Catalagram 110 - Grace
Catalagram 110 - Grace
Catalagram 110 - Grace
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
No. <strong>110</strong> / Fall 2011 / www.grace.com<br />
<strong>Catalagram</strong><br />
®<br />
A Refining Technologies Publication<br />
Hydroprocessing Catalysts from<br />
The Chevron &<strong>Grace</strong>Joint Venture
Happy 10th Anniversary, ART<br />
The ART leadership team (left to right): Scott Purnell, Ryan Heaps, Bruno Tombolesi,<br />
Chris Dillon, Woody Shiflett, John Creighton, Kristen Kopp, Chuck Olsen, Lauren Blanchard,<br />
Darryl Klein, Cecilia Radlowski, Nathan Carpenter, Babu Patrose, Mark Peterson,<br />
Kaidong Chen, Charles Wear, Balbir Lakhanpal and Maureen Birnbaum.<br />
This past spring, we celebrated the 10th anniversary of the formation of Advanced Refining Technologies LLC,<br />
the hydroprocessing joint venture between <strong>Grace</strong> and Chevron Products Company. The venture was created<br />
in 2001 to develop, market and sell a comprehensive line of hydroprocessing catalysts. ART combines<br />
<strong>Grace</strong>’s material science, manufacturing, marketing and sales strength with Chevron’s extensive experience<br />
in operating its own refineries and leadership in technology, design and process licensing.<br />
Through new product innovation and business development, ART has grown significantly since we were<br />
formed. We acquired the Orient Catalyst Company’s hydroprocessing technologies and the HOP catalyst<br />
product line; we have become the largest shareholder in Kuwait Catalyst Company; and we have acquired<br />
and integrated new catalyst technologies from several companies, including Crosfield and Japan Energy.<br />
But most of all, thanks to you, our customers, we have become a leading global supplier of hydroprocessing<br />
catalysts with a complete line of products designed for processing resid feedstocks in fixed-bed, moving bed<br />
and ebullated bed units. We also offer a full line of distillate catalysts, some of which are represented in the<br />
pages of this publication, most notably the SmART Catalyst System ® for ULSD processing.<br />
We look to the future with enthusiasm and optimism. Our foundation will continue to be our commitment to<br />
the global refining industry to provide proven industry-leading catalysts and technical service.<br />
Sincerely,<br />
Scott K. Purnell<br />
Managing Director<br />
Advanced Refining Technologies
2 Hydrocracking<br />
6 The<br />
9 StART<br />
12 Controlling<br />
14 2011<br />
Pretreat Catalyst<br />
Development<br />
By Jifei Jia and Dave Krenzke,<br />
Advanced Refining Technologies<br />
and Theo Maesen, Chevron Lummus<br />
Global<br />
Challenges of Processing FCC<br />
LCO<br />
By Charles Olsen, Brian Watkins and<br />
Greg Rosinski, Advanced Refining<br />
Technologies<br />
Catalyst System Success<br />
Story<br />
By Geri D’Angelo, Advanced Refining<br />
Technologies<br />
Feedstock Contaminants<br />
in Diesel<br />
Hydrotreating Operations<br />
By Dave Krenzke, Advanced Refining<br />
Technologies<br />
NPRA Q&A Answers<br />
No. <strong>110</strong> / Fall 2011 / www.grace.com<br />
®<br />
<strong>Catalagram</strong><br />
A Refining Technologies Publication<br />
<strong>Catalagram</strong> <strong>110</strong><br />
ISSUE 2011<br />
Managing Editor:<br />
Scott K. Purnell<br />
Technical Editor:<br />
Charles Olsen<br />
Contributors:<br />
Geri D’Angelo<br />
Jifei Jia<br />
Dave Krenzke<br />
Charles Olsen<br />
Greg Rosinski<br />
Brian Watkins<br />
Guest Contributors:<br />
Theo Maesen<br />
Please address your comments to:<br />
betsy.mettee@grace.com<br />
Advanced Refining<br />
Technologies<br />
7500 <strong>Grace</strong> Drive<br />
Columbia, MD 21044<br />
410.531.4000<br />
Hydroprocessing<br />
Catalysts Caa<br />
talysts<br />
fr from om<br />
The Chevron<br />
& <strong>Grace</strong><br />
J JJoint<br />
oint<br />
Venture<br />
© 2011 Advanced Refining<br />
Technologies, LLC<br />
10.7.11
Hydrocracking Pretreat Catalyst<br />
Development<br />
Jifei Jia<br />
Lead Research<br />
Engineer<br />
Advanced Refining<br />
Technologies<br />
Richmond, CA, USA<br />
Theo Maesen<br />
Team Leader<br />
Distillates<br />
Chevron Lummus<br />
Global<br />
Richmond, CA, USA<br />
Dave Krenzke<br />
Regional Manager<br />
Hydroprocessing<br />
Technical Service<br />
Advanced Refining<br />
Technologies<br />
Richmond, CA, USA<br />
Introduction<br />
2 SPECIAL EDITION ISSUE No. <strong>110</strong> / 2011<br />
Demands for a cleaner environment have led to more stringent global fuel specifications. The primary target<br />
has been a reduction in sulfur content. Currently on-road diesel fuel must contain no more than 10 parts per<br />
million by weight (wppm) sulfur in both the United States and the European Union. This standard also exists<br />
in a number of Asia Pacific countries. Because the production of ultra low sulfur fuels is primarily met by hydroprocessing<br />
and hydrocracking, continuous improvement in catalyst technology is needed.<br />
In 2001, <strong>Grace</strong> and Chevron combined resources to form Advanced Refining Technologies (ART) in order to<br />
provide world class hydrotreating catalyst technology to the refining industry. Chevron also has a long standing<br />
partnership with Lummus, a leading international engineering company, called Chevron Lummus Global<br />
LLC (CLG). ART and CLG offer a full product line of premium catalysts for upgrading products varying from<br />
heavy oil to distillates. Distillate hydrotreating operations include ultra low sulfur diesel (ULSD) and FCC and<br />
hydrocracker pretreat. Refiners, such as Chevron, use these catalysts to remove sulfur and other contaminants<br />
from petroleum to economically produce more environmentally friendly transportation fuels. As the<br />
most completely integrated source for hydroprocessing technologies and services, ART and CLG can provide<br />
incremental efficiencies at every step in a project.<br />
This paper focuses on ART and CLG’s latest hydrocracker (HCR) pretreat catalyst development.<br />
Hydrocracking Pretreat Catalyst Development<br />
Good hydrodenitrogenation (HDN) activity is the primary function of the HCR pretreat catalyst because organic<br />
nitrogen compounds are detrimental to the performance of the HCR catalyst downstream. The rate<br />
limiting step in the HDN reaction pathway is aromatic ring saturation. This is because the most refractory nitrogen<br />
molecules are compounds like substituted carbazoles in which the nitrogen atom is incorporated into<br />
the aromatic ring at a fairly inaccessible position. The development of an improved catalyst therefore needs<br />
to focus on the catalyst properties that enhance ring saturation. The two critical components for optimizing<br />
catalyst performance are the support properties and active metals deposition technology as discussed in<br />
Reference 1.
