DAVISON

®
CATALAGRAM
NUMBER 87, 2000

In this issue...
Value-Added
FCC’s and Additives
 For catalytic solutions to your
processing challenges worldwide...

Davison Catalysts

FCC Additives

Fluid Cracking Catalysts

Hydroprocessing Catalysts

Automotive and Specialty Catalysts

Davison Catalysts Davison Catalysts Davison Catalysts

W.R. Grace & Co.-Conn. Grace GmbH W.R. Grace Singapore
7500 Grace Drive In der Hollerhecke 1 501 Orchard Road
Columbia, MD 21044 67547 Worms, Germany Singapore 238880
Tel.: (1) 410-531-4000 Tel.: (49) 62 41 4 03-0 Tel.: (65) 737-5488
Fax: (1) 410-531-4101 Fax: (49) 62 41 40 34 55 Fax: (65) 737-5826
In this Issue...

The Effects of Iron Poisoning on FCC Catalysts

by George Yaluris, Ph.D. ..................................................................................................page 3
In this article we provide an update on our research efforts into the deposition of contami-
nant iron on FCC catalyst. A mechanism is described that explains the often observed
changes in catalyst properties, as well as the deterioration of catalyst performance. Based
upon this mechanism, recommendations for dealing with iron problems are presented.

Using FCC Catalyst Coke Selectivity to Improve Unit Profitability

by Thomas F. Petti, Ph.D ................................................................................................page 11
This paper provides an overview of the economic benefits that can be realized by employing a
coke selective FCC catalyst. Application to coke burn limited units, as well as more general
cases, are discussed. A commercial example is also provided that illustrates that the changes
described can be realized in an actual FCC unit.

New Personnel ..................................................................................................................page 18

Calculating the Value of FCC Light Olefin Production

by John T. Haley ..............................................................................................................page 19
This article describes the profitability impact of producing light olefins (propylene) in an FCC
unit. The concept of gasoline conversion value (increased product value obtained by cracking
gasoline to light olefins) is presented. A historic perspective of this value and some market
trends are also described. The economic impact of using light olefin additives in an FCC unit
is also illustrated.

Update: Value-Added Gasoline Sulfur Reduction Strategies

by Jeffrey W. Balko ..........................................................................................................page 25
This section provides an update on our gasoline sulfur reduction technologies. We focused
this discussion on the various ways that a refiner could use these technologies as part of an
economical stategy for producing clean fuels. A number of Value Models are desribed, illus-
trating scenarios under which the use of gasoline sulfur reduction technologies make econom-
ic sense.

A Message from the Editor...

Davison Catalysts recognizes that the measure of our success is determined by the bottom
line of our customers. Therefore, we have dedicated this issue of the Catalagram to value.
We continue to actively pursue improvements in our technologies to help our customers
achieve their performance goals. With the issues of producing clean fuels and petrochemical
feedstocks, as well as environmental concerns complicating the refining landscape, today’s
profitability of a refinery goes well beyond the production of gasoline. We hope that this issue
encourages further discussion on the various ways that Davison Catalysts can help by offer-
ing flexible and profitable solutions to today’s refining concerns.

Tom Petti, Editor

3
2
 The Effects of Iron Poisoning
on FCC Catalysts
By
George Yaluris, Ph.D.
Senior Research Engineer
Davison Catalysts

Poisoning of catalyst with contaminant iron Figure 1, in units that have reported iron related
is not new, but until recently its effects on catalyst problems, equilibrium catalyst ABD can decrease by
performance were not clear. Iron related problems in as much as 0.15 cc/g. The exact magnitude of the
FCC units were largely unreported and unlike Ni or decrease depends on the iron level on the equilibrium
V, little work had been done to understand the effects catalyst, the unit type and operation, and the cata-
of iron on FCC catalyst. Today, because of the lyst used. However, some units1 have been able to
decreasing supply of low iron feeds and the need by operate with high amounts of iron on equilibrium
many units to process heavy local feeds, the effects of catalyst without any serious catalyst performance
iron on the unit operation cannot be ignored. problems and without a decrease in equilibrium cat-
Increasing contaminant iron on equilibrium alyst ABD. In general, pore volume remains unaf-
catalyst can have serious adverse effects including fected or decreases less than 10%. Furthermore, MA
loss of activity and bottoms cracking, an increase of activity and zeolite surface area are rarely affected
CRC in partial burn operations, and increased SOx by iron, and many times an apparent correlation
emissions. The chemical make-up of the FCC cata- with iron levels on equilibrium catalyst is due to the
lyst can be a key factor influencing the ability of a simultaneous increase of other contaminant metals,
unit to handle rising levels of iron contamination. like Na and V.
The effect of iron on equilibrium catalyst There are two types of iron in the FCCU.
properties is not very pronounced. As we show in One type consists primarily of particles of inorganic
Figure 1
ECAT ABD Can Decrease with Increasing Iron Contamination
In Units where Iron Causes Catalyst Performance Problems

Unit A
0.95

0.93 ◆
◆ ◆◆◆ ◆ ◆
0.91 ◆◆◆ ◆ ◆ ◆
◆ ◆
0.89 ◆
◆ ◆◆
0.87 ◆ ◆ ◆ ◆ ◆
◆ ◆ ◆ ◆ ◆ ◆ ◆
ABD, g/cc

0.85 ◆ ◆ ◆◆
◆ ◆ ◆ ◆ ◆ ◆
0.83 ◆◆◆ ◆ ◆ ◆ ◆ ◆ ◆ ◆
◆ ◆◆ ◆ ◆
0.81 ◆◆◆ ◆ ◆ ◆
◆◆ ◆ ◆
0.79 ◆ ◆ ◆

0.77

0.75
0.65 0.70 0.75 0.80 0.85 0.90 0.95
Fe, wt%

3
iron from pipes, storage tanks and other hardware. Deposition of Iron on Equilibrium
This type is usually called "tramp" iron and is gener-
ally benign in FCCU operation. The other type is pri- Catalyst
marily organic iron coming from the feed or from
hardware corrosion by naphthenic acids and other To better understand the mechanism by
corrosive feed components. This type of iron is fine- which the FCC catalyst is affected by iron poisoning,
ly dispersed and has deleterious effects on unit per- we studied by Electron Probe Micro-Analysis
formance. Thus, we have focused our research on the (EPMA) the deposition of iron both on various equi-
latter type of iron and, in this article, we present librium catalysts and on lab deactivated catalysts.
some of the highlights of our results to date and pro- In EPMA elemental maps, areas with high concen-
pose a strategy to address this FCCU catalyst per- tration of the element probed show as red or yellow,
formance issue. while low concentration areas are purple or blue.

Figure 2
Increasing the Amount of Iron Deposited on ECAT
Does Not Result in Iron Penetrating Deeper into the Particle

Fe on ECAT increases by 240%

Both ECATs from the same unit

Fe scans shown

The thickness of Fe and Ca rings is

in the order of 1-3 µm

CaO Ring Detail Fe Ring Detail -- 5µm

Figure 3
Ni Deposition on ECAT from Two Units Representing Two Extreme Cases of
Ni Concentrating on the Outward Areas of ECAT Particles
and Distributing Throughout the Particles

4
 We show one representative example of the Iron Concentration on
distribution of iron on equilibrium catalyst in
Figure 2. The EPMA data unequivocally show that Equilibrium Catalysts Particles
iron always deposits on the external surface of equi- Surface
librium catalyst particles creating rings of high iron
concentration around the particles. We also investi- Since iron deposits on the external surface of
gated if iron would penetrate further inside the FCC the equilibrium catalyst particle, it is reasonable to
particle as iron levels on the equilibrium catalyst assume that the surface concentration of iron is a
increase (chromatographic effect). Analysis of the critical parameter in determining the effects of iron
data, like the example shown in Figure 2, suggests on the catalyst. Using X-ray Photoelectron
that increasing the iron content of the equilibrium Spectroscopy (XPS), we measured the surface compo-
catalyst increases the concentration of iron on the sition of various equilibrium catalysts, mostly from
equilibrium catalyst surface, but not the depth of units reporting problems with iron poisoning of their
iron penetration inside the particle. It appears that catalyst, but also from a few units which have not
iron does not penetrate inside the equilibrium cata- reported any iron related problems, and calculated
lyst particles much deeper than 1-3 µm. surface enrichment ratios for many of the elements
These results show that iron deposits on the
first site of the particle it encounters and continues
to do so during the iron deposition process. Since Table I
iron penetration does not appear to be a strong func-
tion of the catalyst used, we conclude that any pore Surface Enrichment Ratios
structure differences among the catalysts studied do for Elements Present in
not have a major impact on the deposition of iron, Equilibrium Catalyst
and therefore, do not significantly impact catalyst
deactivation by iron. The ratios are calculated by dividing the concentration of the
The characteristic deposition profile of iron element on the surface by its concentration in the bulk.
on equilibrium catalyst particles can help explain the
rather limited effects of iron poisoning on equilibri- Surface Surface
um catalyst properties. Since iron contamination Weight Enrichment
only affects a ring 1-3 µm deep on the surface of a (%) Ratio
particle on average 60 – 70 µm large, most of the par- Na up to 1% 1-2
ticle remains unaffected. Thus, the surface area,
pore volume, and even microactivity of iron poisoned Al 0.4
equilibrium catalysts should not be severely affected
Si 0.8 - 1.2
by iron.
Of the other metal contaminants, calcium is Ca up to 1.2% 2 - 30*
the only one that has the same deposition profile as
iron, forming rings of high Ca concentration on the Fe up to 26% 10 - 18
external particle surface (Figure 2). The presence of
* Ratios of species in small amounts can change
Ca in high concentrations at the same high iron con- significantly from ECAT to ECAT depending on the amount of
centration areas of the particle surface suggests that them present on the fresh catalyst
Ca may be involved in the iron deposition related
phenomena that lead to catalyst deactivation.
Vanadium and sodium deposit throughout the cata- present in equilibrium catalyst. We have compiled
lyst particles. Nickel can concentrate on the external these data in Table I.
surface of equilibrium catalyst particles. However, In general, the surface enrichment ratios in
as we show in Figure 3, depending on the unit, the Table 1 show that there is little difference between
amount of nickel deposited, the residence time of the the surface and bulk concentration for Na and Si.
equilibrium catalyst in the unit, and any catalyst This result is not unexpected considering the well
nickel passivation technology employed, it can even- known mobility of both elements oxides. On the
tually penetrate throughout the particle. Thus, other hand, Al is not very mobile and in general it is
nickel deposition does not cause pore blocking that covered by the depositing elements, primarily iron,
prevents heavy feed molecules from reaching the resulting in a surface enrichment ratio well below
particle interior. one. As expected from the EPMA results, Fe is great-
ly enriched on the particle external surface.

