Catalysts in Petroleum Refining and Petrochemical Industries 1995

M. Absi-Halabi et al. (Editors)

9 1996 Elsevier Science B.V. All fights reserved. 171

MINA ABDULLA REFINERY EXPERIENCE WITH A T M O S P H E R I C RESIDUE

DESULFURIZATION (ARDS)

A. A I - N a s s e r , S. R. C h a u d h u r i , a n d S. B h a t t a c h a r y a

Mina A bdulla Refinery, Kuwait National Petroleum Co., Mina Abdulla, Kuwait.

ABSTRACT
Mina Abdulla Refinery was modernized in the late Eighties. The Crude Processing capacity was
expanded and a number of downstream processing units were installed. Based on the design philosophy
of minimum Fuel Oil production from the Refinery, the new processing scheme included a Hydrocracker
and a Delayed Coking unit with Vac. Gas Oil and Vac. Residue as feed stocks respectively. Light and
Middle Distillates Hydrotreaters along with Atmospheric Residue Desulfurization (ARDS) Unit were
also incorporated in the refinery process scheme..
Three different catalyst systems have been utilized in the ARDS Unit, namely single HDS, dual
HDS/HDM and multiple HDM/HDS/HDS-HDN catalyst combinations. This paper presents the
performance and an analysis of the three systems with particular reference to catalyst life. The Refinery
experience will be useful in selecting improved catalyst type and combinations to optimize run length
and products.

INTRODUCTION
ARDS unit works as the springboard in the new scheme of Mina AbdullaRefinery
operation. Primarily a desulfurization unit, ARDS also reduces the metals, asphaltenes and
nitrogen in the products, thereby, ensuring proper quality of feed for downstream conversion
units. As an additional benefit, ARDS is also a mild hydrocracking process, partially upgrading
high sulfur atmospheric residue to low boiling products like naphtha and distillate.
Choice of the catalyst system and severity of operation, therefore depends on the following
major factors :
- ability of the catalyst system to meet the target product qualities during the run length.
- ability of a catalyst system to ensure a steady economically viable run length.
- stable by-product distillate.

ARDS UNIT G E N E R A L DESCRIPTION

ARDS unit has two trains, each having a Guard Chamber and three main reactors with a
common fractionation section. A simplified flow scheme is shown in figure (1). Other than the
Off-Gases, the three liquid products are Naphtha (C5- 375~ TBP), Distillate (375-680~
TBP) and Low Sulfur Fuel Oil, that is LSFO (680~ + TBP).
Feed stock

The quality of Feed to the ARDS unit generally remains comparable to the specification
used in unit design as shown in Table 1.
172

M A K [ UI'
Ft..,, I'UiI(;F

GAS

v----
NAPHTHA

RESID GAS
FEED MAKEUP OIL
HYDROGEI~ JRGE
FUE L OII

GAS

Figure 1. Process flow diagram of the ARD unit.

Table 1.
Properties of ARD feedstock.

Design Typical
Gravity ~ 12.5 13.5
Sulfur wt. % 4.5 4.3
Con. Carbon wt. % 12.2 max. 12.0
Metals (Ni & V) ppm wt. 88 max. 75
Sodium ppm wt. 3 2.0
Nitrogen ppm wt. 2800 2500
Distillation vol. %
recovery @ 680~ 6 6

Unit Processing Objectives

Processing objectives of ARDS unit may be categorized as below:
1. Feed Preparation for Down Stream Units
- Reduce Nitrogen in VGO to conform to Hydrocracker and FCC units feed
specification requirement.
- Control sulfur in Vacuum Rerun Unit feed in order to obtain a desired coke quality
as well as to be within metallurgical limits of the unit.
 173

2. Fuel Oil Blending

- Produce LSFO of typically 0.5 % wt. sulfur for use as a blending component in the
F.O.
3. Product Slate improvement
In addition to the above, mild hydrocracking associated with the process leads to
production of Distillate and Naphtha thus helping upgradation / optimization of refinery
product slate.
Reactors Arrangement
Each ARDS train has four reactors holding about 28,000 CFT of catalyst. First one is a
Guard Chamber loaded with about 7 % of total catalyst. The other three reactors hold about
3 1 % each of total catalyst.
Quench Hydrogen is available at the inlet/outlet of the Guard Chamber and outlet of
reactors 2 & 3. Quench is utilized for adjusting the reactors inlet temperatures to maintain
desired temperature profiles in the catalyst beds.
The Guard Chamber can be bypassed in case of high pressure drop due to scale or coke
accumulation to enable continued operation. About 20% ( max. ) of the total reactor system
pressure drop is admissible across the Guard Chamber.

