser
Valve
The effect of changing hydrocarbon
Regenerator
partial pressure on hydrocarbon Stabilizer
Column
yields, especially that of light olefin,
Riser Reactor
and gasoline sulfur will be dis- r
Stripper
cussed. Feed
Tanks
Scale Scale
Experimental
Heat
Feed Preheater Exchanger
At the right is a schematic diagram Liquid Product
of the standard DCR setup. The Receivers
range of operating conditions in the Feed Pump
Dispersant Stripping
Steam Steam
DCR is shown in Table X. Operation
of the DCR has been described
previously1. Similar to commercial
FCC units, the DCR is operated in
adiabatic mode. In typical DCR tube heat exchanger. The rate of heat nitrogen lift gas and steam injection
operation, the regenerator tempera- transfer across this exchanger pro- rate constant while increasing the
ture, the riser outlet temperature vides a precise and reliable method to total pressure and feed rate. The
and the feed rate are set. The cata- calculate the catalyst circulation rate. latter case is similar to some com-
lyst circulation rate and thus, the The stabilizer column, also called the mercial FCC unit revamps where
catalyst to oil ratio, is changed by debutanizer column, is operated to the total pressure of a FCC unit is
varying the feed pre-heat tempera- separate C4 minus from the liquid increased to accommodate higher
ture. During operation of the DCR, product, which is condensed and col- feed and catalyst circulation rate.
a metering pump precisely controls lected. The collected liquid is ana-
the feed rate as feed is pumped lyzed by GC (SIMDIS – simulated dis- Table XI shows the three DCR oper-
from the load cell through a pre- tillation) to provide gasoline (ibp - ation conditions. Condition 3 is a
heater. Nitrogen and steam, inject- 430°F), LCO (430-700°F), and 700°F + commonly used DCR operating
ed through a separate pre- bottoms fractions. The gaseous prod- condition, while Conditions 1 and 2
heater/vaporizer, are used as a feed ucts are metered and batch collected are modifications to raise the hydro-
dispersant. Catalyst and product for subsequent analysis by GC. carbon partial pressure closer to
pass from the riser to the stripper the value in commercial FCC opera-
overhead disengager. Products exit We investigated two methods of tions. Since cracking is a molecular
the disengager through a refrigerat- changing hydrocarbon partial pres- weight reduction process, the
ed stabilizer column to a control sure. The first method involved keep- hydrocarbon mole fraction and,
valve which maintains unit pressure ing the total pressure, feed rate, and therefore, partial pressure increase
at the desired level. A section of the steam injection rate constant while along the riser. The molar expan-
stripper-regenerator spent catalyst reducing the nitrogen lift gas. The sec- sion (moles of product/moles of
transfer line consists of a shell and ond method involved keeping the feed) in a typical FCC unit is
between four and five. For the pur-
Table X pose of engineering calculations, it
is common to approximate the
DCR Operating Ranges hydrocarbon mole fraction as equal
Control Parameter Range to 1/3 of the mole fraction at the inlet
and 2/3 of the mole fraction at the
System Pressure < 45 psig
outlet of the riser. The total moles of
Catalyst Charge 1500-4000 g
the hydrocarbon products are cal-
Catalyst Circulation Rate 2500-15000 g/h
culated by using GC analyses of
Feed Rate 350-2000 g/h
the light gases and gasoline PIONA
Feed Types GO, VGO, Resid
and assuming average molecular
Feed Preheater Temperature 120-400°C (250-750°F) weight values of 220 and 350 for
Riser Temperature <590°C (<1100°F) LCO and bottoms, respectively.
Disengager Temperature <746°C (800-1100°F)
Stripper Temperature 427-593°C (800-1100°F)
Stabilizer Column Temperature -34°C (-30°F)
In varying hydrocarbon partial pres- sure of Condition 3 was 2.3 times to the operating conditions of ACE,
sure, we chose operating conditions lower. However, its WHSV was also the operating conditions of the DCR
so as not to greatly change the somewhat greater, due to the higher are much closer to those of the
weight hourly space velocity level of lift nitrogen used. In this case, commercial unit.