5
 Figure 4
SEM Pictures of Equilibrium Catalyst Particles from Two Different Units

High SiO2 catalyst formulation “High accessibility” competitive technology

Figure 5
Optical Microscopy of Lab Deactivated Catalyst with No Added Iron
and Three Equilibrium Catalyst Samples of the Same Catalyst from the
Same Unit Taken at Different Times as Iron Levels Change

Fresh Catalyst (steamed) 90 rel.% contam. 160 rel.% contam. 500 rel.% contam.

Optical Microscope
Increasing Fe Content

The three ECATs are from the same unit collected at different times.

Since Si and Na are not covered by the Morphology and Surface Texture of
depositing Fe, we conclude that in the iron rings on
the equilibrium catalyst particles external surface
Iron Poisoned Equilibrium Catalysts
silica and soda are well mixed with iron. Alumina
does not appear to mix with iron and the other We studied the morphology and surface tex-
oxides. Thus, while phases that contain silica, iron ture of iron poisoned equilibrium catalysts using
and calcium oxides, and soda appear to form easily Scanning Electron Microscopy (SEM) and optical
on the surface of equilibrium catalyst particles, sim- microscopy. Iron has a distinct effect on equilibrium
ilar phases containing alumina and iron oxides seem catalyst morphology and texture. Although there are
to be much more difficult to form. exceptions, the SEM data in Figure 4 clearly show
6
the formation of characteristic nodules on the surface catalysts from the same unit taken at different times
of equilibrium catalyst particles. These nodules as the equilibrium catalyst iron levels changed. The
appear to be present on the surface of equilibrium data show that as the catalyst ages in the unit with
catalyst particles from different units and different iron its surface acquires a "glassy" cover, which
catalyst technologies, including equilibrium cata- under optical microscopy gives the equilibrium cata-
lysts from units using a purportedly “high accessibil- lyst particles a sharper definition and glossy appear-
ity” competitive technology. ance. Increasing levels of iron on the equilibrium
Optical microscopy shows that the texture of catalyst particles further enhances the observed
iron contaminated equilibrium catalyst changes in glassy layer and changes the particles color closer to
another important way. In Figure 5, we show one that of ferric oxide.
example of optical microscopy pictures of equilibrium

Table II
Initial Melting Temperatures for Phase Systems Containing SiO2, Al2O3,
FeO/Fe2O3, Na2O and/or CaO at Ranges of Concentrations which Include
Those That Can Be Found on the Surface of FCC Equilibrium Catalysts

System SiO2 Al2O3 FeO/Fe2O3 Na2O CaO First Liquid

Temperature
(%) (%) (%) (%) (%) (˚C) (˚F)

SiO2 100 1713 3115

Al2O3 100 2050 3722

SiO2Na2O2 Balance 0*-25 780 1436

SiO2-CaO3 Balance 0-36 1436 2617

Al2O3-Na2O4 Balance 0*-5.5 2000 3632

5.5-9.2 1580 2876

>9.2 1410 2570

SiO2-FeO5 Balance 0*-62 1180 2156

SiO2-FeO•Fe2O36 Balance 0*-85 1455 2651

Al2O3-FeO7 Balance 0*-35 1750 3182

SiO2-FeO-Na2O8 Balance 0*-57* 25-0*,† <500 932

Na2O-Al2O3-SiO2- 71.3 4 6.2 18.4 771 1420

Fe2O39 58.5 9 17.3 15.1 943 1729

59.3 16.4 9 15.3 998 1828

47.9 24.1 9.7 18.4 1248 2278

* 0% is not included in the composition range with the indicated first liquid temperature.

† The combination of high FeO and high Na2O concentrations leads to phases that have slightly higher melting
temperatures (~ 667 ˚C/1233 ˚F).

7
The Thermodynamics of Iron- and/or Ca. It takes more than 7% Na to decrease the
melting temperature of alumina to 1410 ˚C (2570 ˚F).
Poisoned FCC Catalyst Surfaces Even a combination of alumina with silica and iron
along with unrealistically high amounts of Na does
Of the two major components of FCC cata- not bring the melting point of phases richer in Al2O3
lysts, silica and alumina, the latter has the higher below 900 ˚C (1652 ˚F). Furthermore, as the alumi-
melting point. However, when Na, Ca and Fe are na concentration increases, the initial melting tem-
present, the melting points of both silica and alumi- perature of the phases formed with silica, iron and
na decrease substantially. In Table II we show a sodium oxides also increases (Table II).
compilation of the temperatures at which the first
liquid appears (initial melting temperatures) for
phases silica or alumina can form with Na, Ca and
The Mechanism of FCC Catalyst
Fe oxides. Both Na and Ca dramatically reduce the Poisoning by Iron
initial melting temperature of silica, although Na is
a much more effective fluxing agent for SiO2 than Ca. Every FCC catalyst particle is constituted of
The effect of FeO on the initial melting point of phas- smaller particles of zeolite, matrix, clay, and the
es with SiO2 is not as severe as that of Na, but it is binder that holds everything together. As a result
more severe than that of magnetite or Fe2O3. the exterior surface of the particle is not homoge-
However, when Na and Fe are combined with SiO2 at neous. Rather it has areas of zeolite, clay and matrix
(typically added alumina), mostly coated with the
concentrations similar to the ones we measured on
binder used. Spot analysis of the binder areas by
the surface of equilibrium catalysts by XPS, the ini-
TEM/EDS shows that Si-Sol binder areas of the par-
tial melting point decreases below 500 ˚C (932 ˚F).
ticle are essentially silica containing only traces of
This temperature is lower than the temperature of
alumina. Similarly, Al-Sol binder and alumina
virtually every riser.
matrix areas are essentially alumina containing only
Contrary to silica, the melting temperature
traces of silica.
of alumina is not affected as severely by Fe, Na

Figure 6
Schematic Representation of Si-Sol Catalyst Deactivation by Iron Poisoning Showing
Iron Deposition, the Formation of Low Melting Temperature
Phases, and Pore Closing in a Small Section of the Surface
of the Spherical Catalyst Particle
Low melting point
phase formed Pore closing and
Fe deposition (glassy layer) nodule formation