Catalyst<br />
Support<br />
Development<br />
Optimizing the catalyst support<br />
starts with the relationship<br />
between the physical<br />
structure of the support and<br />
the catalytic activity for the<br />
hydrocarbon stream being<br />
processed. Figure 1, from<br />
Reference 1, demonstrates<br />
this relationship.<br />
The physical properties of<br />
the optimized support deter-<br />
mine the maximum useful<br />
metal loading and consequently<br />
the maximum potential<br />
activity. Subsequent<br />
steps in the catalyst development<br />
process seek to utilize<br />
as much of this potential activity<br />
as possible. One aspect<br />
of this process is to look at<br />
the metal-surface interaction.<br />
Altering the alumina surface<br />
chemistry with inorganic additives<br />
can have a significant<br />
effect on active site formation<br />
and has been the subject<br />
of an ongoing<br />
investigation by ART and<br />
CLG scientists. ART’s<br />
590DX has benefited from<br />
this study and utilizes a proprietary<br />
inorganic promoter<br />
to facilitate the formation of<br />
active sites during activation.<br />
An example of this work is<br />
shown in Figure 2.<br />
Metals<br />
Deposition<br />
Technology<br />
The current generation of<br />
HCR pretreat catalysts relies<br />
on the formation of DX active<br />
sites during manufacturing<br />
and activation to improve<br />
catalyst activity. The DX<br />
technology utilizes chelates<br />
to optimize metal function<br />
Relative Activity<br />
140<br />
130<br />
120<br />
<strong>110</strong><br />
100<br />
90<br />
80<br />
70<br />
60<br />
Small pores result higher surface area giving a higher<br />
Large pores result in lower surface area, but offer better<br />
Decreasing access to active sites<br />
Increasing access to active sites<br />
FIGURE 1: Relationship Between Pore Size and HDN Activity for VGO<br />
and consequently the activity<br />
and number of active sites.<br />
The chelate binds preferentially<br />
to the transition metal<br />
ion, Ni, and controls the sequence<br />
of metal ion adsorption<br />
during the impregnation<br />
step. This allows the Mo to<br />
adsorb first, followed by Ni,<br />
and reduces the chance of<br />
the Ni interacting with the<br />
alumina support. The Ni associated<br />
with the alumina will<br />
ultimately form NiS which is<br />
inactive for HDN. During activation,<br />
the chelate delays the<br />
sulfidation of Ni relative to<br />
Mo. This allows the complete<br />
formation of the MoS2 slabs<br />
before the Ni ions are sulfided.<br />
Once Ni is released by<br />
the chelate during sulfidation<br />
it moves to the edges of the<br />
MoS2 slabs to form the<br />
highly active Type II sites.<br />
There are many chelates<br />
available and they yield different<br />
performance benefits<br />
for the resulting catalyst. The<br />
ultimate choice is based on a<br />
combination of catalyst activity<br />
and manufacturing compatibility.<br />
Figure 3<br />
demonstrates the impact of<br />
the chelating agent on catalyst<br />
performance.<br />
HDN/HDS Activity Improvement, ˚F<br />
+10<br />
Base<br />
Catalyst Without<br />
Surface Modification<br />
HDN HDS<br />
Catalyst With<br />
Surface Modification<br />
ADVANCED REFINING TECHNOLOGIES CATALAGRAM ® 3<br />
120<br />
<strong>110</strong><br />
100<br />
90<br />
80<br />
70<br />
60<br />
50<br />
Relative Activity<br />
+6<br />
Base<br />
FIGURE 2: VGO HDN/HDS Performance of Catalysts<br />
With and Without Surface Modification<br />
HDN/HDS Activity Improvement, ˚F<br />
+10<br />
Base<br />
Catalyst with<br />
Organic Chelating<br />
Agent B<br />
Catalyst with<br />
Organic Chelating<br />
Agent A<br />
Catalyst without<br />
Organic Chelating<br />
Agent<br />
HDN HDS<br />
FIGURE 3: Chelating Agent Effect on HDN/HDS<br />
Performance of VGO<br />
+6<br />
Base<br />
HDN/HDS Activity Improvement, ˚C<br />
HDN/HDS Activity Improvement, ˚C
ART 590DX Commercial<br />
Performance<br />
590DX was commercialized in the middle of 2009. By mid 2010 it<br />
had already been selected by seven refineries for catalyst refill. Figure<br />
4 shows the HDN performance comparison between 590DX and<br />
the previous generation HCR pretreat, NDXi, in one of the Chevron<br />
refineries. 590DX is 20ºF (11ºC) more active than NDXi after 150<br />
days on stream.<br />
In one of the early uses of 590DX, there was an upset during the<br />
start-up which resulted in an extensive period without hydrogen flow.<br />
This compromised the start-of-run activity of 590DX. Despite the<br />
compromised start-up, the 590DX catalyst continued to activate, and<br />
became as active as the NDXi catalyst in the prior run at 2400<br />
MBBLs on stream. Considering that the start-up of the prior run was<br />
not compromised, and that the feed in the prior run was less refrac-<br />
tory than in the current run, it is clear that 590DX is a remarkably ro-<br />
bust catalyst. The commercial results are summarized in Figure 5.<br />
As a point of comparison, Figure 6 shows a different commercial op-<br />
eration for 2 cycles using the same catalyst grade. One cycle<br />
started-up normally while the other cycle had a significant upset dur-<br />
ing sulfiding. This data demonstrates a more typical result for a sig-<br />
nificant upset during start-up.<br />
As can be seen in Figure 6, the major upset during start-up resulted<br />
in a significant permanent loss of activity and a much higher deacti-<br />
vation rate. This highlights the robust response of 590DX to opera-<br />
tional upsets even during the critical start-up phase.<br />
Summary<br />
The formation of the ART joint venture in 2001 has resulted in a sig-<br />
nificantly increased rate of innovation for HCR pretreat catalysts as<br />
can be seen in Figure 7.<br />
ART’s NDXi, 590DX and the prototype for 591DX are the catalysts<br />
in this series which utilize DX technology. NDXi is ART’s previous<br />
generation catalyst for HCR pretreat applications. The catalyst is<br />
also used in ULSD applications. It has been selected more than 40<br />
times for catalyst refills and start-ups by many refineries since its<br />
commercialization in 2006. Many of the applications were selected<br />
through competitive pilot plant testing. NDXi has demonstrated a<br />
significant HDN activity advantage over the earlier generation catalyst,<br />
ART’s AT580. The base metal loading and alumina support is<br />
similar for AT580 and NDXi which confirms the advantage of<br />
chelate technology to more efficiently utilize the active metals.<br />
4 SPECIAL EDITION ISSUE No. <strong>110</strong> / 2011<br />
WABT, ˚F<br />
+80<br />
+60<br />
+40<br />
+20<br />
Base<br />
-20<br />
-40<br />
NDXi<br />
0 50 100 150 200 250<br />
Days on Stream<br />
590DX<br />
300<br />
+44<br />
+33<br />
+22<br />
Base<br />
FIGURE 4: HDN Performance Comparison Between<br />
NDXi and 590DX<br />
WABT, ˚F<br />
+80<br />
+60<br />
+40<br />
+20<br />
Base<br />
-20<br />
-40<br />
590DX<br />
0 2000 6000 8000 10000 12000<br />
Total Feed MBBLS<br />
NDXi<br />
+11<br />
-11<br />
-22<br />
+44<br />
+33<br />
+22<br />
+11<br />
Base<br />
FIGURE 5: Commercial HDN Performance Comparison<br />
Between NDXi After Smooth Start-up and 590DX<br />
After Compromised Start-up<br />
-11<br />
-22<br />
WABT, ˚C<br />
WABT, ˚C
WABT, ˚F<br />
+100<br />
+80<br />
+60<br />
+40<br />
+20<br />
Base<br />
Base<br />
0 100 200 300 400 500<br />
Days on Stream<br />
Upset During Sulfiding Normal Start-up<br />
FIGURE 6: Comparison of Typical Commercial<br />
Performance Between Normal Start-up and One With<br />
a Major Upset During Sulfiding<br />
ART is now commercializing 591DX, a wider pore version of 590DX,<br />
to handle VGO feeds with higher endpoints. The modified support<br />
properties will provide equivalent activity to 590DX but enhanced<br />
stability with heavier feeds.<br />
References<br />
1. Krenzke, Dave; Vislocky, Jim. Hydrocarbon Engineering<br />
(2007), 12(11), 57-58<br />
+55<br />
+44<br />
+33<br />
+22<br />
+11<br />
WABT, ˚C<br />
HDN Activity Improvement, ˚F<br />
+40<br />
+30<br />
+20<br />
+10<br />
Base<br />
AT580<br />
NDXi<br />
591DX<br />
590DX<br />
1972 1988 1996 2003 2006 2009 2011<br />
FIGURE 7: Hydrocracker Pretreat Catalyst<br />
Generations<br />
ADVANCED REFINING TECHNOLOGIES CATALAGRAM ® 5<br />
+22<br />
+17<br />
+11<br />
+6<br />
Base<br />
HDN Activity Improvement, ˚C
The Challenges of Processing<br />
FCC LCO<br />
Charles Olsen<br />
Director,<br />
Distillate R&D and<br />
Technical Services<br />
Brian Watkins<br />
Manager,<br />
Hydrotreating Pilot<br />
Plant & Technical<br />
Service Engineer<br />
Greg Rosinski<br />
Technical Service<br />
Engineer<br />
Advanced<br />
Refining<br />
Technologies<br />
Chicago, IL, USA<br />
6 SPECIAL EDITION ISSUE No. <strong>110</strong> / 2011<br />
FCC LCO has long been a common component of feed to diesel hydrotreaters. More recently, there has<br />
been greater interest in processing higher quantities of LCO due to economic considerations and to meet<br />
the market demand for ULSD products. LCO has a number of impacts both on the performance of the hydrotreater<br />
and on the resulting ULSD product properties. The extent of the impact depends upon a number<br />
of factors including the amount of LCO in the feed and the catalyst system used in the ULSD unit1 .<br />
It is generally accepted that the addition of LCO to the diesel feed decreases the feed reactivity and requires<br />
an increase in reactor temperature in order to meet the product sulfur target. Figure 8 summarizes<br />
pilot plant data demonstrating this effect. The figure shows the required temperature increase relative to the<br />
straight run feed as a function of product sulfur for feeds containing 15 and 30% LCO. It is clear that even<br />
low levels of LCO impact catalyst activity. At higher product sulfur about 20°F (11°C) higher temperature is<br />
required for 15% LCO and this increases to 40°F (22°C) higher temperature for 30% LCO relative to the<br />
straight run feed. As shown in Figure 8, the activity difference is even greater at lower product sulfur.<br />
In addition to the amount of LCO, the endpoint also has a significant influence on the reactivity of the feed.<br />
Figure 9 compares the performance of two 30% LCO blends relative to the straight run feed. The endpoints<br />
for the 2 LCO’s differ by just over 40°F (22°C). This difference corresponds to an additional 800 ppm of<br />
hard sulfur in the feed for the higher end point LCO, in addition to a 15% increase in PNA’s. The LCO endpoint<br />
effects are apparent even when blended into the straight run feed. The higher endpoint feed requires<br />
about 20°F (11°C) higher temperature relative to the lower endpoint feed at higher product sulfur, and nearly<br />
30°F (17°C) higher temperature for ULSD sulfur levels.<br />
Processing LCO also has a major impact on the product quality of the hydrotreated diesel. A significant<br />
problem relates to the aromatics content, as LCO’s tend to have very high concentrations of naphthalene<br />
type aromatic species which have very low cetane numbers causing the LCO to have relatively low cetane.<br />
The high PNA content also has an impact on diesel product color. This becomes important as end of run<br />
(EOR) is approached since ULSD units processing LCO blends will have product go off color at a lower<br />
temperature relative to a SR feed.