(WHSV), as that in itself could we would need to rationalize the con-
change the selectivity and compli- tribution of hydrocarbon partial pres- Two Davison commercial FCC cata-
cate the interpretation of the results. sure to the selectivity shifts. lysts, labeled Catalyst A and
The slip factor in the riser (ratio of Catalyst B, containing 1.2 and 3.1%
the gas velocity to catalyst velocity), Table XI also compares the current RE2O3, respectively, were used in
estimated by the correlation of DCR operating conditions with that of this study. Both catalysts were
Pugsley and Berruti2, varied from commercial FCC units and ACE. steam deactivated according to
1.7 to 2. These values were consis- Compared to the earlier operating CPS-3 protocol4 at 1480˚F with 500
tent with those reported by Bollas et conditions (Condition 3), Conditions 1 ppm nickel and 500 ppm vanadium.
al.3. Once the slip factor was deter- and 2 are closer to the commercial The chemical and physical proper-
mined, the catalyst holdup (the units, especially in hydrocarbon par- ties of the two catalysts are listed on
amount of catalyst in the riser), cat- tial pressure. Furthermore, compared Table XII. The deactivated unit cell
alyst contact time and WHSV were
readily calculated (Table XI). The
Table XII
catalyst holdup values followed the
trend of pressure drop measure- Properties of Catalysts Deactivated at
ments across the riser. Conditions 1 500ppm Ni/500ppm V CPS-3/1480˚F
and 2 varied in hydrocarbon partial
pressure by a factor of 1.55. Analysis Catalyst A Catalyst B
However, the values of the WHSV, Al2 O 3 , wt.% 40.8 46.9
catalyst-to-oil ratio and conversion RE 2 O 3 , wt.% 1.16 3.05
were essentially identical. Na 2 O, wt.% 0.39 0.28
Therefore, the changes in selectivity Ni, ppm 537 523
could be attributed principally to the V, ppm 520 510
change in hydrocarbon partial pres- Surface Area, m2 /g 219 146
sure. Compared to Conditions 1 ZSA, m2 /g 179 113
and 2, the hydrocarbon partial pres- MSA, m2 /g 40 33
Unit Cell Size, Å 24.24 24.32
24 www.e-catalysts.com
Table XIII
Properties of VGO Feedstock
size measurements of the low and feed rate and catalyst circulation the propane/propylene, n-butane/(1-
high RE2O3 catalysts are 24.24Å total pressure of the unit has to be butylene + trans-2-butylene + cis-2-
and 24.32Å, respectively. Catalyst increased to maintain velocity. butylene) and isobutane/isobuty-
A was also blended with 20% lene. These ratios are shown in
OlefinsUltra® additive, a commer- The plots of catalyst to oil ratio, total Table XIV. In this analysis we are
cially available ZSM-5 additive and C3, total C4, gasoline, LCO, and coke assuming that the C3 and C4 alka-
deactivated according to CPS-3 yields against conversion are shown in nes are the product of hydrogen
protocol at 1480˚F with 500 ppm Figure 32. Increasing HC partial pres- transfer from their parent alkenes
nickel and 500 ppm vanadium. A sure increases dry gas and coke at and ignoring the alkanes formed by
Gulf Coast vacuum gas oil feed the expense of gasoline. The yields of thermal or protolytic cracking. The
was used in this study. The prop- total C3, C4 and LCO remain about the hydrogen transfer reaction of
erties of the feedstock are shown same. The higher coke yield may be isobutene proceeds via a tertiary
on Table XIII. attributed to a higher rate of oligomer- carbenium ion intermediate and
ization, which is a bimolecular reac- thus occurs at a much faster rate
Results and Discussion tion and favored at high pressure. The than the hydrogen transfer reactions
higher dry gas could be the result of of propylene and linear butenes,
Case I oligomerization/recracking. which proceed through a less sta-
ble secondary carbenium ion inter-
In this example, Catalyst A was test- Figure 33 shows that increasing the mediate. All of the hydrogen trans-
ed in the DCR under both HC partial pressure decreases the fer indices increase by a factor of
Conditions 1 and 2. Under yields of propylene, butenes, and 1.5, as the HC partial pressure
Condition 1, the unit pressure was gasoline olefins, while increasing the increases almost proportionally by a
40 psig, the feed rate was 1500 g/h, yield of gasoline isoparaffins. factor of 1.6 from 28 to 44 psia.
the dispersing steam was 30 g/h, Increasing HC partial pressure sub- Thus, all the yield shifts are consis-
and 25 l/h nitrogen was injected to stantially lowers the C3 and C4 tent with an increase in the rate of
help disperse the feed as well as to hydrogen transfer with the increase
olefinicities. These yield shifts sug-
lift the catalyst. Based on the above in HC partial pressure.
gest that the rate of hydrogen transfer
discussion, the time-averaged (1/3
increases with HC partial pressure, as
inlet + 2/3 outlet) hydrocarbon par- Case II
would be expected for a bimolecular
tial pressure under this condition
reaction.