Si-Sol Catalyst aging process with Fe

SiO2 and FeO/Fe2O3 mix together with CaO and Na2O

forming low melting temperature phases

Zeolite Clay Si-Sol

8
 mation and pore closing may occur in these areas.
Si-Sol Catalysts However, in Al-Sol catalysts clay and zeolite particles
are not the main contributors to bottoms cracking.
We show a schematic diagram of the process The alumina binder and the active matrix compo-
of catalyst deactivation for a Si-Sol catalyst by iron nents are where much of the bottoms cracking occur.
deposition in Figure 6. As the catalyst ages, Fe, Ca, The data show that alumina on the external
Na and other contaminant metals deposit on it. On surface of catalyst particles can be covered by the
Si-Sol catalyst the depositing oxides mix with the deposited iron contamination and does not mix with
binder silica forming phases that have initial melting it to form low melting temperature phases. Even if
temperatures as low as the operation temperature of alumina could mix with the depositing iron, the ther-
the riser and/or the regenerator. The oxidation state modynamic analysis shows that alumina phases con-
of the iron on the catalyst surface changes as the cat- taining FeO, Na2O and/or CaO require high temper-
alyst circulates in the unit, but it probably spends atures to melt. These temperatures are much high-
most of the time in a reduced state (Fe2+), further er than the riser and regenerator temperatures.
decreasing the melting point of iron containing phas- Therefore, the surface pores in binder and active
es. The formation of these low melting temperature matrix areas remain essentially open, allowing the
phases facilitates the fluxing of silica in the binder, feed molecules to transport inside the catalytic parti-
thus filling and closing the pores and giving the sur- cle for cracking. Hence, as refinery experience has
face the "glassy" texture we have observed. Even if demonstrated1, these catalysts are resistant to iron
outright melting does not occur, accelerated sinter- poisoning offering good activity maintenance, good
ing, due to the decrease of the material melting tem- bottoms cracking, and no change in equilibrium cat-
perature will have a similar effect. As iron accumu- alyst ABD with iron, even in the face of high levels of
lates on the surface of Si-Sol catalysts, these low iron contamination.
melting temperature phases cover a larger and larg-
er area of the particle external surface, thus closing
more and more pores. Strategies for Addressing Catalyst
The collapse of the surface pore structure Iron Poisoning in the FCCU
where the iron deposits results in the surface
rescinding as the melting and sintering decreases Having the right pore structure that allows
the volume of the affected material. Continuous col- feed molecules to transport inside the catalyst for
lisions between the fluidized particles further shape cracking is critical in any catalyst that delivers high
the softened particle surface. Particles may also activity and bottoms cracking selectivity, particularly
stick together and then break apart forming particles for catalysts slated for resid applications. For the
of irregular shape. All these processes either create FCCU operator the key question is how to maintain
or facilitate the creation of the nodules and valleys that pore structure in the face of iron contamination,
on the catalyst surface that we observed by SEM, and prevent pore closing that results in catalyst per-
and result in less efficient packing of equilibrium cat- formance deterioration. Based on the data we have col-
alyst particles, causing the ABD of iron poisoned lected thus far, we recommend the following strategy.
equilibrium catalysts to decrease.
The mechanism we described here will affect
an area around the particle no deeper than 1-3 µm,
thus leaving most of the particle and its properties
unaffected. However, in the FCC unit it is critical 1. Establish that any catalyst performance
that the feed molecules diffuse inside the particle deterioration observed is indeed due to
and the products diffuse out. The closing of the pores rising iron levels on the equilibrium
by the processes set in motion by iron deposition on catalyst. Sometimes iron may rise after a
the catalytic particle restricts the diffusion of the feed change accompanying loss of activity
feed molecules into particles, thus decreasing activi- and bottoms cracking. Before concluding the
ty. Since the large feed molecules are the ones most increased iron levels have poisoned the cata-
affected by pore closing, bottoms cracking is the cat- lyst, the refiner should establish that the
alyst property more severely affected. performance deterioration is not due to a
change in feed crackability and increases of
Al-Sol Catalysts other metals (e.g., Na, V, Ni) deleterious to
catalyst performance. SEM, EPMA and opti-
The formation of high iron low melting tem- cal microscopy analysis of the equilibrium
perature phases may also occur at high silica areas of catalyst can be used to look for the surface
the particle surface of Al-Sol catalysts. Such areas composition and texture characteristic of
may be exposed zeolite or clay particles. Nodules for- iron poisoned calalyst.

9
2. Try to reduce the iron coming into the 5. Use an appropriately designed iron
unit. Measures that can be employed to resistant Al-Sol catalyst. When the unit
reduce iron coming with the feed is to stop does not have the flexibility to implement
using high iron feeds and/or to buy low iron other solutions, or other solutions fail, Al-Sol
feed to blend with the high iron one. It has catalysts have been proven to provide excel-
been suggested that acids in the feed (e.g., lent resistance to iron contamination1, and
naphthenic acids) can corrode hardware, maintain activity and bottoms cracking even
increasing the iron content of the feed. at iron levels on equilibrium catalyst which
Reducing the acid content, or purchasing low are the highest in the industry.
acid content feeds can reduce hardware cor-
rosion, and thus decrease the amount of iron
in the feed. References

3. Reduce the Na and Ca content of the 1. "Impact of Fe on FCC Resid Processing" in FCC •
feed. Na and Ca act as fluxing agents Additives • Hydroprocessing • Refining
severely aggravating the catalyst poisoning Technology Grace Davison Conference, Singapore,
effect of iron. It is therefore critical that August 16-18, 2000.
2. F. C. Kracek, J. Phys. Chem., 34, 158 (1930)
either low Na and Ca feeds are used, or that
3. B. Phillips and A. Muan, J. Am. Ceram. Soc., 39
the amounts of these metals in the feed are [4], 5 (1915).
reduced by using desalting or other suitable 4. M. Rolin and P. H. Thanh, Rev. Hautes Temp.
processes. Refractaires, 2 [2], 178 (1965).
5. N. L. Bowen and J. F. Schairer, Am. J. Sci., 5th
4. Minimize the riser and regenerator tem- Ser., 24, 200 (1932)
perature, if possible. High temperatures 6. B. Phillips and A. Muan, J. Am. Ceram. Soc., 12
accelerate catalyst deactivation by iron. [9], 415 (1959).
However, lowering the temperature of the 7. I. A. Novokhatskii, B. F. Belov, A. V. Gorokh, and
A. A. Savinskaya, Russ. J. Phys. Chem., 39 [11],
riser or the regenerator can have a signifi-
1498 (1965).
cant impact on the unit operation, and may 8. P. T. Carter and M. Ibrahim, J. Soc. Glass
not be an option for most units. Technol., 36, 156 (1952).
9. D. K. Bailey and J. F. Schairer, J. Petrol., 7, Pt. 1,
125&126 (1966).

Look for interactive versions

of our FCC Value Models to
be added to our website in
the near future.

10
 Using FCC Catalyst
Coke Selectivity
to Improve Unit Profitability

By
Thomas F. Petti, Ph.D.
Marketing Manager, Catalyst Technologies
Davison Catalysts

Coke selectivity is an attribute of FCC cata- of all FCC units. Units that are usually considered
lysts that is frequently discussed. It is normally a coke burn limited are:
"motherhood and apple pie" topic and generally a -operating at air blower capacity
good thing, but it is easy to lose sight of exactly why -constrained by permits related to the
this is so and what it can mean to refiners. We at amount of coke they can burn or
Davison Catalysts have made coke selectivity per- -operating at maximum regenerator temper-
formance the hallmark of the entire FCC catalyst ature and at maximum catalyst circulation.
product line for over a half-century and it continues These units are in a position to directly profit from
to play a prominent role in our new product develop- improvements in coke selectivity.
ment activities. If we consider the definition of coke selectivi-
Why is coke selectivity so important? Let’s
take a closer look at how catalyst coke selectivity Figure 1
gets translated into FCC unit product value through North American FCC Units by Constraint
potentially higher conversion and/or higher charge
rates. We will start by looking specifically at coke 8%
2%
burn limited units and then move to more general
cases where incremental product value can be gener-
ated by producing less coke.
47%

Coke Limited Units

One of the reasons that coke selectivity is
such an important catalyst attribute is that a very 43%
high percentage of FCC units are constrained by
their ability to burn additional coke (Figure 1). In
North America, coke burn limitations affect over half Other Limits

Air Limit
Coke Bum Permit
Regn Temp Limit

11
ty, we can easily see the value that improvements in ratio would lead to higher unit conversion. These
coke selectivity can offer. Figure 2 shows two cata- effects are summarized in Figure 2, noting that
lysts (A and B) that demonstrate different levels of most units run to nearly constant coke yield to main-
coke selectivity performance. Catalyst B produces tain heat balance at a constant set of operating con-
about 15% lower coke than catalyst A at a given con- ditions. The better coke selectivity of Catalyst B
version level. Looking at the impact of a move to results in higher overall conversion, about 4 lv% in
Catalyst B, the unit would respond by producing this example. Note that this example assumes that
lower coke per pass in the riser. This change would the unit is not constrained by catalyst circulation
cool down the regenerator, which would in turn lead (about 25% higher circulation would be required),
to higher catalyst circulation rate to maintain heat and also that the unit is capable of handling the addi-
balance (the slide valve would open to maintain riser tional product volume.
top temperature). What would this example mean in terms of
Assuming that both catalysts are the same additional unit profitability? If we assume that this
activity (lower plot in Figure 2), the higher cat-to-oil FCC unit is running 50,000 B/D with typical Gulf

Figure 2
Comparison of Catalyst Coke Selectivity and the Impact on FCC Unit Conversion

7
Catalyst A
6

5
Coke Yield, wt%

Catalyst B
4

0
68 70 72 74 76 78 80 82
Conversion, lv%
12
Catalyst A
and B
Cat to Oil Ratio, w/w

10

6
▲

Increased
4
conversion

0
68 70 72 74 76 78 80 82
Conversion, lv%

12
Coast economics, we can estimate the value of the We can consider the same example, but look
improved coke selectivity. Table I summarizes these at the effect of improved coke selectivity when feed
results. Here we see that this unit generates $30.55 of rate is increased, rather than running to higher con-
product value for each barrel of feed. Since feed is version. In this case, assume that when operating on
worth approximately $25.00/BBL, the upgrade value is Catalyst B, the conversion is dropped back down to
$5.55/BBL of feed. When the same scenario is consid- the base 74 lv% level. We have assumed that feed
ered with the more coke selective Catalyst B, we can preheat is increased, but a decrease in riser top tem-
estimate that the upgrade value will increase in pro- perature will give similar results. In this case,
portion to the yields of the individual products. This is approximately 15% higher feed rate can be run to the
worth about $0.07/Conversion-BBL (incremental lv% unit without consuming more air than the base case.
conversion times the feed rate). Here we see that the Table II summarizes the economics of this case.
4 lv% higher conversion translates into $5.85/BBL Here, we see an even higher incremental value, due
feed, or an increase of about $0.30/BBL. For this unit, to the upgrade value and the assumption that we
this would offer $15,000/day in incremental unit prof- have not bumped into any other constraints. In each
its (i.e., an extra $3,000 in profit for each ton of case we see the significant value that the more coke
Catalyst B). Notice that this incremental product selective catalyst is delivering.
value more than covers the entire cost of any FCC cat-
alyst on the market today.