Unit Operating Variables & Modalities

The ARD trains are generally operated at design capacity of 33,000 BPD each. The
Recycle Gas (85% H2) rates are maintained as high as the system permits - usually giving a
Gas : Oil ratio between 4500 - 5000 SCF/BBL at the reactor system outlet. The reactors
system pressure is maintained at the design level between 1800 - 2000 psig. There is facility for
water injection in the charge heater coils to improve catalyst performance.
Generally, a fiat temperature profile is maintained at the outlet of the three main reactors,
with GC outlet remaining ~ 20 - 30~ lower. Maximum exotherm allowed in the three main
reactors is around 50~ GC exotherm is generally limited to 25~ This is to ensure a more
controlled and uniform deactivation of the catalyst system.
Some catalyst systems are characterized by higher pressure drop in the reactors than the
other. To elaborate a little, pressure drop in the reactors is a function of shape (i.e. cylindrical,
spherical, trilobe or quadralobe) of catalyst and its size. It has been observed that pressure drop
can be brought within operating limits by varying Gas / Oil ratio or changing oil viscosity.
However, reduction in pressure drop by Gas/Oil ratio adjustment is mostly at the cost of
expected life of the catalyst. Further discussion on this is followed in the next paragraph.

C A T A L Y S T SYSTEMS USED AT MINA ABDULLA R E F I N E R Y

Three distinct catalyst combinations have been used at Mina Abdulla ARDS unit.
A : Mono Catalyst - HDS
B : Dual Catalyst - HDM + HDS
C : Multiple Catalyst - HDM + HDS + l I D S / H D N
General specifications of the catalysts are given in Table 2.
174

Table 2.
Catalyst Specifications.
Catalyst System A B C

Catalyst type HDS HDM HDS HDM HDS HDN

Combination % 100 52 48 32 52 16

Active metal CoMo CoMo NiMo Mo NiMo NiMo

Tri-/
Shape Quadra- Quadra lobes Quadra lobes /
lobes Cylindrical
Size 1/20" 1/20" 1/20", 1/32"
1/4" 1/8" 1/4" 1/8" 1/5" 1/10"

All the three systems employed graded bed in order to tackle the reactor pressure drop,
especially at the front end where the operating temperatures are lower and catalyst encounters
the virgin stock.
Most of the catalysts used are tri-lobes or quadra-lobes which are known to ensure better
wetting of the catalyst surface with consequent improvement in liquid flow pattern and
sustenance of catalyst activity. On the other hand, cylindrical shaped catalysts though results in
lesser void fraction has poorer wetting characteristics.
Catalyst size selection is also of utmost importance. More catalyst can be packed in
available volume with lesser catalyst size. However, reaction section hardware may pose a
limitation. A minimum size of 1/20" seems to be optimum for Mina Abdulla Refinery reactor
system.
Last but not the least, criterion for good performance of a catalyst system is its crush
strength. A refinery catalyst unit is liable to encounter upsets during its run length. A quality
that a refiner looks for in the catalyst is its ability to sustain such upsets without any adverse
impact on its performance.

PERFORMANCE OF THE CATALYST SYSTEMS

Catalyst System A: HDS
During initial stages of operation the catalyst was very active i.e., SOR WABT was about
25~ lower than predicted. The rise in WABT in first four months was continuous @ about
6~ Subsequently, the catalyst activity declined sharply(WABT rise 10-12 ~
The pressure drop across the main reactors remained low (35-55 psi) and steady throughout
the run length.
 175