was 44 psia. Under Condition 2, the Catalyst A was blended with 20%
unit pressure was 25 psig, the feed OlefinsUltra® additive and tested in
The interpolated yields at 73 wt.%
rate was 1000 g/h, whereas the the DCR under Conditions 1 and 2.
conversion are listed on Table XIV. A
steam and nitrogen flow rates were The catalyst to oil ratio, total C3, total
convenient way to gauge the hydro-
the same as that of Condition 1.
gen transfer rate is to look at the paraf- C4, gasoline, LCO, and coke yields
The hydrocarbon partial pressure
fins to olefins ratio of C3, as well as lin- against conversion plots are shown
under this condition is 28 psia. The
above comparison is very similar to ear and branched C4 compounds5. in Figure 34. The yields of C3=,
a common revamp of a commercial The hydrogen transfer indices are C4=, gasoline olefins and gasoline
FCC unit where in order to increase defined as the ratios of isoparaffins, as well as the olefinici-
7 11
6.0
6
10
5.5
5
9
5.0
Gasoline, wt.% LCO, wt.% Coke, wt.%
4.5
52
23
4.0
51 22
3.5
21
50
20 3.0
49
19 2.5
48
70.0 72.5 75.0 70.0 72.5 75.0
Conversion, wt.%
Figure 33
Effect of DCR Operating Conditions on the Olefins Yield and Olefinicity of Catalyst A
6.8 30.0
4.5
6.4 27.5
0.72 26
0.84
0.68 24
0.82 22
0.64
0.80 0.60 20
26 www.e-catalysts.com
Table XIV
Interpolated Yields at 73 wt.% Conversion Over Catalyst A
Condition 1 Condition 2
Ratio of HC Pressure
HC Partial Pressure, psia 44 28 1.6
Cat to Oil 6.1 6.5
H2 Yield , wt.% 0.05 0.05
C 1 + C 2 's , wt.% 2.8 2.4
C 2 = , wt.% 0.8 0.8
Total C 3 , wt.% 5.9 5.9
C 3 = , wt.% 4.7 5.0
Total C 4 , wt.% 10.5 10.6
iC 4 , wt.% 3.0 2.5
nC 4 , wt.% 0.8 0.6
Total C 4 = , wt.% 6.7 7.5
iC 4 = , wt.% 1.9 2.4
Gasoline, wt.% 50.1 50.8
G-Con P, wt.% 4.0 3.7
G-Con I, wt.% 24.8 22.2
G-Con A, wt.% 30.5 30.2
G-Con N, wt.% 10.9 11.5
G-Con O, wt.% 29.6 32.2
G-Con RON EST 92.2 92.0
G-Con MON EST 79.8 79.2
LCO, wt.% 20.5 20.7
Bottoms, wt.% 6.3 6.2
Coke, wt.% 3.5 3.0
Hydrogen Transfer Index Ratio of HT Index
C 3 /C 3 = 0.24 0.17 1.4
nC 4 /(1C 4 = + t 2 C4 = +c 2 C 4 =) 0.18 0.12 1.5
iC 4 /iC 4 = 1.56 1.05 1.5
Figure 34
Effect of DCR Operating Conditions on the Yields of
Catalyst A with 20% OlefinsUltra® Additive
8 12.5
16.5
12.0 16.0
7
15.5
11.5
6 15.0
11.0 14.5
Gasoline, wt.% LCO, wt.% Coke, wt.%
41 22
4.0
21
40 3.5
20
3.0
39 19
2.5
18
38 2.0
70.0 72.5 75.0 70.0 72.5 75.0
Conversion, wt.%
11.0 37.5
11.4
35.0
10.5 11.1
32.5
10.0 10.8
30.0
9.5 10.5
C3 = / Total C3 C4 = / Total C4 G-Con I, wt.%
0.80
0.900
21.0
0.885 0.75
19.5
0.870
0.70 18.0
0.855
16.5
0.840 0.65
15.0
70.0 72.5 75.0 70.0 72.5 75.0
Conversion, wt.%
ties of C3, C4, and gasoline are ZSM-5. The rate of hydrogen transfer, position are shown in Figure 37. For
shown in Figure 35. As in the case as estimated by the hydrogen transfer both the high and low unit cell size
without OlefinsUltra® additive (Case indices, described above, increases catalysts, the response of the LPG
I), increasing HC partial pressure approximately proportionally to HC and gasoline olefin yields to the
increases coke and dry gas and partial pressure. changes in DCR conditions are very
dramatically decreases gasoline similar to that observed in Case 1,
and LPG olefinicity. The C3 olefinic- Case III namely increasing HC partial pres-
ity of ca. 0.84 at the higher HC par- sure decreases LPG and gasoline
tial pressure is much more realistic In this example, Catalysts A and B, olefins and olefinicity. The hydrogen
and close to the commercially having unit cells size values of 24.24Å transfer indices increased by a fac-