Table I
Summary of Economic Benefit of Improved FCC Catalyst Coke Selectivity

Catalyst A Catalyst B

Feed Rate, BPD 50,000 50,000

Catalyst Adds, TPD 5 5

Conversion, Iv% 74 78
Product Value, $/BBL Feed $30.55 $30.85
Feed Value, $/BBL $25.00 $25.00
Upgrade Value, $/BBL Feed $5.55 $5.85

Incremental Value of Catalyst B, $/BBL - $0.30

Incremental Value of Catalyst B, $/Day - $15,000
Incremental Value of Catalyst B, $/Ton of Catalyst - $3,000

Table II
Summary of Economic Benefit of Improved FCC Catalyst Coke Selectivity

Catalyst A Catalyst B

Feed Rate, BPD 50,000 57,500

Catalyst Adds, TPD 5 5

Conversion, Iv% 74 74
Product Value, $/BBL Feed $30.55 $30.55
Feed Value, $/BBL $25.00 $25.00
Upgrade Value, $/BBL Feed $5.55 $5.55

Incremental Value of Catalyst B, $/BBL - $0.00

Incremental Value of Catalyst B, $/Day - $41,625
Incremental Value of Catalyst B, $/Ton of Catalyst - $8,325

13
Units Not Coke Burn Constrained constraints must not limit the ability to either gen-
erate higher conversion or to run higher feedrates.
Units that are not necessarily constrained by The unit would need to handle more product volume
coke burn may also be able to take advantage of in either case. It may also need to circulate more cat-
improved coke selectivity. The unit heat balance will alyst and tolerate a lower regenerator temperature
need to be maintained, so the required coke must be without losing the ability to burn the catalyst clean.
generated in the unit. Certainly, it is preferable to However, even if these situations do not apply, other
generate less coke per unit conversion with a catalyst changes in catalyst design can allow an FCC unit to
and meet the heat demand with higher conversion, enjoy the benefits of improved coke selectivity.
thereby producing more of the valuable products in Often refiners who operate near their cata-
the process (Table I). Alternately, higher feed rate lyst circulation limit or who run with low regenera-
at lower conversion and coke yield can also be used to tor temperatures believe that they cannot tolerate a
produce the required coke (Table II). However, unit more coke selective catalyst. Their main goal

Figure 3
Comparison of Catalyst Coke Selectivity
and Activity and the Impact on FCC Unit Conversion

7
Catalyst B
6

5
Coke Yield, wt%

Catalyst C
4

0
72 74 76 78 80 82 84 86
Conversion, lv%
14
Catalyst B
12
Cat to Oil Ratio, w/w

10
Catalyst C
8

6
▲

Increased
4
conversion
2

0
72 74 76 78 80 82 84 86
Conversion, lv%

14
becomes to increase the regenerator temperature in A Commercial Example
order to decrease catalyst circulation or improve
regeneration. Although it is true that higher regen- Can we really see shifts of this sort in a com-
erator temperatures may be required to improve mercial FCC unit? Let’s take a look at an example
their operation, poorer coke selectivity will not give where a catalyst change was made, similar to those
them the flexibility to improve the operation (i.e., to that we have been discussing here.
increase conversion or throughput). However, a com- This commercial FCC unit is a UOP side-by-
bination of higher activity catalyst and better coke side design, and typically runs a hydrotreated gas oil
selectivity often can offer significant profitability feedstock. The unit is combustion air and catalyst
opportunities. circulation limited and was seeking to maximize unit
Let’s take a look at how this works. If we conversion. The unit was running a competitive cat-
first consider the impact of changing to a poorer coke alyst technology before switching to a Davison FCC
selectivity catalyst in order to achieve the goal of a technology that offered improved coke selectivity and
warmer regenerator, we can simply use the same two higher activity. A comparison of the operation on the
catalysts we discussed above in Figure 2, but this two catalyst technologies is summarized in Table
time in reverse. The operation is running on III. From the table, we can easily identify the
Catalyst B at 78 lv% conversion, but higher conver- improved coke selectivity. The new catalyst is capa-
sion cannot be achieved with higher riser tempera- ble of achieving +7 lv% higher conversion at the same
ture or lower feed preheat because the unit cannot base coke yield. This is the definition of coke selec-
circulate more catalyst. A switch to Catalyst A tivity. Note that the feed charge rate is nearly con-
might be considered to generate more coke to heat up stant through the comparison periods. Also note that
the regenerator, relieving the circulation constraint. the improved coke selectivity is observed in the equi-
As we see in Figure 2, this move will lower the cat- librium catalysts properties as a decline, i.e.,
to-oil ratio and catalyst circulation rate at constant improvement, in coke factor, which is a measure of
feed, but at the expense of 4 lv% conversion. coke selectivity.
Although it may be possible to recover some of this The activity increase is also evident from the
conversion loss with higher reactor temperature or table. Although the conversion increased from the
lower feed preheat, it is unlikely that we finish bet- base, the unit’s catalyst circulation limit would not
ter than the starting 78 lv% conversion. allow the cat-to-oil increase. In the table, we see that
Now let’s consider another approach. If, the cat-to-oil actually decreases. Higher conversion
instead of moving to the poorer coke selectivity at lower cat-to-oil is the definition of higher catalyst
Catalyst A, the refiner moved to a more active and activity. Notice also that the regenerator tempera-
coke selective Catalyst C, we get the scenario shown ture increased on the unit due to the increase in
in Figure 3. In this case, the improved coke selec- activity. The equilibrium catalyst microactivity also
tivity allows the unit to run to higher conversion shows the activity increase, although the change is
than on the base Catalyst B. Here again we see an smaller than that observed on the unit. Keep in mind
increase of 4 lv% conversion (to 82 lv%) on the more that the equilibrium catalyst is tested on a standard
coke selective Catalyst C, which is about 15% more feed, so activity changes will be directionally rele-
coke selective. Unlike in Figure 2, however, the vant while the magnitude may be feed dependent.
Catalyst C is also more active. By definition this This example commercially demonstrates the effects
means that Catalyst C will yield higher conversion at presented in Figure 3, although the activity
a given cat-to-oil ratio. As a result, the higher con- increase here more than compensates for the
version does not demand a higher cat-to-oil ratio, improved coke selectivity, so cat-to-oil actually
which the unit cannot deliver. The higher activity decreased. This catalyst change has resulted in sig-
Catalyst C is able to yield the 82 lv% conversion at nificantly higher unit conversion, despite the multi-
the base cat-to-oil ratio, demanding no higher cata- ple constraints that were previously limiting this
lyst circulation from the unit. In this case, we see unit. As shown in Table III, the economic impact
that even with a catalyst circulation limit or low has been extremely positive.
regenerator temperature, a properly designed cata-
lyst with improved coke selectivity will deliver sig-
nificant value. Since this case very closely mirrors A Few Comments on Coke
the discussion from the last section, the economic Selectivity
benefits will be similar to those presented in Table
I: an extra $0.30/BBL of feed or $3,000/ton of cata- Throughout this discussion we have been
lyst. talking about coke selectivity as though it were a sin-
gle catalyst property (either you have it or you don’t).