The naphtha and distillate makes were significantly high. The catalyst behavior was
explainable in view of the operating requirement i.e., primarily LSFO sulfur. Since the catalyst
was solely HDS type, a high degree of desulfurization was obtained at lower temperature at
SOR. On the other hand, the feed metals poisoned the HDS catalyst very fast, initially at the
front end and then the subsequent beds. Once this happened, the WABT had to be raised
steeply to attain target LSFO Sulfur.
Due to the high fouling rate, the catalyst cycle length was lower than expected.
Catalyst System B: HDM / HDS
The catalyst showed high rate of deactivation (9~ WABT ) during first 2
months or so followed by a relatively steady period of 5-6 months (4~ WABT). The
EOR was characterized by steep temperature rise.
After six months of operation, the GC was by-passed in both trains. Pressure drop across
the main reactors remained almost steady throughout the run (55-75 psi). Presumably the
effect of catalyst fouling was countered by the increase in reactors temperatures from SOR to
EOR.
The behavior of HDM/HDS combination was as expected. As the amount of liDS catalyst
was much less than that in Catalyst System-A and desulfurization activity of front end
catalyst (HDM) declined fast, WABT had to be raised progressively right from the SOR to
obtain target LSFO Sulfur. At mid-of-run, both HDM and HDS catalysts were at their
optimum temperature levels and a rather steady activity (& LSFO Sulfur) was observed for
some time. This was expected as the HDM catalyst in the front end of the reactor system
protects the HDS catalyst from getting fouled up by metal deposition.
At EOR, the scavenger HDM catalyst was exhausted and metals in feed started poisoning
the HDS catalyst. This phase was characterized by a steep rise in WABT to maintain LSFO
Sulfur. The drift of exotherm from Reactor-1 to Reactor-2, as observed, was due to saturation
of HDM catalyst and consequent shiit of relatively more virgin feed (& reaction) to the next
catalyst bed.
The pressure drop characteristics of this batch remained stable throughout the run which
allowed comfortable Gas/Oil ratio. This did contribute to the overall relatively good
performance of this catalyst in terms of cycle length.
Catalyst System C: HDM/HDS/HDS- HDN
An initial period of stabilization (about two months) was experienced with this catalyst
system, when high (8~ catalyst deactivation rate was observed. After this, a period of
low catalyst deactivation rate followed (4-5~ month WABT). Regarding the reactors
exotherms, the same trend as with Catalyst System-B was observed. First main reactor showed
about 50% of total delta temperature which drifted to the next reactor at EOR.
As expected, the behavior of Catalyst System-C was similar to that of Catalyst System-B
in view of similar HDM catalyst guard at the front end. However, the reactors pressure drop
was much higher from the SOR itself, enforcing a lower Gas/Oil ratio that resulted in enhanced
catalyst fouling and deactivation. The higher pressure drop was related to the characteristics
(i.e. size and shape) of the catalyst. The products yield was, as expected, similar to that of
Catalyst System-B.
176

Table 3.
Comparative Performance of Catalyst Systems.

ATTRIBUTES UNIT CAT-A CAT-B CAT-C

Effective run length Months 8.8 10.5 10

Effective catalyst life BBL/Ib 8.1 11 9.5
Operating LSFO Sulfur %wt. 0.65(0.5-0.9) 0.58(0.53-0.73) 0.65(0.5%0.$2)
Demetallation %wt. 50 70 82
Denitrification % wt. 35 40 39
Naphtha Dist. Yield % Vol. 16-33.2 20-25 19-30
VGO % Vol. 43.5 40-45 46
Make-up Hydro[en SCFB 1005 1015-1125 1015-1050
Rx. Pressure drop psi. 45 (65)* (80)*
* GC was bypassed partially or totally

CATALYST PERFORMANCE ANALYSIS

The performance of the three catalyst systems are summarized in Table 3. The three
systems operated with the same feed and nearly the same target Sulfur in LSFO. Their most
significant differences in performance were as follows:
In Catalyst System-A, the metals in the feed poisoned the catalyst very fast imposing
higher severity on the remaining part which deactivated as fast. This was apparent in the
progressive and continuous increase of WABT throughout the cycle length (Fig.2). Though the
WABT, to start with, was lower than catalyst systems B & C, rate of deactivation was
very high resulting in shorter run length. Other relevant operating and performance parameters
are shown in Fig. 3 & 4. However, this system generated more distillate towards EOR which
again could be attributed to the characteristics of the catalyst system.
As was expected, demetallation/denitrification of the feed stock was substantially lower
than that for Catalyst Systems B and C. Even though product pattern with this system was
superior, short catalyst cycle length and consequently limited On-Stream Factor (OSF)
weighed heavily against it. Short catalyst cycle length was operationally uneconomical due to
frequent shut downs.
Catalyst System-B was a significant improvement in terms of cycle length. The Demet
catalyst at front end of the reactor system acted as metals scavenger and protected the activity
of HDS catalyst down below. In general, the run was also characterized by almost constant
distillate production from SOR uptill EOR situation. It is worth mentioning here that this
Catalyst System-B tolerated a prolonged shut down without any appreciable deterioration in
performance.
Catalyst System-C incorporated Denitrification catalyst intended to lower Nitrogen in
VGO (Hydrocracker Feed stock). The advantages over Catalyst System-B, if any, was not
clear. Higher Distillate stability was expected but could not be substantiated. Distillate yield
was comparable with that of Catalyst System-B.
 177