observed values. and 24.31Å, respectively, were tested tor of two as HC partial pressure
under DCR Conditions 1 and 3. increased by a factor of 2.3 (Table
The interpolated yields at constant Condition 3 featured a unit pressure of XVI). Thus, as in Case I, the change
conversion of 73 wt.% are shown on 25 psig, 1000 g/h feed rate, 30 g/hour in the hydrogen transfer indices are
Table XV. Remarkably, increasing steam, and 128 l/h nitrogen. The main nearly proportional to the change in
the HC partial pressure from 28 to difference between Condition 1 and HC partial pressure. This suggests
44 psia decreases the propylene Condition 3 was the greater amount of that the change in HC partial pres-
yield by 1 wt.% absolute and nitrogen lift gas used in Condition 3, sure is mainly responsible for the
decreases the butylenes yield by which not only decreased the HC par- yield changes while the shifts in
0.6 wt.% absolute. It is known that tial pressure by a factor of 2.3, from WHSV may be responsible for the
the addition of ZSM-5 increases 44 to 19 psia, but also increased the shifts in conversion at a given cata-
LPG olefins by cracking gasoline WHSV by a factor of 1.4. The effect of lyst to oil ratio.
range olefins6-8. Increasing the rate the change in WHSV will be dis-
of hydrogen transfer, by increasing cussed. The yields shifts due to changing
the HC partial pressure, depletes unit cell size are consistent with
the gasoline range olefins and The main yields are shown in Figure what has been reported in the liter-
decreases the effectiveness of 36, while the LPG and gasoline com- ature9,10, namely that the higher UCS
28 www.e-catalysts.com
Table XV
Interpolated Yields at 73 wt.% Conversion Over Catalyst A with 20% OlefinsUltra® Additive
Condition 1 Condition 2
Ratio of HC Pressure
HC Partial Pressure, psia 44 28 1.6
Cat to Oil 6.7 6.7
H2 Yield , wt.% 0.04 0.05
C 1 + C 2 's , wt.% 3.0 2.6
C 2 = , wt.% 1.4 1.3
Total C 3 , wt.% 11.4 12.0
C 3 =, wt.% 9.7 10.7
Total C 4 , wt.% 15.8 15.5
iC 4 , wt.% 3.9 3.3
nC 4 , wt.% 1.0 0.8
Total C 4 = , wt.% 10.9 11.4
iC 4 = , wt.% 4.1 4.3
Gasoline , wt.% 39.2 39.8
G-Con P , wt.% 4.0 3.8
G-Con I , wt.% 19.6 17.1
G-Con A , wt.% 33.7 33.9
G-Con N , wt.% 8.9 9.2
G-Con O , wt.% 33.5 35.8
G-Con RON EST 95.2 94.9
G-Con MON EST 81.5 80.9
LCO , wt.% 19.6 19.8
Bottoms , wt.% 7.3 7.1
Coke , wt.% 3.3 2.9
Hydrogen Transfer Index Ratio of HT Index
C 3 /C 3 = 0.17 0.12 1.5
nC 4 /(1C 4 = + t 2 C4 = +c 2 C 4 =) 0.15 0.11 1.4
iC 4 /iC 4 = 0.95 0.76 1.2
Table XVI
Interpolated Yields at 75 wt.% Conversion Over Catalyst A and B
Catalyst A Catalyst B
Condition 3 Condition 1 Condition 3 Condition 1
64 72 80
C/O Ratio Total C3 wt.% Total C4 wt.%
6.5 12
7
6.0 11
6
5.5 10
5 5.0 9
4.5 8
Gasoline wt.% LCO wt.% Coke wt.%
54.0 25.0
6
52.5 22.5
5
51.0
20.0 4
49.5
17.5 3
48.0
15.0 2
64 72 80 64 72 80
Conversion, wt.%
Figure 37
Variaton of Olefins Yield and Olefinicity with Unit Cell Size and DCR Operating Conditions
64 72 80
C3 = wt.% Total C4 = wt.% G-Con O wt.%
5.6 35
7.5
5.2 30
7.0
4.8 25
6.5
4.4 6.0 20
4.0 5.5 15
C3 = / Total C3 C4 = / Total C4 G-Con I wt.%
0.8 36
0.87
0.84 0.7 32
0.81 0.6 28
0.78 24
0.5
0.75
20
0.4
64 72 80 64 72 80
Conversion, wt.%
30 www.e-catalysts.com
Scheme 1 4.D. Wallenstein, R.H. Harding, J.R.D. Nee,
L.T. Boock, “Recent Advances in the
Deactivation of FCC Catalysts by Cyclic
R R
Propylene Steaming (CPS) in the Presence
R -C4 + H2S and Absence of Metals,” Appl.Catal. A:
General 204 (2000) 89.