15
Unfortunately, it is not always this straightforward. returned to the regenerator. These four types of coke
In commercial operations the coke yield on the unit are often referred to as:
is impacted by a number of sources of coke. These • Catalytic coke
include the catalyst’s base coke selectivity, the coke • Contaminant coke
generated from contaminant metals, Conradson car- • Feed added coke
bon of the feedstock, and unstripped hydrocarbons • Cat-to-oil coke

Table III
Summary of Economic Benefit of Improved Catalyst Coke
Selectivity and Activity

Competitive Davison Catalyst

Catalyst Improved Coke
Selectivity
and Higher Activity

Equilibrium Catalyst Properties

Microactivity 73 75
Coke Factor 1.2 1.0
Gas Factor 1.7 1.5

Equilibrium Catalyst Ni, ppm Base Base + 200

Equilibrium Catalyst V, ppm Base Base

Total Surface Area, m2/gm 150 190

Zeolite Surface Area, m2/gm 90 150
Matrix Surface Area, m2/gm 70 38

Unit Yields Delta from Base

Dry Gas, lv% Base -0.5%
LPG, lv% Base +3.0%
Gasoline, lv% Base +5.0%
LCO, lv% Base -3.9%
Slurry, lv% Base -3.1%
Coke, wt% Base Base
Conversion, lv% Base +7.0%

Unit Data Delta from Base

Catalyst Additions, T/D Base Base
Feed Rate, B/D Base +700
Feed Gravity, ˚API Base +0.8
Feed Preheat Temperature, ˚F Base +12
Riser Top Temperature, ˚F Base Base
Regen Dense Temperature, ˚F Base +34
Cat-to-Oil, w/w Base -1.0

Economic Impact - West Coast Economics Delta from Base

Incremental Value of Catalyst Change, $BBL - $0.60
Incremental Value of Catalyst Change, $/Day - $46,000
Incremental Value of Catalyst Change, $/Ton of Catalyst - $7,700

16
The catalyst design can play a role in managing all of customers and coke selectivity continues to be an
these types of coke. Certainly in light feed situa- active area of research. Our goal is to continually
tions, the catalytic coke plays the dominant role. develop new products and manufacturing processes
However, in heavier feed applications (e.g., resid that raise the performance bar, allowing our cus-
feeds), the contaminant coke and feed added coke tomers to profit from improved unit performance and
also play a large role. Regardless of the type of appli- operational flexibility.
cation, improvements in coke selectivity can yield the Although we also investigate catalyst objec-
benefits that have been described here. However, tives other than coke selectivity, sacrificing coke
different application types may require different selectivity performance is never considered an
approaches to achieve coke selectivity, including bet- acceptable tradeoff. Each of our research areas is
ter catalytic selectivity, improved tolerance to con- attacked with coke selectivity as a performance
taminant metals, or better catalyst strippability. It objective. Similarly, as different raw materials are
is very important to consider the impact of coke screened, lower cost alternatives are not considered
selectivity in any application, and to design a cata- if coke selectivity performance might be compro-
lyst that is specifically tailored to the service. mised.
Davison Catalysts has commercialized six
new technologies in the year 2000. These products
Davison FCC Catalyst Technologies are each designed for specific target applications,
and each of these new catalyst families also delivers
We started this discussion by stating that the best available coke selectivity of any product in
coke selectivity performance is an absolute require- its class.
ment for all FCC catalysts that Davison offers to our

"Advanced Catalyst Technology for "Superior Stability and Selectivity

Stability and Selectivity" for Resid Processing"

Coke
Selective
Technologies
"Superior Zeolite and Matrix "The Next Generation of Premium
Activity for Maximum Performance Vanadium Tolerant FCC Catalysts"

"Advanced Technologies "Optimum Coke Selectivity

for Low Olefin and Sulfur Fuels" and Metals Tolerance"

17
 Davison Catalysts Introduces
Six New Members
of Its Refining Sales/Service Team
Alfred F. Jordan has been Lori T. Boock is Technical
named to the newly created Sales Representative for
position of National Sales Refining Catalysts located
Manager, Refining Cata- in Houston. Previously
lysts, Columbia. Al, who Senior Research Engineer
joined Grace in 1979, has for Refining Catalysts, Lori
held numerous sales, man- joined Davison in 1992.
ufacturing, financial, and Lori received a Ph.D. in
technical positions in chemical engineering from
Grace’s various product the University of Delaware,
lines including Technical an M.S. in environmental
Sales Representative for engineering from the
Davison’s Silicas product line, Production California Institute of Technology and a B.S.Ch.E.
Coordinator for FCC Catalysts, and Senior Financial from the University of Pennsylvania.
Analyst. Prior to joining the Sales/Service Team, Al
was Curtis Bay Works Manager and Manufacturing
Manager for Grace’s Polyolefin Catalyst product line.
He holds a B.S.Ch.E. from Drexel University and an Raul Barerra, Technical
M.B.A. from Loyola College. Al and Jack Olesen, Sales Manager, is the most
National Sales Manager, Refining Catalysts, recent addition to our Latin
Houston, will report to Kent Davis, Director, North American group. Raul, pre-
American Sales/Service for Refining Catalysts. viously with Brown & Root
Services as a systems engi-
neer, received a B.S. in both
The newest members of chemical and industrial
Davison’s Refinery Cata- engineering from the
lysts’ Sales/Service Team Instituto Technológico de
are Paul D. Wendt and Buenos Aires. A native of
Shahab Parva, both Tech- Argentina, Raul has also
nical Sales Representa- been employed by Shell CAPSA there.
tives located in Houston.
Previously, Paul was locat-
ed at the ExxonMobil refin-
ery in Baytown, Texas, Natalie C. Petti joins
where he held a number of Davison as Technical Sales
technical and operations Representative, located in
positions. Paul holds a Columbia. A B.S.Ch.E.
B.S.Ch.E. and an M.E. in graduate of the University
chemical engineering from of Delaware, Natalie was
Texas A&M University. formerly Process Engineer
Shahab, who has a at the Motiva refinery in
B.S.Ch.E. from West Delaware City, Delaware.
Virginia University, joins Prior to joining Grace, she
Davison from Pennzoil was in process engineering
Quaker State Company, and business development
where he was a process for W. L. Gore and Associates, manufacturers of
engineer. Gore-Tex®.

18
 Calculating the Value
of
FCC Light Olefin Production
By
John T. Haley
Marketing Manager, Light Olefins Technologies
Davison Catalysts

Market Forces well into the future, providing incentives for more
refiners to fill the gap.
With Asian economies back on track, signifi- FCC units currently supply nearly one-third
cant opportunities exist for refiners able to capitalize (32%) of the world’s propylene, and that percentage
on the increasing worldwide demand for propylene. is expected to increase. Of the approximately 31.7
After hitting a ten-year low in the first quarter of million tons of refinery-produced propylene, only
1999, propylene values rallied to a record level this 16.3 million tons, slightly more than half (52%), is
past summer. Although propylene values have cor- consumed for petrochemical production (Figure 1)1.
rected somewhat from this high, they remain above A major component of durable goods, polypropylene
historical ten-year peak values. With demand out- demand continues to improve as a result of improve-
stripping supply, values will likely remain strong ments in the health of most economies.

Figure 1
2000 World Propylene Supply

To Petrochemicals From Refineries

51.2 Million Tons 31.7 Million Tons

Steam Cracker
To Chemical
66%
52% LPG/Fuel
26%
▲
▲

Dimersol/ Polygas
2%

Alkylation
20%
Other From Refinery
2% 32%

19
 Even with the economic recovery in full refinery, or FCC-produced propylene, for the foresee-
swing, per capita polypropylene consumption in most able future.
Asian, African and South American economies
remains significantly lower than the world average
(Figure 2)2. Although polypropylene capacity Gasoline Conversion Value
expansions have outpaced the expected 7% growth in
polypropylene demand, propylene supply will remain Propylene has often been called a by-product
tight. of FCC unit operation since it was not intended as
Compounding the shortage of propylene the primary product. But, with the increased role of
capacity, naphtha cracker feedstock values are at an the refinery in the propylene market, many refiners
all time high, squeezing cracker profits and, in some are making intentional changes in their operations
cases, causing rate cutbacks. As the market was to improve propylene production. Of course, the
heating up, several naphtha crackers were taken improvement in propylene comes at the expense of
down for turnarounds. The net effect was a depletion other product yields and the value of the incremental
of reserves, driving spot prices significantly higher propylene yield must be weighed against the value of
than contract prices, and in some cases, leaving those lost yields.
petrochemical manufacturers short of feedstock.
Although many refineries could not take advantage
Figure 2
Polypropylene Consumption Per Capita

30

25

20 ▲
Difference Between
15 Average World Consumption
and
10 "Developed" World Consumption
▲
5

0
Korea
Taiwan

Mexico
U.S.