Required Temperature Increase, ˚F<br />
120<br />
100<br />
80<br />
60<br />
40<br />
20<br />
Straight Run: 37.7 API & 1.10 wt.% Sulfur<br />
15% LCO blend: 34.0 API & 1.10 wt.% Sulfur<br />
30% LCO blend: 30.7 API &1.08 wt.% Sulfur<br />
0<br />
0<br />
0 50 100 150 200 250 300 350 400 450<br />
Product Sulfur, wppm<br />
Straight Run 15% FCC LCO 30% FCC LCO<br />
Figure 8: Impact of LCO Content on Hydrotreater<br />
Performance<br />
Required Temperature Increase, ˚F<br />
120<br />
100<br />
80<br />
60<br />
40<br />
20<br />
Straight Run: D2887 endpoint 747˚F (297˚C)<br />
Low EP LCO blend: D2887 endpoint 755˚F (402˚C)<br />
High EP LCO blend: D2887 endpoint 774˚F (412˚C)<br />
0<br />
0<br />
0 50 100 150 200 250 300 350 400 450<br />
Product Sulfur, wppm<br />
Straight Run High EP LCO Low EP LCO<br />
Figure 9: LCO Endpoint Effect on Catalyst Activity<br />
Increase in Cetane Index or API Gravity<br />
9.5<br />
9.0<br />
8.0<br />
7.5<br />
7.0<br />
6.5<br />
6.0<br />
5.5<br />
5.0<br />
10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0<br />
Percent LCO in Feed<br />
Cetane Index Increase API Increase<br />
Figure 10: Cetane Increase Observed in a Commercial<br />
ULSD Unit<br />
67<br />
56<br />
44<br />
33<br />
22<br />
11<br />
67<br />
56<br />
44<br />
33<br />
22<br />
11<br />
Required Temperature Increase, ˚C<br />
Required Temperature Increase, ˚C<br />
A survey of commercial operating units shows there are a number of<br />
operating parameters which influence cetane improvement and vol-<br />
ume swell in a diesel hydrotreater, most notably hydrogen partial<br />
pressure, LHSV and feed API gravity (i.e. amount of LCO). Gener-<br />
ally speaking, as LHSV decreases, the potential cetane improve-<br />
ment and volume swell increase. Commercial ULSD experience<br />
shows that for LHSV’s around 1 hr -1 or less, cetane increases (as<br />
measured by cetane index, ASTM D-976) of about 10 numbers are<br />
achievable, provided the H 2 pressure is high enough when process-<br />
ing LCO blends. At higher LHSV’s (greater than about 1.7 hr -1 ) the<br />
potential cetane improvement decreases to about 4 numbers or less.<br />
Not surprisingly, higher pressure units tend to achieve much larger<br />
cetane increases. It has been observed that the cetane uplift is typi-<br />
cally less than 6 numbers when the unit pressure is less than 1000<br />
Psig (69 Barg) while the cetane uplift increases to 8-10 numbers as<br />
pressure increases beyond 1000 Psig (69 Barg). A more detailed<br />
discussion can be found in Reference 2.<br />
Figure 10 is a summary of commercial data from a ULSD unit using<br />
a SmART Catalyst System ® which operates at about 1300 Psig (90<br />
Barg) and slightly under 1 LHSV. The feed varies from 100%<br />
straight run to 70%+ FCC LCO. As the figure shows, the percent<br />
LCO in the feed has a big effect on the cetane upgrade. At high<br />
LCO levels the cetane index increase achieved in this unit ap-<br />
proaches 9 numbers, compared to less than 7 numbers for lower<br />
LCO levels.<br />
As might be expected, there is a significant cost in hydrogen to<br />
achieving very high cetane increases from feeds containing LCO.<br />
Figure 11 shows how the H 2 consumption increases with increasing<br />
cetane uplift from the same commercial ULSD unit. For cetane<br />
number increases of 6-7 numbers the H 2 consumption is just over<br />
850 SCFB (143 Nm 3 /m 3 ). Notice, however, that achieving increases<br />
greater than about 8 numbers come at a very large increase in H 2<br />
consumption; at 9 numbers of cetane improvement the H 2 consump-<br />
tion is about 1150 SCFB (194 Nm 3 /m 3 ), an increase of about 30%<br />
compared to a 6-7 number increase. Note the similar behavior for<br />
API increase as a function of LCO content and H 2 consumption.<br />
Another product quality issue when processing LCO containing<br />
feeds is the diesel color. It is generally accepted that the species re-<br />
sponsible for color formation in distillates are polynuclear aromatic<br />
(PNA) molecules. Some of these PNA’s are green/blue and fluores-<br />
cent in color, which is apparent even at very low concentrations.<br />
Certain nitrogen (and other polar) compounds have also been impli-<br />
cated as problems for distillate product color and product instability.<br />
Work conducted by Ma et.al. 3 concluded that the specific species re-<br />
sponsible for color degradation in diesel are anthracene, fluoran-<br />
thene and their alkylated derivatives. Other work completed by<br />
Takatsuka et.al. 4 demonstrated that the color bodies responsible for<br />
diesel product color degradation were concentrated in the higher<br />
ADVANCED REFINING TECHNOLOGIES CATALAGRAM ® 7
oiling points in the diesel (>480°F or >249°C) suggesting that color<br />
can be improved by adjusting the diesel endpoint.<br />
PNA’s such as these are readily saturated to one and two ringed<br />
aromatics under typical diesel hydrotreating conditions at start of run<br />
(SOR), but as the temperature of the reactor increases towards<br />
EOR, an equilibrium constraint is reached whereby the reverse de-<br />
hydrogenation reaction becomes more and more favorable. At some<br />
combination of ‘low’ hydrogen partial pressure and ‘high’ tempera-<br />
ture, PNA’s actually begin to form (or reform) resulting in a degrada-<br />
tion of the diesel product color.<br />
Figure 12 summarizes data from a commercial ULSD unit using ART<br />
catalysts. The data show that the product color exceeded 2.5<br />
ASTM, the pipeline color specification for diesel, at reactor outlet<br />
temperatures above 730°F (388°C). The feed to this unit contained<br />
30-50% LCO and it was operated at 1.0 LHSV and 850 Psig (59<br />
Barg) inlet pressure.<br />
Figure 13 summarizes some pilot plant data which was generated as<br />
part of a larger color study using spent CDXi, a premium CoMo cata-<br />
lyst for ULSD 5 . The figure shows a comparison of the diesel product<br />
color achieved from a SR feed and a 30% FCC LCO blend at con-<br />
stant H 2/Oil ratio and two pressures. Processing the SR feed results<br />
in acceptable color over a wide range of temperatures and for both<br />
pressures shown. The product from the LCO blend, on the other<br />
hand, goes off color (>2.5 ASTM) above 730°F (388°C) at 800 Psig<br />
(55 Barg) while at 1200 Psig (83 Barg) the temperature can exceed<br />
760°F (404°C) before going off color. This clearly demonstrates the<br />
significant impact that H 2 partial pressure has on diesel product color<br />
when processing LCO containing feeds.<br />
Processing LCO as part of the feed to a ULSD unit can be challeng-<br />
ing since the quality, quantity and endpoint of LCO affect catalyst activity<br />
and product properties. These challenges can be overcome<br />
with proper choice of catalyst system and an understanding of the<br />
impact LCO has on both unit performance and ULSD product quality.<br />
Advanced Refining Technologies is well positioned to provide assistance<br />
on how best to maximize unit performance and to take<br />
advantage of opportunities to successfully process more LCO into<br />
ULSD.<br />
References<br />
1. B. Watkins and C.Olsen, 2009 NPRA Annual Meeting, paper<br />
AM-09-78<br />
2. G.Rosinski and C.Olsen, <strong>Catalagram</strong> ® 106, Fall 2009<br />
3. X. Ma et. al., Energy and Fuels, 10, pp 91-96 (1996)<br />
4. T.Takatsuka et.al., 1991 NPRA Annual Meeting, Paper AM-91-<br />
39<br />
5. G.Rosinski, B.Watkins and C.Olsen, <strong>Catalagram</strong> ® 105, Spring<br />
2009<br />
8 SPECIAL EDITION ISSUE No. <strong>110</strong> / 2011<br />
Increase in Cetane Index or API Gravity<br />
H2 Consumption, Nm<br />
135<br />
9.5<br />
145 155 165 175 185 195<br />
3 /m3 9.0<br />
8.5<br />
8.0<br />
7.5<br />
7.0<br />
6.5<br />
6.0<br />
5.5<br />
5.0<br />
800 850 900 950 1000 1050 <strong>110</strong>0 1150 1200<br />
H 2 Consumption, SCFB<br />
Cetane Index Increase API Increase<br />
Figure 11: H 2 Consumption Increases with Cetane<br />
Uplift<br />
Product Color, ASTM<br />
304 316 327 338 349 360 371 382 393 404 416<br />
3.5<br />
3.0<br />
2.5<br />
2.0<br />
1.5<br />
1.0<br />
0.5<br />
Outlet Temperature, ˚C<br />
0.0<br />
580 600 620 640 660 680 700 720 740 760 780<br />
Outlet Temperature, ˚F<br />
Figure 12: ULSD Product Color From a Commercial<br />
ULSD Unit<br />
Product Color, ASTM<br />
338 349 360 371 382 393 404 416 427<br />
4.0<br />
3.5<br />
3.0<br />
2.5<br />
2.0<br />
1.5<br />
1.0<br />
0.5<br />
Straight Run<br />
30% FCC LCO<br />
Temperature, ˚C<br />
0.0<br />
640 660 680 700 720 740 760 780 800<br />
Temperature, ˚F<br />
Low Pressure<br />
High Pressure<br />
High Pressure<br />
Low<br />
Pressure<br />
Figure 13: Comparison of Product Color for SR and<br />
30% LCO
StART® Catalyst System<br />
Success Story<br />
Geri D’Angelo<br />
Senior Technical<br />
Service Engineer<br />
Advanced Refining<br />
Technologies<br />
Chicago, IL, USA<br />
Hydrotreating coker naphtha poses some unique challenges for the refiner. All coker naphthas contain a<br />
high concentration of olefins/diolefins, and naphthas from delayed cokers also contain silicon. In addition,<br />
these stocks have a significantly higher concentration of sulfur and nitrogen than straight run naphthas.<br />
Even though the typical processing scheme involves blending the coker stock with straight run naphtha, is-<br />
sues of high exotherms, pressure drop increase and silicon poisoning often control the cycle length.<br />