S S cracking
HT
5.Cheng, W-C., Suarez, W., and Young, G. W;
catalyst makes higher gasoline, nature, increase with increasing HC “The effect of catalyst properties on the
lower octane, lower LPG and gaso- partial pressure. The hydrogen trans- selectivities of isobutene and isoamylene in
FCC,” AIChE Symposium Series, 291
line olefins. These trends are fer index, defined as the paraffin/olefin (1992) 38.
observed at both DCR conditions. ratio of C3, linear C4 and branched C4
The effect of unit cell size and HC species increase almost linearly with 6.K. Rajagopalan, G.W. Young, in Fluid
partial pressure on the rate of Catalytic Cracking − Role in Modern
HC partial pressure. It has been
Refining, M.L. Occelli (Ed.), ACS Symposium
hydrogen transfer appears to be demonstrated that the effectiveness of Series 375 (1988) 34.
simply additive. The rate of bimole- ZSM-5 additives is lessened at high
cular reactions can be increased by HC partial pressure due to the deple- 7.X. Zhao, T.G. Roberie,”ZSM-5 Additive in
increasing acid site density as well Fluid Catalytic Cracking. 1. Effect of Additive
tion of gasoline range olefins via
Level and Temperature on Light Olefins and
as increasing HC partial pressure. hydrogen transfer reactions. The con- Gasoline Olefins,” Ind. Eng. Chem. Res. 38
The ratios of the hydrogen transfer centration of gasoline sulfur species (1999) 3847.
indices of Catalyst B to Catalyst A decreases at higher HC pressure,
are about the same at both low and 8.R.J. Madon, “Role of ZSM-5 and
again due to higher rate of hydrogen
Ultrastable Y Zeolites for Increasing Gasoline
high HC partial pressure (Table transfer. Recent advancements in Octane Number,” J. Catal. 129 (1991) 275.
XVI). DCR operation enable more realistic
simulation of commercial FCCU oper- 9.L.A. Pine, P.J. Maher, W.A. Wachter,
Figure 38 shows the concentration “Prediction of Cracking Catalyst Behavior by
ation.
a Zeolite Unit Cell Size Model,” J. Catal. 85
of gasoline sulfur (including all thio- (1984) 466.
phene species with a boiling point References
below 430°F, tetrahydrothiophene, 10.G.W. Young, W. Suarez, T.G. Roberie, W.C.
and benzothiophene) for both 1.G.W. Young, G.D. Weatherbee, “FCCU Studies Cheng, “Reformulated Gasoline: The Role of
with an Adiabatic Circulating Pilot Unit,” AIChE Current and Future Catalysts,” NPRA Annual
Catalysts A and B under the two Meeting, AM-91-34, 1991.
Annual Meeting, November, 1989.
DCR operation conditions.
Gasoline sulfur concentration 2.S. T. Pugley, F. A. Berruti, “A Predictive 11.R.H. Harding, R. Gatte, J.A. Whitecavage,
decreases with increasing unit cell Hydrodynamic Model for Circulating Fluidized R.F. Wormsbecher, “Reaction Kinetics of
Bed Risers,” Powder Technol., 89 (1996) 57. Gasoline Sulfur Compounds,” in
size and with increasing HC partial Environmental Catalysis, J.N. Armor (Ed.),
pressure. These results suggest 3.G. M. Bollas, I. A. Vasalos, A. A. Lappas, D. American Chemical Society, Symposium
that the reduction of gasoline sulfur Iatridis, “Modeling Small-Diameter FCC Riser Series 552 (1994) 286.
follows the trend of increase hydro- Reactors, A Hydrodynamic and Kinetic
Approach,” Ind. Eng. Chem. Res., 41 (2002) 12.F. Can, A. Travert , V. Ruaux , J.-P. Gilson ,
gen transfer activity and are consis- F. Maugé , R. Hu, R.F. Wormsbecher, “FCC
5410.
tent with the previously proposed Gasoline Sulfur Reduction Additives:
mechanism, shown below11,12. Mechanism and Active Sites,” J. Catal. 249
(2007) 79.
Scheme 1
Figure 38
High rate of hydrogen transfer Effect of UCS and DCR Operating Conditions
speeds up this reaction by promot-
on Gasoline Sulfur Concentration
ing the formation of the reaction
intermediate, tetrahydrothiophene.