Japan

W. Europe

S. America

Argentina

Brazil

SE Asia

China

Africa/ME

E. Europe

India

World

1998 1999 2004

There are several methods to improve FCC

of this market due to location or infrastructure con- propylene yields. The three most common methods
straints, other FCC operators could not take advan- are:
tage of the market because they lacked the equip-
ment necessary for increased propylene recovery, 1. increased riser (reaction) temperature;
despite the relatively low cost of refinery recovery 2. changes in FCC catalyst formulation; and
improvements. Propylene recovery improvements 3. FCC additives designed specifically to
for a typical refinery are estimated at $75MM versus increase the yield of LPG olefins.
a new steam cracker at $500MM; or new refinery
propylene capacity at $800MM.1 Increasing riser temperature will typically
With light olefin demand at an all time high, increase undesirable, low value gas make, and
the market is still struggling to rebuild reserves and changes in catalyst will only affect LPG olefin pro-
come into balance. Undoubtedly, many refiners duction to a small degree. On the other hand, FCC
missed the opportunity of the decade to take advan- additives designed for LPG olefin production can sig-
tage of this market. However, a significant opportu- nificantly alter the FCC yield slate and improve prof-
nity still exists as the market remains strong for
20
itability in a very controlled, predictable way without tion of gasoline and, in some cases, from the destruc-
an increase in dry gas make. Further, the nature of tion of LCO, propylene values alone can be mislead-
FCC additives allows refiners to respond rapidly to ing. The value of the increased LPG olefin yield
spot markets. must be weighed against the value of the lost gaso-
Since their introduction into the market by line.
Davison Catalysts in 1984, light olefin additives have This incremental value from the conversion
gained wide acceptance in the refining industry as of gasoline to LPG olefins with a light olefin additive
the most cost-effective method of producing incre- is referred to as the "gasoline conversion value". The
mental LPG olefins. Davison Catalysts currently gasoline conversion value can be easily calculated,
supplies four grades of light olefin additives: given a refiner’s gasoline, propylene, and butylene
values. Basically, the gasoline conversion value is
OlefinsPlus® calculated by subtracting the value of the lost gaso-
For high dispersion / low use operations line from the value of propylene and butylene pro-
duced from the lost gasoline.
OlefinsExtra™
For mildly circulation limited operations In other words,

OlefinsMax® Gasoline Conversion Value =

For high activity / maximum stability (activity Propylene Gained % Value + Butylene Gained %
retention) Value - Gasoline Lost % Value
Also minimizes dilution effects, leading to
improvements in conversion Mass balanced, the total propylene + butylene is
equal to the mass of gasoline lost. So if product val-
Butimax® ues are converted to mass measurements, the calcu-
For true butylene selectivity and lation simplifies to,
alkylation feedstock applications
Gasoline Conversion Value =
The appropriate use of these additives (X) Propylene Value + (1-X)Butylene Value –Gasoline
depends on the specific refiner’s objectives and unit Value
operating constraints. Additionally, since this class
of product generates LPG olefins from the destruc-

Figure 3
Historical U.S. Gasoline to LPG Olefins Conversion Value

12.0
Gasoline to LPG Olefin Conversion Value

10.0

8.0
(cents/pound)

6.0

4.0

2.0

0.0
Jan-90

Jan-91

Jan-92

Dec-92

Jan-94

Jan-95

Jan-96

Dec-96

Jan-98

Jan-99

Jan-00

Dec-00

21
This is a simple model since light olefin additives Example
also significantly increase the octane of the gasoline.
Thus, the value of the remaining gasoline is higher In the example, shown in Table I, the gaso-
than the starting value. line conversion value and the resultant economic
Average gasoline values are available from a impact are calculated for a typical 50,000 B/D FCC
number of sources. CMAI, Chemical Week, and unit in North America. Product values are U.S. Gulf
Bonner & Moore are, among others, good sources. Coast, June 2000, values quoted by CMAI.3
Typically, propylene is quoted regionally, on a cents In this example, for every gallon of gasoline
per pound basis by grade with contract and spot val- destroyed in the production of propylene and buty-
ues distinguished as appropriate. lene, the refiner makes between $0.517 and
$0.62/gal, depending on the grade they are able to
There are three general grades of propylene, defined produce. Therefore, for every volume percent gaso-
as follows, line destroyed, the refiner makes approximately an
additional $11,000 to $13,000 per day. Note, this
• Polymer Grade: minimum 99.5% Propylene does not include the incremental octane value of the
• Chemical Grade: minimum 92% Propylene gasoline of about 0.6 (R+M)/2 per volume percent
• Refinery Grade : 60-85% Propylene gasoline loss.
A typical refiner will take significant advan-
Butylene values are based on their alkylation tage of light olefin additive technology by converting
value. Therefore, a refiner’s ability to take advantage 2 to 4 vol% of gasoline to propylene and butylene.
of light olefin additives also depends on the refiner’s Some refiners are converting more than 8 vol% gaso-
ability to utilize or market the additional butylene line, and with the newer technologies, some might
produced by these additives. consider more aggressive applications.
Published values are meant as a guide. In the current market, producers of all propy-
Actual product values depend on local market condi- lene grades might take advantage of this technology.
tions; refinery location relative to infrastructure, But even during downturns in the propylene market,
such as ports, pipelines, or rail lines; location relative e.g., the downturn at the end of 1998, gasoline con-
to petrochemical plants; and refinery configuration, version to propylene and butylene would still be eco-
such as wet gas compressor capacity and splitter nomical for producers of higher-grade (polymer and
capacity. possibly chemical grade) propylene. With relatively
small capital expense or modifications to existing
equipment, many refiners would be able to partici-
pate more completely in this growing market.

Table I
Commercial Example of Gasoline Conversion Value

Feed Rate 50,000 B/D

Feed Gravity 24.5˚ API

Yield Value
Propylene (Polymer Grade) 4.2 wt.% $0.275/lb
Propylene (Chemicalr Grade) 4.2 wt.% $0.260/lb
Propylene (Refinery Grade) 4.2 wt.% $0.248/lb
Butylene (Alkylation Value) 6.0 wt.% $0.223/lb
Gasoline (Unleaded 87 (R+M)/2) 52.0 wt.% $0.155/lb $0.966/gal
Gasoline Gravity (56.5˚ API)

Gasoline Conversion Value

Polymer Grade Case $0.099/lb $0.620/gal
Chemical Grade Case $0.090/lb $0.564/gal
Refinery Grade Case $0.083/lb $0.517/gal

22
Additive Selection For example, to produce the same propylene
yield, a refiner could use just 5% of high activity
The selection of an appropriate additive is OlefinsMax rather than 12.5% of a low activity
equally important. One of the most critical selection additive like OlefinsPlus. As shown in Figure 4,
criteria is activity and Davison continues to lead the the advantage of the higher activity additive is that
industry with the most active, commercially avail- the unit can achieve constant conversion at 8% lower
able products. cat-to-oil ratio, due to less dilution. This is particu-
Activity is becoming more important as refin- larly important as units increase the level of additive
ers begin to recognize the value of maximizing propy- in inventory, and catalyst circulation constraints
lene. These additives act essentially as an inert rel- become more relevant. To maintain constant equilib-
ative to the primary cracking function of the FCC rium catalyst activity, the typical refiner would be
catalyst, diluting the catalyst inventory and, there- able to lower fresh catalysts addition substantially
fore, diluting the overall activity of the catalyst. by using the more active light olefins additives.
While many refiners are able to make some minor Because the response of equilibrium activity to fresh
modifications to the catalyst circulation rate to over- catalyst make-up is non-linear, avoiding the incre-
come the dilution effect at low levels of additive in mental 7.5% dilution allows a disproportionately
inventory, the effect becomes more problematic as large 15% decrease in fresh catalyst additions to
refiners push additive levels higher in response to maintain constant equilibrium activity. Depending
increased propylene demand. on catalyst cost and daily catalyst additions, this sav-
ings could be very significant.

Figure 4
Effect of Dilution on Cat/Oil Requirement to Maintain Constant Conversion
(Pilot Plant Data)

9.0

•
C/O Ratio for Constant 72.5 % Conversion

8.5

8.0
•
7.5

Lower Cat/Oil
• ▲

7.0
•
▲

Less Additive
Dilution
6.5

0 5 10 15 20 25 30 35

Additive Level in Inventory, wt%

23
Conclusion ished inventories, spot prices have recently run $0.04
to 0.07/lb higher than contract prices.4
Refineries are in the best position ever to Davison was the first to introduce light olefin
take advantage of the high value propylene market. additives for use in the FCC unit in 1984. Since
Although the propylene market will likely see some then, Davison has continued to lead the industry as
correction this year, demand should continue to an innovator of high quality light olefin additives
increase at about 6% per year. Current non-refinery with the introduction of high activity products like
propylene production will not be able to keep pace OlefinsMax®, the most cost effective light olefin
with demand, increasing capacity utilization rates. additive available. Continuing research and devel-
Additionally, profitability of non-refinery sources of opment in this important area, and alternative mar-
propylene, like naphtha crackers, are subject to fluc- ket approaches, may allow refiners to push the limits
tuation in feedstock costs. When feedstock costs are of propylene production from the FCC to previously
high, naphtha cracker operator’s incentive to operate unattainable levels.
at or near full capacity is diminished.
Although refineries able to produce polymer
grade propylene are in the best position to take 1 Zinger, S. J., "The Critical Role of the Refinery in the
advantage of the current market, market conditions Propylene Market", 2000 World Petrochemical
are also ripe for participation by producers of chemi- Conference, Houston, TX, March 29-30, 2000
cal and refinery grade propylene. The price spread 2 Harris, G., "Polypropylene - Critical Mass or Critical
between polymer and refinery grade propylene has Mess?", 2000 World Petrochemical Conference, Houston,
recently closed to $0.02 to 0.03/lb and chemical grade TX, March 29-30, 2000
prices are running only $0.015/lb lower than polymer 3 CMAI Monomers Market Report, June 30, 2000

grade. With the tight supply situation and dimin- 4 CMAI Monomers Market Report, June 15, 2000

Look for interactive versions

of our FCC Value Models to
be added to our website in
the near future.