Advanced Refining Technologies (ART) has long had a strong position in naphtha hydrotreating technology.<br />
AT535 is one of the keys to the technology with an outstanding commercial track record with over 300 users<br />
worldwide in both straight run and coker naphtha service. In addition to the active catalyst, ART supplies ac-<br />
tive silicon guard materials to economically supplement the silicon capacity and provide activity grading. Activity<br />
grading is an important aspect of coker naphtha processing. The high heat release resulting from olefin<br />
saturation can cause polymerization and a subsequent pressure drop problem. By grading hydrogenation<br />
activity from low to high (active guard to catalyst) the temperature rise is spread out over a larger portion of<br />
the catalyst bed and the potential for polymerization is mitigated.<br />
As previously mentioned, units that process coker naphtha streams typically contain silicon which comes<br />
from the delayed coker. In order to suppress foaming at the coker, an anti-foam agent such as silicone oil<br />
(i.e. polydimethylsiloxane) is used. The silicone oil breaks down during the coking process to lower molecular<br />
weight fragments consisting of modified silica gels. These fragments end up primarily in the naphtha<br />
range fraction, although small quantities are also found in the kerosene and diesel fractions.<br />
Silicon can also be found in synthetic crudes because the process of making synthetic crude often involves<br />
a coking step. Many crude suppliers have also used additives containing silicon in drilling processes, and<br />
ADVANCED REFINING TECHNOLOGIES CATALAGRAM ® 9
pipeline companies are using silicon containing additives injected<br />
into the crudes for both flow enhancing performance and foaming is-<br />
sues. As a result, even refiners who don’t have coking capabilities<br />
may run into Si contamination issues.<br />
ART introduced the StART TM (Silicon tolerance by ART) Catalyst<br />
System in 2002 as a way for refiners to extend the cycle length of<br />
their naphtha treaters, while at the same time facing increased inci-<br />
dence of Si contamination. This technology combines a state of the<br />
art silicon guard material, AT724G, along with the active HDS and<br />
HDN catalyst AT535. The StART technology has been widely ac-<br />
cepted in the industry for the combination of high Si capacity and<br />
high activity in naphtha service. Figure 14 compares the Si capacity<br />
of AT724G and AT535 with some competitor catalysts. AT535, by itself,<br />
has essentially the same Si capacity as the competitor catalysts<br />
while AT724G has over 30% higher Si capacity.<br />
Figure 15 shows a commercial example of the how the StART sys-<br />
tem can increase the cycle length over competitive silicon tolerant<br />
catalysts. In this case, the StART system more than doubled the<br />
cycle length over several previous cycles that used catalyst from<br />
competitor A. This refiner was processing coker naphtha and living<br />
with very short cycles of between 7 and 13 months. After installing<br />
the StART System the cycle length more than doubled. Based on<br />
this significantly improved performance this refiner continues to use<br />
a StART system in their coker naphtha unit.<br />
Another recent success story involves a naphtha hydrotreating<br />
(NHT) unit in the Gulf Coast region which had been experiencing<br />
cycle lengths of between 12-18 months due to silicon poisoning of<br />
the catalyst. ART worked with the refiner to design a catalyst system<br />
which would extend the cycle length to 24 months or more.<br />
The NHT unit consists of two reactors in series and produces the<br />
feed for a downstream reformer. The feed sulfur and nitrogen get as<br />
high as 2500 ppm and 140 ppm respectively, and the product speci-<br />
fications for sulfur and nitrogen are
vation was consistent with the estimated Si levels discussed previ-<br />
ously, and both indicated the lead reactor was approaching the cata-<br />
lyst Si capacity.<br />
The performance of the unit was outstanding and easily exceeded<br />
the expected 24 month run length. After 40 months on stream, the<br />
estimated Si loading on the catalyst combined with the exotherms<br />
shown in Figure 17 indicated that it was time to change out the lead<br />
reactor. The lead reactor catalyst was replaced with fresh AT724G<br />
and AT535. The NHT was re-started with the new catalyst in the lead<br />
reactor and the AT535 in the lag reactor which had already been in<br />
use for 40 months. At the time of the lead reactor catalyst change,<br />
the NHT was still producing on spec product and the deactivation<br />
rate of the catalyst was 0.3-0.5°F/month (0.2-0.3°C/month).<br />
ART received spent catalyst samples from the lead reactor which<br />
were lab regenerated. These were analyzed for a variety of poisons<br />
including silicon. The analysis showed that the catalyst picked up<br />
significant amounts of silicon as well as arsenic. Figure 18 shows<br />
the silicon and arsenic profile through the catalyst bed in the lead re-<br />
actor. The analysis shows that silicon and arsenic got well into the<br />
catalyst bed, consistent with the observation of the exotherm shifting<br />
to the lower part of the bed. The results show that AT724G picked<br />
up 20 wt% silicon at the top of the catalyst bed and around 15 wt%<br />
Si at the bottom of the AT724G bed. This indicates there was a little<br />
more silicon capacity in the AT724G, but Si was definitely breaking<br />
through to the active catalyst. The amount of silicon in the spent<br />
AT535 samples was 8 wt% near the top of the catalyst bed and < 2<br />
wt% at the bottom of the bed. The spent catalyst data show that lit-<br />
tle if any Si and As made it all the way through the lead reactor into<br />
the lag reactor. While there may have been some capacity left to<br />
trap more of both Si and As, the right decision was made to change<br />
out and protect the lag reactor catalyst from poisoning.<br />
The use of ART’s StART catalyst system in this unit was a tremen-<br />
dous success. The cycle length was increased from 12-18 months<br />
to over five years, with full unit turnaround expected in two more<br />
years.<br />
The SOR temperature after the lead reactor catalyst change-out was<br />
the same as the first StART loading despite the fact that the lag reactor<br />
was not changed out. The catalyst deactivation in the current<br />
cycle continues to run between 0.3-0.5°F/month (0.2-0.3°C/month)<br />
and has currently been on stream for 18 months since the changeout<br />
of the lead reactor. The unit is on track to run through the end of<br />
2013 when it is anticipated to shut down due to silicon poisoning. At<br />
the end of the current cycle the lag reactor will have been on line for<br />
7 years.<br />
Percent Delta Temperature Rise<br />
100%<br />
90%<br />
80%<br />
70%<br />
60%<br />
50%<br />
40%<br />
30%<br />
20%<br />
10%<br />
0%<br />
Top Mid 1<br />
Days on Stream<br />
Mid 2 Mid 3 Bottom<br />
Figure 17: Reaction Exotherms Move Through The<br />
Catalyst Bed<br />
Wt% Si on Catalyst<br />
25%<br />
1.0%<br />
Silicon 0.9%<br />
20%<br />
AT724G<br />
Arsenic 0.8%<br />
0.7%<br />
15%<br />
0.6%<br />
0.5%<br />
10%<br />
0.4%<br />
AT535<br />
0.3%<br />
5%<br />
0.2%<br />
0.1%<br />
0%<br />
0%<br />
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0<br />
Fraction of Lead Reactor Catalyst Volume<br />
Figure 18: Silicon and Arsenic Profiles Through the<br />
Catalyst Bed<br />
ADVANCED REFINING TECHNOLOGIES CATALAGRAM ® 11<br />
Wt% As on Catalyst
Controlling Feedstock Contaminants<br />
in Diesel Hydrotreating Operations<br />
Improving Unit Performance Requires Strategies for<br />
Avoiding Rapid Deactivation of Hydrotreating Catalyst<br />
from Contaminants<br />
Dave Krenzke<br />
Regional Manager<br />
Hydroprocessing<br />
Technical Service<br />
Advanced Refining<br />
Technologies<br />
Richmond, CA, USA<br />
12 SPECIAL EDITION ISSUE No. <strong>110</strong> / 2011<br />
Higher feedstock volumes contaminated with catalyst poisons such as those listed in Table 1 are being<br />
processed in high complexity refining facilities. Many of these hydrocarbons are from sources recently introduced<br />
into the global crude market in significant quantities over the past five years. For example, Canadian<br />
Synbit and Dilbit crudes will come to make up a significant fraction of feedstocks to North American refineries.<br />
North American refiners and their technology partners are just now discovering the challenges encountered<br />
with upgrading these “cheap” crudes to ULSD specifications.<br />
In other producing regions such as deep offshore Brazil, preliminary reports indicate that hydrocarbons from<br />
the Tepi oil field, which has been called the greatest oil discovery in the past 100 years, may be relatively<br />
high in nitrogen content. What has also been noticeable are the new challenges certain refiners are facing<br />
with the “redistribution” of certain crudes. For example, the heavy Venezuelan crudes (e.g., Merey Crude:<br />
16° API, 11.5 UOP K-factor, 2.7wt% S, 3600 ppm N, etc.) that U.S. Gulf Coast refiners began processing<br />
back in the 1980s are finding a new market in emerging refining regions such as India. These new highcomplexity<br />
facilities may nonetheless be unfamiliar with their contaminants.