230
220
Conclusions
210
Gasoline Sulfur (ppm)
Brian Watkins n keeping up with refiners’ including Co and Ni. It has been
Supervisor, Laboratory Technology I demand for superior technology shown that when applied correctly,
chelates can promote the formation
Charles Olsen
and premium performance,
Advanced Refining Technologies of Type II metal sulfide sites (see
introduced its line of ultra high activity Catalagram® 96, 2004). ART’s pre-
Worldwide Technical Services DX series of catalysts. ART’s series of mium CDXi has proven its perform-
Manager DXTM catalysts has exceeded refinery ance advantage with greater stabili-
David Krenzke
expectations in their ability to tolerate ty and exceptional ability to utilize a
difficult feed blends in demanding minimum amount of hydrogen to
ULSD applications. Key to that effort provide refiners with consistent
Technical Services Manager is maximizing the utilization of the ULSD production.
active metals on the catalyst through
Advanced Refining Technologies ART’s chelate chemistry. This impreg- In keeping with this tradition, ART is
nation technology offers outstanding releasing its newest generation of
potential for significantly improving ultra high activity CoMo DXTM cata-
metals utilization in catalysis due to a lyst, 420DXTM. Figure 39 compares
superior ability to control metal ions the activity of a variety of CoMo cat-
125
fur molecules, yet avoid the unnec-
essary additional saturation of the
100 monoaromatic compounds which
can increase hydrogen consump-
tion.
75
AT405 CDXi 420DXTM
ART’s dedicated staff of
HDS HDN researchers has continued to inves-
tigate ways to improve catalytic per-
formance, and surface acidity has
been identified as an important
Figure 40 property. It is generally accepted
IR Spectra of the Support Material for ART 420DXTM Catalyst that higher surface acidity increas-
es reactions controlled through ring
26 saturation such as nitrogen and hin-
1451.67
24
LC243-260-1
dered sulfur removal. This acidity
22 has also been shown to affect the
20 interaction of active metals with the
18 alumina surface. ART was able to
16
1624.43 exploit this in the design of ART
14
1616.20 420DXTM catalyst. This catalyst uti-
12
1494.12 lizes similar impregnation technolo-
10 gy as CDXi, but is built on a modi-
8 fied alumina carrier which results in
6 a dramatic increase in activity. The
1700 1650 1600 1550 1500 1450 IR chart in Figure 40 shows a dou-
Wave numbers (cm -1) ble peak at 1624 and 1616 wave
numbers as well as one at 1451
which are believed to indicate the
presence of Lewis acid sites. This
Figure 41 feature was not as prevalent in the
Comparison of CDXi and ART 420DXTM Catalyst spectra of the CDXi support and
at High Pressure confirms the incorporation of sur-
face acidity in the new support.
(980 hydrogen partial pressure, 2500 H2/Oil) While the acid sites give ART
420DXTM catalyst superior perform-
650 ance for both HDS and HDN activi-
ty, they are not strong enough to ini-
640 tiate any cracking reactions under
WABT, ˚F (10ppm sulfur
and 1ppm nitrogen)
680
660
650
HDS HDN
CDXi 420DXTM
Figure 43
Comparison of CDXi and ART 420DXTM Catalyst
at ULSD Conditions
300 29
180 25
120 23
60 21
0 19
600 620 640 660 680
Temperature, ˚F
420DXTM CDXi
ditions ART 420DXTM catalyst clearly tial pressure and 1500 Scfb H2/Oil This enhanced sulfur removal activ-
outperforms CDXi by over 20°F at using a feedstock containing cracked ity offers refiners greater flexibility
10 ppm sulfur on a difficult feed material. in meeting their HDS activity
containing 30% cracked stocks. requirements while minimizing
The primary benefit of ART 420DXTM hydrogen consumption using ART
The performance gains seen at high catalyst is that the improved HDS and 420DXTM catalyst as a stand alone
pressures are also available to refin- HDN activity does not result in an catalyst or in combination with
ers operating at lower unit pressure increase in aromatic saturation and ART’s premium NDXi catalyst in a
and hydrogen circulation. Figure 42 consequently does not increase SmART System® for producing
shows the benefits of using ART hydrogen consumption. As can be ULSD from difficult feeds.