24
 Update: Value-Added Gasoline Sulfur
Reduction Strategies
By
Jeffrey W. Balko
Marketing Manager, Clean Fuels Technologies
Davison Catalysts

As Davison Refining Catalyst representa- The Cost of Sulfur Reduction

tives around the world visit with refiners, the pri-
mary message we receive is the desire to produce
In the United States alone it is estimated
cleaner fuels more cost effectively. Most recently, the
that Tier 2 gasoline sulfur regulations will cost refin-
focus has been on removing sulfur from hydrocarbon
ers $3-5 billion annually, while 15 ppm sulfur diesel
fuels and, as a result, the call for economical sulfur
is expected to have an annual price tag of at least
reduction solutions has been paramount.
this much. Only time will tell the ultimate con-
Davison foresaw this need and has been
sumer price impact of these additional refining costs.
devoting significant resources to developing and
However, one needs only look at the Phase 2 RFG
commercializing catalysts and additives for FCC
and CARB gasoline premiums relative to
gasoline sulfur reduction. These efforts have materi-
Conventional Gasoline (CG) this past summer to rap-
alized into three commercially proven gasoline sulfur
idly appreciate the costs associated with producing
reduction materials: SuRCA™ and GSR-1™ world-
cleaner fuels. Depending on the region of the United
wide and GFS™ in Europe. In order to help our cus-
States, RFG premiums relative to CG this summer
tomers identify economic opportunities in the gaso-
were in the $0.07-0.12/gal range. Certain areas of
line sulfur reduction area, Davison has also devoted
the country, such as the Midwest, saw this differen-
substantial efforts to understanding the short and
tial soar as high as $0.25-0.30/gal due to multiple
long-term value of FCC gasoline sulfur reduction. In
production and supply issues.
order to provide more comprehensive technical serv-
For this reason, it is worthwhile to think
ices and help our customers improve profitability,
about the economic value of gasoline sulfur reduction
several Clean Fuels Value Models have been devel-
from the current and future standpoint. In many
oped. While these Value Models are no substitute for
situations it is expected that the value of a 20-30%
refinery LP’s, they can help in identifying opportuni-
lower sulfur FCC gasoline stream would far out-
ties that should be further explored with an LP.
weigh the typical catalyst cost of $0.002-0.003/gal of
This article is an overview of some of
FCC gasoline or roughly $0.0008-0.0012/gal of the
the economic opportunities associated with catalytic
total gasoline pool.
FCC gasoline sulfur reduction, the value associated
with these opportunities, and an update on the com-
mercial performance of our gasoline sulfur reduction
products.
25
The Current Value of Gasoline Sulfur LCO. While refiners will sometimes undercut the
heavy end of the FCC gasoline into the LCO stream
Reduction to increase distillate product yields due to favorable
economics, this undercutting is also sometimes done
Based on current regulations there are sev- purely to meet sulfur specifications. Davison has
eral ways in which refiners can take advantage of seen several cases where refiners were forced to
incremental FCC gasoline sulfur reduction. These undercut FCC gasoline during times of excellent
are: gasoline/#2 oil margins in order to meet annual aver-
age complex model requirements for conventional
1. Reduced FCC gasoline undercutting into LCO gasoline. The catalytic sulfur reduction provided by
to comply with sulfur specs SuRCA or GSR-1 allows refiners to increase the
2. Increased production of higher valued RFG or gasoline endpoint and regain lost gasoline barrels.
CARB gasoline Figure 1 shows the graphical output of our Gasoline
3. More cost effective production of RFG or Undercutting Value Model. Input data to this model
CARB includes the gasoline sulfur distribution curve for a
4. More cost effective compliance with annual given refiner as determined by sulfur specific
complex model conventional gasoline require GCAED (Gas Chromatagraph with a sulfur Atomic
ments Emission Detector); the value spread between gaso-
5. Processing of less expensive (higher sulfur) line and #2 oil; and the expected gasoline sulfur
FCC feedstocks reduction. The model then calculates the allowable
6. Production of additional barrels of FCC gaso increase in gasoline cutpoint back to the base sulfur
line within total gasoline pool sulfur con level and the associated increase in gasoline volume
straints and value. In Figure 1 the three curves illustrate
7. Trading off with other US/CARB complex how different gasoline cutpoint increases associated
model variables. with different levels of sulfur reduction would trans-
late to yield value for this 50,000 B/D FCC unit. In
One of the most obvious benefits for using order to demonstrate the sensitivity of this value to
SuRCA or GSR-1 to reduce gasoline sulfur is the gasoline-LCO spread, values are plotted versus this
potential for reduced gasoline undercutting into differential. The data clearly shows how even at a

Figure 1
Refiner A - Cost of Compliance

◆
Total Incremental Yield Value ($/day)

$20,000
Gasoline Sulfur Reduction results in 20˚F T90 Increase
▲
■ Gasoline Sulfur Reduction results in 30˚F T90 Increase
$18,000
▲ Gasoline Sulfur Reduction results in 40˚F T90 Increase ▲
$16,000 ■
$14,000 ▲
■
$12,000
▲ ■ ◆
$10,000
■
◆
$8,000 ▲ ◆
$6,000
■ ◆
▲ ◆
■
$4,000
◆
$2,000
0 1 2 3 4 5 6 7

Gasoline/#2 Oil - Value Differential ($/bbl)

26
fairly typical differential of $3.00/BBL, the economic other potentially constrained gasoline properties.
value can easily reach $5,000-$9,000 per day for this Figures 2 and 3 were developed using the Federal
50,000 B/D FCC unit. EPA Complex Model. Figure 2 shows the impact of
In our August 2000 Catalagram News, an gasoline sulfur concentration on the reduction of
economic case was presented for using incremental major pollutants such as NOx, toxics and VOC’s
gasoline sulfur reduction achieved through use of (volatile organic compounds). It is clear to see that a
SuRCA or GSR-1 to blend additional barrels of high- reduction in gasoline sulfur concentration results in
er value RFG. This case was developed using our reductions of all three pollutants. The reduction in
RFG Value Model. An even simpler way to look at VOC’s could be used to provide some RVP relief,
the potential value of a gasoline sulfur reduction is to while translating the NOx reduction down to Figure
see what complex model relief it would allow for 3 shows that olefins level could be increased, result-
Figure 2
Complex Model - Effect of Sulfur Level

15
▲
▲
RFG Phase 2 Reduction/ (Increase)

10 ◆ ▲
◆◆▲
◆
in Emissions (%)

5 ▲
◆
■■■
■
■ ▲
◆
0 ▲
■
■ ◆ ■
▲
◆ ■
▲ ■
-5 ▲ ■
◆ ▲ ▲
■
-10
◆
◆
-15
◆
-20
0 100 200 300 400 500 600 700 800 900
Sulfur (ppm)

◆ Total Toxics Reduction ■ Total VOC Reduction ▲ NOx Reduction

Figure 3
Complex Model - Effect of Olefin Level

1.5
RFG Phase 2 Reduction/ (Increase)

▲ ▲
1.0 ▲
◆ ■
◆ ▲ ■
0.5 ◆ ▲ ■
◆ ■
◆ ▲ ■
in Emissions (%)

0.0 ◆■
■ ◆
■ ▲ ◆
-0.5 ■ ◆
■ ▲ ◆
-1.0 ■
◆
-1.5 ▲

-2.0 ▲
-2.5
▲
-3.0
4 6 8 10 12 14 16
Olefins (vol%)

◆ Total Toxics Reduction ■ Total VOC Reduction/ (Increase) ▲ NOx Reduction

27
ing in further VOC reduction and potential RVP The Longer Term Value of Gasoline
relief.
Another potential economic benefit of cat- Sulfur Reduction
alytic sulfur reduction is the processing of poorer
quality, lower cost feedstock components. Since There are also several ways that refiners can
some refiners process crude mixes comprised of lower economically benefit from incremental FCC gasoline
sulfur and higher sulfur crudes, such as ANS and sulfur reduction in the longer term. These are:
SJV on the U.S. West Coast, the relative costs of
these crudes and the need to meet product sulfur 1. Generation and use (or sale) of Tier 2 ABT
specs can sometimes mean good value for the refiner. credits and allotments;
Consider the theoretical example where pool gasoline 2. As part of a compliance strategy for ultra-low
sulfur was the limiting variable and a 30% FCC gasoline sulfur specs to avoid additional capital
gasoline sulfur reduction allows the refiner to spending;
increase SJV processing by 5,000 B/D at the expense 3. Compliance with "mid-point" sulfur regula-
of ANS. If the differential between these crudes is tions in countries like Canada and Australia;
only $3.00/BBL, this would translate to an incre- and
mental crude cost savings of $15,000 per day, typi- 4. Reduce operating costs of HDS options (hydro
cally more than enough to pay for even a premium gen and octane loss) for compliance with ultra-
catalyst technology. low gasoline sulfur specifications.