Contaminant Feed Guidlines Common Source Remedies<br />
Si < 1.0 wppm<br />
The contaminants found in these feeds can be particularly harmful to<br />
hydrotreating catalyst activity. In anticipation of the challenges in-<br />
volved in processing these types of feedstocks, there has been a no-<br />
ticeable trend towards designing new diesel hydrotreating units at<br />
higher hydrogen partial pressures to compensate for catalyst activity<br />
loss from these poisons. Instead of operating at higher H 2 partial<br />
pressures, older units can be modified to meet Euro 4 or ULSD tar-<br />
gets, for example, with higher catalyst volume and lower space ve-<br />
locity.<br />
Contaminant Control<br />
In general, silicon (Si) carryover from Si-based antifoam agents used<br />
in delayed coker operations will plug up catalyst porosity. Sodium<br />
(Na) and calcium (Ca) can originate from seawater exposure, poor<br />
desalting or caustic sources in the refinery. Arsenic (As) is becoming<br />
more of a common issue as more synthetic crudes and crudes from<br />
Africa and Russia are processed. It is present at low levels throughout<br />
the whole boiling range for synthetic crudes, which is why arsenic<br />
traps may need to be employed. Depending on the future cost<br />
of crudes, the shale oil based hydrocarbons from Colorado that first<br />
drew interest in the 1980s, that are relatively low in sulfur, may again<br />
become marketable in large quantities. However, they are high in As<br />
Anti foam from<br />
delayed cokers<br />
Na, Ca < 0.5 wppm Sea water; caustic<br />
As < 250 ppb<br />
Pb, P < 0.5 wppm<br />
Ni + V<br />
Fe<br />
< 1.0 wppm total<br />
for Ni + V + Fe<br />
< 1.0 wppm total<br />
for Ni + V + Fe<br />
Crudes from W. Africa,<br />
Russia, synthetic crudes<br />
Gasoline slop tanks,<br />
imported feeds<br />
Resid; heavy feeds<br />
Soluble: corrosion; insoluble:<br />
unfiltered particulates<br />
C7 insolubles MCR < 100 wppm < 0.5 wt% Resid; heavy feeds<br />
Table 1: Feed Contaminants Commonly Found in Crudes<br />
Guard catalysts<br />
Improved desalting;<br />
Guard catalysts;<br />
Don’t send spent<br />
caustic to feed<br />
tanks or units<br />
High Ni guard catalysts<br />
Don’t process feeds<br />
containing Pb or P<br />
Better feed distillation;<br />
Guard Catalysts;<br />
Bed grading<br />
Inert guard material<br />
with high void<br />
space; Fe-traps (soluble);<br />
Top bed skimming<br />
Better feed<br />
distillation;<br />
Guard catalysts;<br />
Bed grading<br />
and Ni. Also, phosphorous (P) is becoming a problem with some<br />
feeds. The presence of nickel (Ni) and vanadium (V) in heavier<br />
feeds has always been a significant concern to hydrotreating operations.<br />
Their presence may also mean entrainment is occurring.<br />
These Ni + V contaminants are found in significant concentrations in<br />
deep-cut VGO streams (e.g., 1-2 ppm), which is why end-point control<br />
is critical with heavy feedstocks. In hydrotreating operations, either<br />
one of these elements will behave like coke and plug up the<br />
catalyst. Si will also plug up catalyst pores. Although Fe in naphthenic<br />
crudes may not be as significant a problem as Ni + V poisoning<br />
of hydrotreating catalyst, it nonetheless is a corrosion precursor<br />
and leads to FeS formation. Soluble Fe generated from naphthenic<br />
or acidic-based crudes can lead to corrosive products entering catalyst<br />
beds.<br />
ADVANCED REFINING TECHNOLOGIES CATALAGRAM ® 13
Answers to the 2011 NPRA<br />
Q&A FCC Questions<br />
Contaminants/Analytical<br />
2. Are there any standard sampling and analytical methods that<br />
can be used in the refinery labs to accurately determine the sili-<br />
con content in the feed to the coker naphtha hydrotreater?<br />
Geri D’Angelo<br />
Accurately measuring silicon in naphtha streams can be done but it<br />
takes a bit of work to get a representative sample of the naphtha.<br />
The silicon in the coker naphtha depends on the type and amount of<br />
antifoam chemical at the delayed coker unit. Delayed cokers have<br />
cycles ranging anywhere between 8 - 24 hours. The coker unit is<br />
continually producing a coker naphtha stream during these cycles<br />
which is typically being sent from the fractionator straight into the<br />
naphtha hydrotreater feed drum. The antifoam chemical is usually<br />
not added for the entire coker cycle. This means that the silicon in<br />
the naphtha stream will vary with the timing of the coker cycle. In<br />
order to get a representative amount of silicon in the coker naphtha<br />
stream a composite should be made of hourly samples mixed to-<br />
gether for the time of the cycle. For example, for an eight hour cycle,<br />
eight samples would be mixed and the composite sample analyzed<br />
for silicon. To measure the silicon an ICP-MS (Inductively Coupled<br />
Plasma Mass Spectrometry) instrument can be used. This instru-<br />
ment/method can measure very low silicon concentrations.<br />
Fouling/Poisons<br />
4. What has been your experience with antimony and phospho-<br />
rous poisoning on hydrotreating catalyst performance? What<br />
is the maximum level?<br />
Charles Olsen<br />
Phosphorous (P) contamination in oil has been traced to frac fluids<br />
that are often used in crudes from the Western Canadian Sedimentary<br />
Basin. The source is diphosphate esters which are soluble in<br />
the crude oil. Refineries that run large percentages of light Western<br />
Canadian crude have reported crude column and crude furnace fouling<br />
for many years. Improvements made to crude columns to minimize<br />
fouling have transitioned the depositing of phosphorous to the<br />
downstream hydrotreaters.<br />
14 SPECIAL EDITION ISSUE No. <strong>110</strong> / 2011<br />
Other sources of phosphorous include gasoline slop tanks, imported<br />
feeds and lube oil wastes. If phosphorous does manage to make its<br />
way into the hydrotreater it will poison the active sites of the catalyst<br />
causing a loss in activity. A level of 1 wt% of phosphorous on the<br />
catalyst results in roughly 10°F (6°C) loss in activity. ART recommends<br />
that a feed content of < 0.5 wppm be maintained whenever<br />
possible, as well as the use of feed filters to assist in trapping of<br />
phosphorous sediment.<br />
Historically, phosphorous contamination has not been very common<br />
but, with the increasing use of opportunity crudes, it is being ob-<br />
served more frequently. A recent example is summarized in the<br />
table below which shows the results of some spent catalyst analysis<br />
from a diesel unit. This unit experienced extremely rapid catalyst<br />
deactivation shortly after start up. It was so severe that within sev-<br />
eral months the unit required an unplanned turnaround and fresh<br />
catalyst was installed. The spent catalyst analysis indicates the cat-<br />
alysts were exposed to high levels of several poisons including<br />
sodium and phosphorous. The contaminants penetrated well into<br />
the catalyst bed. The level of contaminants indicates that the cata-<br />
lyst in the top half of the bed had lost over 60°F (33°C) of activity.<br />
Sulfiding<br />
6. What is the minimum hydrogen sulfide required in the recy-<br />
cle gas for units with low sulfur feed? Do refiners inject sulfur<br />
compounds to maintain a minimum concentration?<br />
Gordon Chu<br />
There is no minimum hydrogen sulfide requirement as long as the<br />
feed contains some sulfur as the sulfided catalyst is very resistant to<br />
sulfur loss under normal process conditions. We are not aware of<br />
any refiners adding sulfur compounds to maintain a minimum H 2S<br />
concentration during the process cycle.<br />
Na, wt% P, wt%<br />
GSK-9 1.71 4.43<br />
GSK-6A 0.21 2.29<br />
Top Catalyst 1.66 2.29<br />
Table 2: Commercial Example of Phosphorus<br />
Poisoning
Start-Up and Shut-Down<br />
7. Is there any harm adding cracked stocks too quickly after<br />
break-in following catalyst activation? What is typical introduc-<br />
tion rate?<br />
Ben Sim<br />
Introducing cracked stocks too early after sulfiding will cause notice-<br />
able loss in activity. Coke precursor molecules in cracked feeds will<br />
have a tendency to form coke over the fresh and highly active sites<br />
on the catalyst. Delaying the introduction of cracked stocks for at<br />
least 3 days after sulfiding will allow the catalyst activity to be passi-<br />
vated which helps to minimize these effects.<br />
After running for three days on straight run the cracked material should<br />
be added to the feed stream gradually. ART typically recommends<br />
adding the cracked feed in small increments every shift making sure<br />