420DXTM catalyst at 10 ppm sulfur seen in Figure 43, ART 420DXTM cata-
and lower pressure. A clear 15°F lyst shows equal aromatic conversion
advantage over CDXi is apparent to that of CDXi at lower product sulfur.
even at only 580 Scfb hydrogen par-
Catalagram 103 Spring 2008 35
Successful Implementation of
State-Of-The-Art ULSD/Dewaxing
Technology at Irving Oil,
Saint John, NB
Mike Beshara
Project Manager, Irving Oil
Greg Rosinski ew regulations for Ultra-Low to maximize ULSD yield since low
Technical Services Engineer,
Advanced Refining Technologies N Sulfur Diesel (ULSD) in
Canada and the United States
cloud point is not required at that
time of year. Irving Oil also wanted
Charles Olsen
took effect in June 2006, reducing the to minimize the hydrogen consump-
on-road diesel sulfur content from 500 tion so that feed rate could be max-
Worldwide Technical Services to 15 ppmw. Anticipating the new sul- imized within make-up hydrogen
Manager, Advanced Refining fur regulation, Irving Oil decided to constraints.
Technologies convert the existing VGO
Ben Prins
Hydrocracker/LCO Desulfurizer at the In the summer of 2004, Irving Oil
Saint John refinery in New Brunswick, contacted several catalyst suppliers
Canada to an LCO /heavy diesel including ART and Süd-Chemie Inc.
Senior Process Engineer, ULSD unit. to begin the catalyst selection
Fluor process for the revamped unit. ART
Garry Jacobs
Technical Requirements is a supplier of top-tier hydrotreat-
ing catalysts, but does not have a
Technical Director, Fluor Fluor Corporation provided engineer- dewaxing catalyst in its portfolio.
Alan Birch
ing, procurement and supported tech- Süd-Chemie offers premium
nology selection. The project included dewaxing catalyst technology but
increasing feed capacity from 30,000 does not have hydrotreating prod-
Account Manager,
to 45,000 BPSD while at the same time ucts. ART and Süd-Chemie joined
Süd Chemie
producing ULSD from a feedstock together to offer a complete pack-
Ernst Köhler
containing up to 50% LCO. Irving Oil age.
wanted to make 7 ppm sulfur diesel
Global Product and required at least 30°F improve- The high level of LCO in the feed,
Manager-Zeolites, ment in cloud point for winter diesel. combined with high unit operating
Süd Chemie During the summer months they want- pressure and the need to minimize
ed to “turn off” the dewaxing function hydrogen consumption made it a
36 www.e-catalysts.com
challenge to design the appropriate Table XVII
catalyst system to meet the desired Feedstock Properties
product characteristics. This was
further complicated by the high
level of nitrogen in the feed. Type 50% LCO
API 22.8
Hydrodesulfurization and saturation Sulfur, wt.% 1.16
of olefins and aromatics are very Nitrogen, wppm 409
exothermic reactions and bed activ- Aromatics, vol.%
ity must be controlled to avoid Mono- 17.8
excessive temperature rise. The unit Di- 21.4
is equipped with inter-bed quench Poly 11.1
facilities to control the overall tem-
SimDist (D2887)
perature levels. Hydrodewaxing
(HDW) is endothermic and the IBP, °F 305
dewaxing activity is controlled 50% 588
through the bed inlet temperature, FBP 798
using lower temperature to “turn off”
the dewaxing catalyst activity.
a very high temperature rise. These mine the expected hydrogen con-
constraints dictated that the HDW cat- sumption and activity, ART pilot test-
It is easiest to control the tempera-
alyst should be placed below at least ed three catalyst systems: 100%
ture to the first bed which makes it a
one of the hydrotreating catalyst beds. NiMo, 50%/50% NiMo/CoMo
convenient position to place the
SmART System® and 100% CoMo
HDW catalyst. However, there are
The key question became how to using Irving’s feed which is listed in
several problems with installing the
achieve the desired product specifi- Table XVII.
HDW catalyst in that location. Most
cations while minimizing hydrogen
dewaxing catalysts are sensitive to
consumption. NiMo catalysts tend to ART selected CDXi, a premium high
nitrogen compounds so they must
have higher activity for saturating aro- activity CoMo catalyst for ULSD and
be protected by a NiMo catalyst.
matics and removing nitrogen while AT505 which is a high activity con-
LCO contains a high concentration
CoMo catalysts tend to give lower ventional NiMo catalyst and prede-
of olefins, and that can deactivate
hydrogen consumption through less cessor to NDXi.
the HDW catalyst quickly due to
aromatics saturation.
olefin polymerization and related
Figure 44 compares the HDS activi-
coking. The required operating
Hydrotreating Catalyst System ty observed for each catalyst sys-
temperature window for the HDW
tem. Under these conditions the all
catalyst is not compatible with man-
ART’s SmART Catalyst System® Series NiMo system is clearly the most
agement of the overall temperature
offers custom system design to meet active for sulfur removal, followed by
profile, as the feed olefins and
individual refiner constraints and the SmART System® and finally the
‘easy’ sulfur react rapidly, producing
objectives. To help Irving Oil deter- CoMo catalyst, CDXi.