D A Y S ON 01L v s W A B T ( n o r )

750 ~ : ! ' ~ ,
730 ~ ' ~ ~
710 , . . ~ . . ~ _ . ~ . . ~ , t ~ ;
690' - - - t - 1. I __-"-'~ ! !
~ ~ i I
670 I i ; '
eso , t i I,
0 25 50 100 150 200 250 300 350
DAYS ON OIL

D A Y S ON 01L vs H2 C O N S U M P T I O N

I.~ 1100
1000
~ k

d
i j
900
i

~ 800
r

O 700
m
50 100 150 200 250 300 350
DAYS ON OIL

D A Y S ON 01L vs G A S / O I L R A T I O @ G C I N L E T

~ 5000i
t i " 1
4soo;
~ 3500.
~ 3000
2500
0 50 100 150 200 250 300 350
DAYS ON OIL

CAT. SYSTEM A ,t,, CAT. SYSTEM B 9 CAT. SYSTEM C

Figure 2. Days on oil vs. Performance and operating parameters.

178

D A Y S ON OIL v s A P !

14.3
14.1
13.9
13.7
13.5 ........

i . . . . . . . . . . .
13.3
0 50 100 150 200 250 300 350
DAYS ON OIL

D A Y S O N OIL v s S U L F U R

4.354-4 t
4.3
4.25
Z ~ 4.2 if. .~ , --\. I / , , , ~, ..:.~. ~ _ ~ ~ - - ~
U,,I "\ I// ',. / "'\~~ '
" I ~ "

i" "I"
4.1
n~ 4.05
4
i 1 ' '_! .......
0 50 100 150 200 250 300 350
DAYS ON OIL

D A Y S O N OIL v s M E T A L S

n~ 75
70 F i I i t ~ i
CL
~~. 65
~ 6o
ss
~ so 7
0 50 100 150 200 250 300 350
DAYS ON OIL

"- CAT. SYSTEM A l CAT.SYSTEM B # CAT.SYSTEM C i

Figure 3. Days on oil vs. feed quality.

 179

D A Y S O N OIL v s L S F O S U L F U R

0.g
0,8 /

~0.7 4 ' ! / "

~ 06
rr~
nU.! O.5
i !
0.4
0 50 100 150 200 250 300 350
DAYS ON OIL

D A Y S O N OIL v s L S F O M E T A L S

3O I !

. 20 ~

10- 1 I ,
)

50 160 150 200 250 300 350

DAYS ON OIL

D A Y S O N OIL v s D I S T . Y I E L D

30
=~ 25
~ ~o
LU;~ 15
1(?
50 100 150 200 250 300 350
DAYS ON OIL

= CAT. SYSTEM A -" CAT.SYSTEM B | CAT. SYSTEM C

Figure 4. Days on oil vs. product quality and distillate yield.

180

OPTIMUM CATALYST BLEND

To safeguard against metals poisoning, some section of reactors volume is necessarily
allocated to the Demet catalyst. The clue to optimum catalyst combination, therefore, is in
developing improved catalyst and a balance among various types, i.e.
a. Search for higher metal removal capacity HDM catalyst to leave more reactor volume to
HDS catalyst.
b. Search for improved HDS catalyst that gives:
i. Superior activity and stability.
ii. higher tolerance to metals deposition.
c. Improve physical characteristics of catalysts to maximize Compacted Bulk Density (CBD)
within mechanical limitation of the reactor system.

CONCLUSIONS
1. Mono catalyst system, i.e. only HDS, by itself is a poor performer for the feed stock
processed in ARDS unit.
2. A dual catalyst system for ARDS units processing heavy stock, i.e. combination of HDM
and HDS catalysts gives a more stable and longer catalyst cycle length, more suited to
refiner's over-all operational plan.
3. Advantage of a multiple system, i.e. combination ofHDM, HDS and HDN combination
catalyst, over that of a dual-catalyst system is not conclusive. However, HDN catalyst is
expected to impart color stability to naphtha and distillate products.
In short operating experience with different catalyst systems helps in selecting improved
catalyst combinations to optimize run length, product yields and qualities.