the reactor exotherm remains under control and within acceptable limits<br />
before increasing the cracked feed amount any further.<br />
Loading and Unloading<br />
9. Have you successfully dumped, screened and reloaded spent<br />
hydrotreating or hydrocracking catalyst without regeneration<br />
during a turnaround? Can you share any best practices during<br />
this operation to avoid problems on restart?<br />
Greg Rosinski<br />
Spent hydroprocessing catalyst is pyrophoric due to small particu-<br />
lates of iron sulfide scale that are present, so care must be taken to<br />
minimize the exposure of the spent catalyst to air. In addition, spent<br />
sulfided catalyst has some coke on it and it will slowly oxidize in air.<br />
If the spent catalyst is exposed to air, it will slowly heat up and, if iron<br />
sulfide is present, it will combust which may ignite the coke or other<br />
residual hydrocarbon on the catalyst.<br />
The key to this procedure is to have competent and experienced<br />
personnel performing the required tasks. The reactor must be thoroughly<br />
swept of hydrocarbons, and a nitrogen purge should be kept<br />
on the reactor at all times. During the unloading, the screener and<br />
the dump nozzle should be continuously purged. The containers<br />
that will hold the catalyst during unloading should be blanketed with<br />
nitrogen or have dry ice placed inside until ready for loading. The<br />
containers should not be open to the atmosphere. The loading<br />
should be done under inert conditions with experienced personnel.<br />
When preparing your procedure, make sure to involve your refinery<br />
EH&S group and give careful consideration to all aspects of the<br />
process to ensure that you take all the precautions necessary.<br />
Kerosene/LCO Processing<br />
10. In treating kerosene, what factors play into the decision to<br />
use hydrotreating versus sweetening processes such as caustic<br />
treating?<br />
Dave Krenzke<br />
The decision to use hydrotreating or a sweetening process depends<br />
on the types of sulfur in the kerosene and the product sulfur target.<br />
Hydrotreating can remove all types of sulfur compounds and there-<br />
fore the sulfur content of the product is only limited by the process<br />
conditions and catalyst activity. The sweetening process only re-<br />
moves mercaptan sulfur so the product sulfur is limited to the non-<br />
mercaptan sulfur in the feed.<br />
11. What LCO 95% distillation point do you target for optimizing<br />
ULSD production? Do you see a significant catalyst life penalty<br />
with increased LCO cut point?<br />
Brian Watkins<br />
The addition of LCO to a ULSD hydrotreater has several effects<br />
such as increased hydrogen consumption, higher required reactor<br />
temperatures and possibly shorter cycle time. Figure 19 summa-<br />
rizes some of pilot plant data comparing a SR and a SR/LCO feed<br />
blend. It shows that the SR diesel requires a 43°F (24°C) increase<br />
in temperature to go from 100 ppm sulfur down to 10 ppm sulfur.<br />
The 20% LCO blend requires almost 20°F (11°C) higher temperature<br />
to achieve the same product sulfur relative to the SR feed. The<br />
product from the LCO blend also has a 2 to 3 number lower API<br />
compared to the SR product, and hydrogen consumption increases<br />
significantly for the LCO blend due to saturation of additional pol-<br />
yaromatic compounds found in the LCO. These latter consequences<br />
set limits on the amount of LCO which can be processed<br />
and still meet product cetane specifications and also meet hydrogen<br />
availability constraints.<br />
One option to re-gain some of the lost activity in adding additional<br />
LCO is to change the end point of the LCO in the feed. ART com-<br />
pleted pilot plant testing on an LCO from the same FCC which had<br />
been cut at two different end points. Table 3 lists the analysis of the<br />
two LCO feeds and shows that the end points differed by about 40°F<br />
(22°C). The decrease in endpoint lowers the total sulfur by almost<br />
1000 ppm and total nitrogen decreases by 129 ppm.<br />
A comparison of activity on the two LCO feeds blended at 30% by<br />
volume into SR feed is shown in Figure 20. Over 30°F (17°C) higher<br />
temperature is required to treat the higher endpoint feed to meet the<br />
ULSD specification. This difference in activity corresponds to a significant<br />
decrease in the hydrotreater cycle length.<br />
ADVANCED REFINING TECHNOLOGIES CATALAGRAM ® 15
The addition of LCO has a major impact on activity for both the low<br />
and high endpoint LCO materials. The required temperature in-<br />
crease for ULSD in going from 0 to 30% LCO for the lower endpoint<br />
material is about 1.2°F (0.7°C) per percent LCO. Processing the<br />
higher endpoint LCO increases the required temperature to about<br />
1.4°F (0.8°C) per percent LCO. Figure 21 demonstrates this more<br />
clearly in the form of a plot of the required temperature increase as a<br />
function of LCO content. Notice from the chart that the activity ef-<br />
fects are not exactly linear with increasing LCO content. The first<br />
15% LCO has a larger impact on activity than the next 15%.<br />
Volume Gain/Conversion<br />
12. What sets the volume gain in ULSD units? How much does<br />
lowering the space velocity increase the volume gain? How<br />
much volume gain can be expected for each feed component?<br />
Charles Olsen<br />
There are a number of parameters which influence volume gain in a<br />
ULSD unit. Hydrogen partial pressure and LHSV are two key oper-<br />
ating conditions which have a large effect on the product volume in-<br />
crease. Catalyst selection also plays an important role since at<br />
higher pressures NiMo catalysts have a higher aromatics saturation<br />
activity compared to CoMo catalysts.<br />
Figure 22 shows the total volume yield on a fresh feed basis that has<br />
been achieved in several commercial diesel hydrotreaters as a func-<br />
tion of unit LHSV. Generally speaking, as LHSV decreases the po-<br />
tential volume swell increases. At a LHSV around 1 hr -1 or less,<br />
total product volume increases of 6-7% or more are achievable (pro-<br />
vided the H 2 pressure is high enough), while at a LHSV greater than<br />
about 1.7 hr -1 the total product volume increase is about 1-2% .<br />
Type<br />
LCO<br />
(low FBP)<br />
16 SPECIAL EDITION ISSUE No. <strong>110</strong> / 2011<br />
LCO<br />
(high FBP)<br />
API 18.31 15.31<br />
Sulfur, wt% 0.948 1.041<br />
Nitrogen, ppm 708 837<br />
Aromatics, lv% 66.86 68.81<br />
mono-,lv% 22.65 18.44<br />
poly-, lv%<br />
Dist., D2887, ˚F/C<br />
44.21 50.37<br />
IBP 249/121 256/124<br />
10% 425/218 432/222<br />
50% 531/277 550/288<br />
70% 600/316 620/327<br />
90% 677/358 699/371<br />
FBP 772/411 812/433<br />
Table 3: Comparison of Boiling Point Reduction on<br />
LCO<br />
Required Temperature Increase,˚F<br />
70<br />
60<br />
50<br />
40<br />
30<br />
20<br />
10<br />
0<br />
0<br />
0 50 100 150<br />
Product Sulfur, ppm<br />
SR<br />
LCO<br />
Figure 19: Activity Comparison on SR and Blended<br />
SR/LCO<br />
Required Temperature Increase,˚F<br />
160<br />
140<br />
120<br />
100<br />
80<br />
60<br />
40<br />
20<br />
0<br />
0<br />
0 100 200 300 400 500 600<br />
Product Sulfur, ppm<br />
SR 30% Hi FBP LCO<br />
30% Lo FBP LCO<br />
Figure 20: Impact of Endpoint Reduction on<br />
Hydrotreating Performance<br />
WABT Increase to Achieve<br />
Product Sulfur, ˚F<br />
50<br />
45<br />
40<br />
25<br />
35<br />
20<br />
30<br />
25<br />
15<br />
20<br />
15<br />
10<br />
10<br />
5<br />
5<br />
0<br />
0<br />
0 5 10 15 20 25 30 35<br />
% LCO<br />
Hi EP LCO Lo EP LCO<br />
Figure 21: Activity Comparisons at different LCO FBP<br />
and Concentration<br />
35<br />
30<br />
25<br />
20<br />
15<br />
10<br />
5<br />
80<br />
70<br />
60<br />
50<br />
40<br />
30<br />
20<br />
10<br />
Required Temperature Increase,˚C<br />
Required Temperature Increase,˚C<br />
WABT Increase to Achieve<br />
Product Sulfur, ˚C
Total Product Volume, % of Fresh Feed<br />
<strong>110</strong><br />
109<br />
108<br />
107<br />
106<br />
105<br />
104<br />
103<br />
102<br />
101<br />
100<br />
0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50<br />
Refiner A Refiner B Refiner C Refiner D<br />
Refiner E Refiner F<br />
LHSV, hr1<br />
Figure 22: Effect of LHSV on Volume Swell in<br />
Commercial Units<br />
Total Product Volume, % of Fresh Feed<br />
27.6<br />
<strong>110</strong><br />
41.4 55.2 68.9 82.7 96.5 <strong>110</strong>.3 124.1 137.9 151.7<br />
109<br />
108<br />
107<br />
106<br />
105<br />
104<br />
103<br />