9.0 Feed Cetane Index is 35.5 Dewaxing Catalyst and its Impact
on the Process Design
8.0
7.0
With the CoMo/NiMo ratio resolved,
the design proceeded to the
6.0 amount and placement of the
HYDEX®-G HDW catalyst. This
5.0 required a detailed evaluation of the
620 630 640 650 660 670 overall system with respect to heat
Temperature, °F release, quench capabilities and
catalyst requirements.
38 www.e-catalysts.com
Figure 48 ferences in deactivation
ULSD/Dewaxing Process Flow and rates).
• Selected combinations of the
Catalyst Loading Scheme Example above.
Recycle Hydrogen
Amine Unit
The results of this analysis revealed
H2S + NH3 + H2 that Bed 3 was the optimal location
Hydrogen for the HDW catalyst.
+25
The position of the HYDEX®-G in the
catalyst load was dictated by prac- +20
for Dewax WABT
+15
analysis. Practical constraints
+10
included:
+5
• The catalyst bed volumes in 0 Target Range for Cloud Point
Design Range
-40 30
The quench capability in the com-
mercial reactors is used to control -60 20
the bed temperatures to avoid over- SmART System®
SmART System®
treating and achieve the ultimate -80 + HYDEX®-G 10
Sulfur ppmw
goal of 7 ppm sulfur with control of
-100 0
the degree of dewaxing. The pilot Operating Temperature, ˚F
plant data indicates that increasing
the temperature by 40°F above the
threshold temperature increases the
40 www.e-catalysts.com
CPI with HYDEX®-G to 90°F, much Figure 52
more than necessary. At this high Commercial Unit Normalized Reactor Temperature
level of CPI the yields of naphtha
and light ends also increase which
can have an adverse effect on the 700
economics of the operation. 690
However, this reserve of activity 680
ensures the system is in balance
670
with regards to stability of the HDS
Temperature, °F
and HDW functions guaranteeing 660
Please welcome...
Shawn Abrams has joined Grace Davison as Vice President and General
Manager of Refining Technologies. Shawn will have direct responsibility for
the Refining Technologies Global Organization. Shawn joins us with over 20
years of management experience in global industrial markets at Evonik
Industries AG (formerly Degussa); where he held a variety of progressive
leadership positions including Director of Sales in Frankfurt, Germany, and
Vice President and General Manager of Asia Pacific for their Bleaching and
Water Chemicals Business Unit. For the past five years, Shawn has served
as a member of the Degussa Executive Group as Senior Vice President and General
Manager of their Active Oxygens Business Unit.
Shawn graduated from Lehigh University with a B.S. in Mechanical Engineering, and com-
pleted his MBA at the Thunderbird School of Global Management. Reporting to Shawn will
be
42 www.e-catalysts.com
Grace Davison
Multi-Loader System
Adam Kasle
It has long been understood that opti- mental regulations.
mum catalyst performance and unit
stability in an FCC unit are greatly Additive System Justification
Technical Sales Manager assisted by continuous and steady
Grace Davison Refining injection of FCC catalyst and additives The continuous goal of every FCC
Technologies throughout the course of the day. operator is to optimize unit perform-
However, this goal has been difficult ance and maximize profitability. As
for many refiners to achieve due to physical unit constraints are pushed
inadequate or unreliable addition sys- with mechanical solutions and cata-
tems that have not been up to the task. lyst formulations are tuned, refiners
The challenge has been further com- are looking more at unit control to
plicated in recent years as refiners derive further value from their
must now operate within multiple envi- assets. One such area of control is
ronmental regulations and depend on the maintenance of a continuously
stable operation and product addition optimized catalyst condition in the
rates to do so. One solution that has unit through stable, ongoing addi-
been rapidly gaining acceptance in tion of fresh catalyst. This allows the
the industry is the Grace Davison unit to provide the optimum yield
Multi-Loading System. This four-in- slate throughout the day without the
one catalyst addition system provides degradation associated with bulk
an accurate and reliable method for catalyst loading. Proper cracking
adding both cracking catalyst and catalyst management has begun to
additives to the FCC unit allowing the
44 www.e-catalysts.com
Commercial Experience The decision was made to install a Multi-Loader allowed the refiner to
Fresh
Additive Cat
Tote-Bin Hopper
55 gal.
Drum
46 www.e-catalysts.com
SAVETHEDATE