Figure 4
Gasoline Pool Sulfur Concentrations for North American FCCU’s

40

35
Number of Units within Range

30

25

20

15

10

0
<=30 <=80 <=150 <=200 <=300 >300

Pool Gasoline Sulfur Level Range (ppm)

28
 Figure 4 shows the range of gasoline pool ning in 2004. These credits and allotments are gen-
sulfur levels for North American refiners. These erated by either "pre-reducing" sulfur relative to the
data come from a Davison Catalysts database kept refiner’s ‘97/’98 baseline (2000-2003) or reducing sul-
up-to-date through periodic customer surveys and fur level below target levels in 2004-2006. The obvi-
represent about 70% of the North American refining ous benefit of this program is that a participating
industry. There are several important points to be refiner can delay the implementation of a capital
drawn from these data. solution and the associated operating costs for one to
First, there are clearly several refiners that three years. Refiners who have current pool sulfur
may be able to meet ultra-low sulfur gasoline specifi- levels below 300 ppm (below the Tier 2 2004-2005
cations with minimal capital investment. For these cap) and especially those with pool sulfur levels near
refiners, the benefits of technologies like SuRCA and and below 120 ppm (at or near 2004/2005 corporate
GSR-1 might easily mean the difference between average standards) are well-positioned to take
building or not building a product post treater. In advantage of this program.
fact, several U.S. refiners have trialed and are plan- The value of these credits and allotments will
ning on SuRCA, GSR-1 or improved technologies as ultimately be decided by the market and will most
part of their long term compliance strategy. likely fall somewhere between the "cost of genera-
Second, many of these refiners are in an tion" of the credits/allotments and the "cost of com-
excellent position to take advantage of the Tier 2 pliance" with Tier 2 gasoline sulfur specifications.
ABT (Averaging Banking and Trading) Credit and While the cost of compliance will vary from refiner to
Allotment program. Much has been written about refiner, many references have been made to costs in
this program and only a very simplified and limited the range of $0.02 to 0.04/gal of total gasoline pool.
overview is provided here. In the simplest terms, Figure 5 is a simplified example developed
this program allows refiners who can meet the 300 using an ABT Credit Value Model and shows the
ppm sulfur cap beginning in 2004 to use credits to potential economic value of credits generated
meet the 30 ppm sulfur standard beginning in 2005 through the use of a Davison’s SuRCA or GSR-1.
and allotments to meet the corporate average stan- This example shows the number of days Refiner C
dards (120 ppm in 2004 and 90 ppm in 2005) begin- with 100,000 B/D of gasoline production can post-

Figure 5
ABT Credit Value for Refiner C

900 120

Days of Delay for 30 ppm S Gasoline

800
ABT Credit Value (MM$)
100
Days of Delay for 30 ppm S Gasoline

700
ABT Credit Value (MM$)
600 80

500
60
400

300 40

200
20
100

0 0
10 15 20 25 30 35

% Gasoline Sulfur Reduction

29
pone production of 30 ppm sulfur gasoline. Figure approximately 30% light and heavy gasoline sulfur
5 also illustrates the approximate cost savings of reduction. Since that time, two additional applica-
postponement for various levels of FCC gasoline sul- tions of SuRCA have occurred, with preliminary
fur reduction and associated credit generation. The results indicating roughly 25-30% light and heavy
days of postponement are calculated based on the gasoline sulfur reduction in both cases. Results of
assumption that Refiner C commenced use of a these applications will be published in the near
SuRCA catalyst on July 1, 2000 and continues to use future. The SuRCA family of catalyst contains
the catalyst through 2006. The cost savings are cal- Davison’s GSR-4 technology and can be designed
culated assuming a "cost of compliance" of $0.02/gal with any Davison catalyst family as the starting
for 30 ppm sulfur gasoline. It is also assumed in the point. It is important to note that four of the five
example that the refiner’s starting sulfur level is commercial applications of SuRCA thus far have
equivalent to their ‘97/’98 baseline (the baseline for been cases where the base catalyst being replaced
credit generation) and that the 300 ppm sulfur cap, was a very high unit cell size catalyst that would
120 ppm (2005), and 90 ppm (2004) corporate aver- have already decreased the gasoline sulfur selectivi-
age standard can be met. ty by roughly 10%.
There are currently refiners using Davison GSR-1 gasoline sulfur reduction additive has
sulfur reduction technology to generate ABT credits. now been used in 13 FCC units in North America and
It is also worthwhile to note that there will be some Europe and is currently in use in five. Recent appli-
refiners who are ideally positioned to be "credit cations have shown consistent sulfur reductions of
users" and some who are better positioned to be about 20% with additive usage levels in the 5-10%
"credit generators". Davison understands this range. The performance per usage level is dependent
dynamic and is able to help refiners meet their cred- on several variables.
it strategies and demands more cost effectively by Based on key findings from commercial
looking outside their own organizations. applications of GSR-1 and extensive research and
development efforts over the last two years, Davison
Update on Davison Premium is currently in the process of commercially evaluat-
ing a substantially improved sulfur reduction addi-
Gasoline Sulfur Reduction FCC tive. Two commercial trials of this new additive com-
Catalyst and Additives Technologies menced this summer and, based on extremely favor-
able results, one of the refiners has already decided
The year 2000 has been a very busy one for to continue using the material on an ongoing basis.
Davison gasoline sulfur reduction catalysts and addi- While results from these commercial trials will be
tives. published in the near future along with additional
Our 2000 NPRA paper, co-authored with information on this new sulfur reduction additive,
Murphy Oil, showed how the use of SuRCA at some limited laboratory information is shown in
Murphy’s Meraux, Louisiana FCC unit resulted in Table I.
The GFS family of catalysts, developed in
Europe, is currently being commercialized and has
Table I already provided a 25-30% sulfur reduction over a
REUSY base catalyst in one application. Several
Sulfur Reductions Base Ecat GSR-1.1 GSR-1 additional commercial trials are ongoing.

Light Cut Sulfur Base 45% 33% Summary

Heavy Cut Sulfur Base 21% 16%
Cut Gasoline Sulfur Base 37% 27%
Total Sulfur Base 22% 16% As worldwide gasoline sulfur levels are
pushed to ultra-low levels, the costs to produce these
fuels will clearly rise. The use of commercially avail-
Results generated in ACE Pilot unit able FCC catalytic sulfur reduction technology, such
testing with VGO feed as SuRCA, GSR-1 and GFS, is one way that refiners
can minimize these production cost increases. These
Base “Total” gasoline sulfur of 1300 ppm, base technologies have been commercially proven, and
“Cut” gasoline sulfur of 800 ppm when applied properly, can be a very cost effective
part of clean fuels production.

30
Announces New FCC Catalyst Solutions
Four new product families have been launched to address the growing need
for custom, value-added technology in FCC processing. These products have
been commercially and laboratory tested for superior performance in their tar-
get applications. Because they deliver premium performance, these new cata-
lysts families are available on a limited basis. They are:

Formulated for high activity retention

with superior coke selectivity and
bottoms cracking for short contact
time (SCT) or gas oil operations

Specifically designed for improved activity

and superior selectivity under severe operat-
ing conditions

Offers superior stability, activity, and selectivi-

ty for resid operations

The next generation of premium,

vanadium-tolerant FCC catalysts

31
 Grace Davison
Hydroprocessing
Catalysts and Applications

Straight 400 Series

▼

Run 500 Series

Naphtha
▼
Coker 400 Series
▼

CK500 Series
Crude Tower

Naphtha
▼
Kerosene Gasoline
▼

▼
& Jet 400 Series Solvents
Jet Fuel
Diesel CK400 Series Kerosene

▼
▼

CK500 Series Diesel Fuel

Heating Oil
Lube Oils
Hydrocracker
White Oils
▼

CK500 Series
Pretreat
Waxes
VacuumTower

▼
FCC CK400 Series
▼

Pretreat CK500 Series

▼
Lube Oil
500 Series
& Speciality

Residual Oil GR® Series and other

▼

Custom catalysts Other Processes

Treating

32
 Announcing e-Catalysts...
Logistics Supply
Vertical Support Chain
Community Management

Secure
Transactions

sm
Comprehensive
Searchable
Catalog
Secure
Information
Exchange

The versatile, all-purpose tool to meet

the e-business needs
of the catalyst industry
e-Catalysts is the new “one-stop shop” for catalyst
manufacturer s , their customer s and supplier s . A ver tical
mar ketplace with both focused technical depth and industr y
breadth, it provides an Inter net por tal to ena ble transactions ,
infor mation exchange and suppor t for the catalysts industr y.
e-Catalysts includes a searcha ble “exper t-system” data base to
allow purchaser s to find infor mation, identify supplier s and
make buying decisions .

For more information, or to become a member, contact us at

[email protected]

e-Catalysts, Inc.
700 Dresher Road ®

Horsham, Pennsylvania 19044

Phone: (215) 315-3269
 W.R. Grace & Co.-Conn., 7500 Grace Drive, Columbia, MD USA 21044
Phone: (1) 410-531-4000; Fax: (1) 410-531-4101
[email protected]

©2000 W. R. Grace & Co. - Conn.

We hope the information given here will be helpful. It is based on our best knowledge and we believe it to be true and accurate.
Please read all statements, recommendations or suggestions herein in conjunction with our conditions of sale which apply to
all goods supplied by us. We assume no responsibility for the use of these statements, recommendations or suggestions, nor
do we assume any responsibility in the event that our goods or any use of such goods infringe any patents.