102<br />
101<br />
Refiner A Refiner B Refiner C Refiner D<br />
Refiner E Refiner F<br />
Reactor Inlet Pressure, Barg<br />
100<br />
400 600 800 1000 1200 1400 1600 1800 2000 2200<br />
Reactor Inlet Pressure, Psig<br />
Figure 23: Effect of Unit Pressure on Volume Swell<br />
Total Product Volume, % of Fresh Feed<br />
<strong>110</strong><br />
109<br />
108<br />
107<br />
106<br />
105<br />
104<br />
103<br />
102<br />
101<br />
0.934 0.924 0.914 0.904 0.894 0.884 0.874 0.864 0.854 0.844 0.834 0.824 0.814<br />
100<br />
18.0 20.0 22.0 24.0 26.0 28.0 30.0 32.0 34.0 36.0 38.0 40.0 42.0<br />
Refiner A Refiner B Refiner C Refiner D<br />
Refiner E Refiner F<br />
Feed Specific Gravity<br />
Feed API Gravity<br />
Figure 24: Feed Gravity Has a Significant Impact on<br />
Volume Swell<br />
Of course LHSV is not the only parameter which can influence the<br />
volume swell. H 2 partial pressure also has a significant effect. Fig-<br />
ure 23 summarizes the total product volume yield as a function of<br />
unit pressure for the commercial units shown in Figure 22. Not sur-<br />
prisingly, higher pressure units tend to achieve a much higher level<br />
of volume swell. In these examples, the volume increase is typically<br />
less than 3% when the unit pressure is less than 1000 Psig (69<br />
Barg). The total volume swell increases to 4-7% as the unit pres-<br />
sure increases beyond 1000 Psig (69 Barg). The data in Figures 22<br />
and 23 also suggest there is a practical limit to the volume swell<br />
achieved from typical hydrotreating. A comparison of the volume<br />
swell achieved by Refiners A and B shows they are roughly the<br />
same for both units despite the large difference in operating pres-<br />
sure at similar LHSV.<br />
The volume swell also varies significantly with feedstock. Figure 24<br />
summarizes how the total product volume yield correlates with the<br />
gravity of the feed. In general, the product volume swell increases<br />
as the feed API decreases or specific gravity increases. In other<br />
words, as more FCC LCO is added to the feed the potential volume<br />
swell from hydrotreating increases.<br />
As mentioned previously, the catalyst will also have an impact on the<br />
degree of volume swell achieved in a hydrotreater. It is well known<br />
that NiMo catalysts have higher aromatic saturation activity than<br />
CoMo catalysts, and therefore a NiMo catalyst is expected to deliver<br />
greater volume swell. Figure 25 summarizes pilot plant data which<br />
demonstrates this. These data were generated using a 25% LCO<br />
containing feed, and shows that the NiMo catalyst results in 1-2<br />
numbers higher total product volume increase compared to the<br />
CoMo catalyst.<br />
Total Product Volume, % of Fresh Feed<br />
105.0<br />
104.5<br />
104.0<br />
103.5<br />
103.0<br />
102.5<br />
102.0<br />
101.5<br />
0.50 0.70 0.90 1.10 1.30 1.50 1.70 1.90 2.10 2.30<br />
LHSV, hr -1<br />
NiMo CoMo<br />
Figure 25: The Catalyst Type Effects Volume Swell<br />
ADVANCED REFINING TECHNOLOGIES CATALAGRAM ® 17
13. What considerations are being given to include mild hydroc-<br />
racking in your high pressure ULSD unit?<br />
Robert Wade<br />
Many refiners consider complimenting their existing ULSD HDT cat-<br />
alyst with a hydrocracking catalyst to improve cold flow properties.<br />
This is accomplished through end point reduction. The hydrocrack-<br />
ing catalyst used is usually the type that preferentially cracks large<br />
straight chain paraffins. For properly designed catalyst systems this<br />
will be economic as heavier feeds may be processed while meeting<br />
stringent diesel cold flow properties, and simultaneously not severely<br />
reducing diesel selectivity.<br />
Some refiners find it economic to retrofit their existing ULSD units so<br />
that they can run full range VGO. This option is very dependent on<br />
the design pressure of the reactors as the fouling rate will be consid-<br />
erably higher for a VGO service. In addition, the capital expense re-<br />
quired to meet recovery section requirements may be a major<br />
consideration.<br />
Hydrogen Optimization/HPNA’s<br />
14. With limited hydrogen availability for desulfurization of<br />
diesel, what criteria influence the optimization of hydrogen consumption<br />
between the FCC Pretreat and ULSD units? What catalytic<br />
options exist to achieve the desired balance of<br />
consumption?<br />
Greg Rosinski<br />
For any given feed, hydrogen consumption is a function of hydrogen<br />
partial pressure, LHSV, H 2/Oil and catalyst. For the most part, the<br />
first three variables are fixed for a given unit, since throughput re-<br />
duction is not an economical choice. Thus, catalyst selection is one<br />
of the few variables which refiners are willing to consider changing.<br />
CoMo catalysts have lower hydrogen consumption than NiMo cata-<br />
lysts due to lower aromatic saturation activity. At equivalent product<br />
sulfur, using all CoMo catalyst in the FCC Pretreater will lower the<br />
hydrogen consumption with a longer cycle in terms of HDS activity,<br />
but at the cost of lower conversion in the FCC and higher LCO<br />
yields. Using all NiMo catalyst in the FCC Pretreater will result in<br />
higher FCC gasoline yields and lower LCO yields due to higher PNA<br />
saturation, but a shorter cycle life in terms of HDS activity.<br />
With regards to the ULSD units, if the unit is high pressure, using a<br />
NiMo catalyst will result in higher aromatic and PNA saturation. This<br />
may be beneficial if cetane upgrade is desired; however, there may<br />
be a diminishing return on hydrogen for the incremental cetane upgrade<br />
over a CoMo catalyst.<br />
18 SPECIAL EDITION ISSUE No. <strong>110</strong> / 2011<br />
ART can help optimize both FCC Pretreater and ULSD performance<br />
based upon the refiner’s needs, including hydrogen consumption,<br />
cetane uplift, and cold flow properties. ART provides the ApARTTM and SmART ® staged catalyst systems for FCC Pretreat and ULSD<br />
applications, respectively. ART has helped many refiners manage<br />
hydrogen consumption in both units by using staged catalyst systems<br />
utilizing NiMo, CoMo and NiCoMo catalysts optimized to enhance<br />
HDS, HDN or HDPNA activity for a given feedstock.<br />
Furthermore, ART’s relationship with <strong>Grace</strong> Davison can enhance<br />
the unit optimization to include the FCC unit as well as the FCC pretreater<br />
and the ULSD unit. Utilizing the technical resources of both<br />
ART and <strong>Grace</strong> Davison, the refiner can gain a more comprehensive<br />
understanding of the interactions and dependence of these units on<br />
each other in terms of hydrogen consumption and product property<br />
enhancement.<br />
Hydroprocessing<br />
16. Pre-hydrotreated feeds and crudes look easy to process on<br />
paper. Why is it more difficult than expected to process pre-hydrotreated<br />
feeds in a hydroprocessing unit?<br />
Brian Watkins<br />
Opportunity feedstocks, having already been processed through<br />
conventional refinery processes, pose unexpected challenges to re-<br />
finers wishing to incorporate them into the distillate pool. Some of<br />
these streams have proven to be significantly more difficult to<br />
process underscoring the fact that it is important to understand the<br />
potential impacts of processing new feed streams in order to avoid<br />
unpleasant surprises. Significant differences in feed reactivity for<br />
various pre-processed feed components are not necessarily antici-<br />
pated from observing the usual bulk feed analyses.<br />
When considering the use of synthetic crudes an understanding of<br />
the upstream processing is important. Production of synthetic fuels<br />
involves a combination of several processes in order to accommo-<br />
date downstream processing. These upstream processes include<br />
coking or an ebullating bed resid operation, followed by a hydrotreat-<br />
ing or hydrocracking operation in order to produce a lighter grade<br />
material. These hydrocracking units tend to operate at severe con-<br />
ditions in conjunction with high hydrogen partial pressures. At these<br />
conditions, the removal of all the easy, less refractory sulfur is read-<br />
ily achieved, and the majority of the multi-ring aromatics are satu-<br />
rated. This leaves a product that is relatively low in sulfur and PNA’s<br />
and, when added to the feed to a ULSD unit, gives rise to a surpris-<br />
ingly difficult feedstock to process. These products are then<br />
blended in with other heavier materials as a diluting or cutting stock<br />
and sent downstream as synthetic crude.
Light SR Gas Oil LCO EB Diesel Synthetic Diesel FB Diesel<br />
Sulfur, wt% 1.11 0.17 0.017 0.07 0.006<br />
Nitrogen, wppm 138 203 135 261 71<br />
API Gravity 32.37 24.09 31.24 31.96 31.42<br />
Aromatics, vol%<br />
Total 24.18 64.61 41.26 36.58 44.83<br />
mono 14.47 31.97 36.55 32.52 41.95<br />
poly 9.71 32.64 4.71 4.06 2.88<br />
Dist., D2887, ˚F/˚C<br />
0.5 286/141 219/104 303/151 206 / 97 285 / 141<br />
10 492/256 367/186 407/208 349 / 176 379 / 193<br />
50 601/316 470/243 589/309 517 / 269 515 / 268<br />
90 731/388 578/303 707/375 662 / 350 652 / 344<br />
99.5 799/426 644/340 759/404 738 / 392 775 / 413<br />
Sulfur Distribution<br />
Thiophenes 8 0 0 0 0<br />
Benzothiophenes (BT) 2 69 0 0 0<br />
Substituted BT’s 2793 1083 0 26 0<br />
DiBenzothiophenes (DBT) 222 <strong>110</strong> 0 7 0<br />
Substituted DBT’s 3453 436 72 266 15<br />
4,6 DM-DBT 199 0 78 29 43<br />
C 3-DBT 4410 0 17 370 24<br />
Table 4: Diesel Feedstock Analysis<br />
Likewise, the use of diesel range products from an H-Oil ® , LC-FIN-<br />
ING unit or fixed bed residuum desulfurizer can also have a signifi-<br />
cant impact on downstream diesel catalyst activity for similar<br />
reasons. The general properties of these types of diesel feeds often<br />
indicate that they may be relatively easy to hydrotreat due to their<br />
low sulfur content and higher API gravity, which is often similar in ap-<br />
pearance to straight run (SR) materials. Table 4 lists the properties<br />
for several of these diesel feeds including the diesel product frac-<br />
tions from an ebullating bed residuum (EB) unit, a fixed bed<br />
residuum (FB) unit and a diesel fraction from Canadian synthetic<br />
crude.<br />
ADVANCED REFINING TECHNOLOGIES CATALAGRAM ® 19
Figure 26 shows the activity difference between a SR and a blended<br />
SR/Synthetic diesel. Note that at higher product sulfur, the two feed-<br />
stocks respond similarly to each other. As the application becomes<br />
more demanding, the required reactor temperature increases dramatically<br />
for the synthetic diesel feed compared to the SR feed.<br />
The blended feed requires more than 25°F (14°C) higher temperature<br />
relative to the SR to achieve ULSD sulfur levels.<br />
It is reasonable to expect that the upstream hydroprocessing of the<br />
synthetic diesel material results in a feed that behaves similarly to<br />
other previously hydrotreated feedstocks like those from the EB and<br />
FB residuum applications.<br />
Advanced Refining Technologies can work closely with the refining<br />
technical staff to help plan for processing opportunity feeds such as<br />
those discussed above. One of the keys is being aware of the potential<br />
impacts processing certain feeds will have on unit performance.<br />
Feeds which have been previously processed present unique<br />
challenges and ART ® is well positioned with its experience at providing<br />
customized catalyst systems for ULSD applications. Opportunity<br />
feeds provide yet another objective to consider when designing the<br />
appropriate catalyst system to maximize unit.<br />
ART featured in August issue of<br />
Hydrocarbon Engineering<br />
The August 2011 issue featured ART on its cover and in the lead ar-<br />
ticle “Yield of Dreams” by ART’s Brian Watkins and Chuck Olsen and<br />
<strong>Grace</strong>’s David Hunt.<br />
The collaboration discusses the key relationships between FCC pre-<br />
treat and FCC unit operations and their corresponding catalyst sys-<br />
tems in maximizing the distillate pool. Both processes must be<br />
reoptimized as the refiner moves from gasoline to distillate produc-<br />
tion to ensure maximum profitability.<br />
The article emphasizes the need for refiners to follow an integrated<br />
approach to managing the catalysts and operation of the FCC pretreater<br />
and FCCU.<br />
The complex relationship between the FCC pretreater and the<br />
FCCU underscores the importance of working with a catalyst technology<br />
supplier that has the capability to understand the interplay<br />
between the hydrotreating performance of the FCC pretreater and<br />
the performance, yield structure and product sulfur distributions of<br />
the FCCU.<br />
For copies of the article, contact betsy.mettee@grace.com<br />
20 SPECIAL EDITION ISSUE No. <strong>110</strong> / 2011<br />
HYDROCARBONENGINEERING HYDROCARBOON ENGINEERING<br />
AUGUST2011 AU G GUST2011<br />
www.energyglobal.com<br />
www. energygloba .comm<br />
Required Temperature Increase,˚F<br />
90<br />
80<br />
70<br />
60<br />
50<br />
40<br />
30<br />
20<br />
10<br />
0<br />
0<br />
0 20 40 60 80 100 120<br />
Product Sulfur, ppm<br />
Volume 16 Number 8 - August 2011<br />
SR Synthetic<br />
Figure 26: Activity Comparison of the SR and<br />
Synthetic Diesel Blend<br />
TThe<br />
he<br />
W WWorld’s<br />
orld’<br />
s LLeading<br />
eadin<br />
g SSupplier<br />
upplier<br />
oof<br />
f HHydroprocessing<br />
y ydroprocessing<br />
CCatalysts<br />
atalysts<br />
DDistillate<br />
istillat<br />
e HHydrotreating<br />
ydrotreating<br />
●<br />
FFixed<br />
ixed<br />
B BBed<br />
ed<br />
RResid<br />
esid<br />
HHydrotreating<br />
ydrotreatin<br />
g ●<br />
EEbullating<br />
bullating<br />
BBed<br />
ed<br />
RResid<br />
esi<br />
d HHydrocracking<br />
ydrocracking<br />
LLet<br />
et<br />
AART<br />
R RT<br />
bbe<br />
e a<br />
ppart<br />
art<br />
oof<br />
f yyour<br />
our<br />
ssolution.<br />
olution.<br />
FFor<br />
or<br />
mmore<br />
ore<br />
iinformation<br />
nformati<br />
o on<br />
oon<br />
n oour<br />
ur<br />
pproducts,<br />
roducts,<br />
pplease<br />
lease<br />
ccontact<br />
ontact<br />
uus.<br />
s.<br />
AAdvanced<br />
d v vanced<br />
RRefining<br />
efining<br />
TTechnologies<br />
ech<br />
n nologies<br />
●<br />
77500<br />
500<br />
G<strong>Grace</strong><br />
race<br />
D DDrive<br />
rive<br />
●<br />
CColumbia,<br />
olumbia,<br />
MMD<br />
D 2 221044<br />
1044<br />
UUSA<br />
SA<br />
V<br />
+ +1.410.531.4000 1.<br />
410.<br />
531.<br />
4000<br />
F<br />
+ +1.410.531.8246 1.<br />
410.<br />
531.<br />
82<br />
4 46<br />
W<br />
aartcatalysts.com<br />
rtcatalysts.<br />
com<br />
50<br />
45<br />
40<br />
35<br />
30<br />
25<br />
20<br />
15<br />
10<br />
5<br />
Required Temperature Increase,˚C
-· < '<br />
ART's 420DX Catalyst<br />
Leading the Pack in ULSD Processing<br />
Confirmed by independent lab testing<br />
According to independent lab testing*, ART's 420DX ranks highest in activity among six leading<br />
ULSD catalysts. 420DX is the newest CoMo member of ART's DX® Catalyst Platform. Depending<br />
on operating conditions and objectives, 420DX can be applied in either standalone service or stacked<br />
with ART's NDXi, the high-performance NiMo analog, to achieve the additional benefits of ART's<br />
SmART Catalyst System® technology.<br />
Our technical experts can discuss your specific ULSD operation and tailor a SmART Catalyst System<br />
to meet your deep treating requirements. Contact us to learn more about how to extend your<br />
cycle, improve upset recovery and optimize H 2 consumption in your ULSD unit with ART® catalyst<br />
technology.<br />
Advanced Refining Technologies, LLC<br />
7500 <strong>Grace</strong> Drive<br />
Columbia, MD 21044 USA<br />
v +1.410.531.4000<br />
F +1.410.531.4540<br />
W artcatalysts.com<br />
*Independent lab results available upon request. Advanced Refining Technologies•
© 2011 Advanced Refining Technologies, LLC<br />
SmART Catalyst System ® is a registered trademark in the United States and/or other countries, of W. R. <strong>Grace</strong> & Co.-Conn.<br />
ApART ®, StART ® and DX ® are registered trademarks of Advanced Refining Technologies, LLC.<br />
ART ® and ADVANCED REFINING TECHNOLOGIES ® are trademarks, registered in the United States and/or other countries,<br />
of Advanced Refining Technologies, LLC.<br />
This trademark list has been compiled using available published information as of the publication date of this brochure and may not<br />
accurately reflect current trademark ownership.<br />
artinfo@grace.com<br />
www.e-catalysts.com<br />
The information presented herein is derived from our testing and experience. It is offered, free of charge, for your consideration,investigation and<br />
verification. Since operating conditions vary significantly, and since they are not under our control, we disclaim any and all warranties on the results<br />
which might be obtained from the use of our products. You should make no assumption that all safety or environmental protection measures are<br />
indicated or that other measures may not be required.