You are on page 1of 6

HYDROCARBON PROCESSING SEPTEMBER 2011

I


59
REFINING DEVELOPMENTS SPECIALREPORT
Achieve success in
gasoline hydrotreating
Case history describes achieving top performance
in FCC gasoline hydrotreater
K. SANGHAVI, Alon USA, Big Spring, Texas; and
J. SCHMIDT, Axens North America, Inc., Houston, Texas
S
uperior FCC gasoline hydrotreating performance is achiev-
able by selecting the optimal process scheme to minimize
octane loss. Enlisting help from a refnery process consultant
(PC) and technology licensor and collaborating early in the design
stage, further ensures the success in determining the better design
for the facility. Consequently, maintaining cost-efective solutions
for a staged project investment and operating the worlds short-
est FCC main fractionator subjected Alon Big Spring Refnery
(BSR) with difcult project challenges. Te roadmap used for a
two-phase project and the lessons learned during Phase I (Interim
Case) contributed to the successful implementation of Phase II
(Ultimate Case). By knowing the key process and operational
principals, the Alons Big Spring new hydrotreater yields world
class performance with an excellent economic advantage.
Case history. In early 2002, Alon, being an owner of a single
refnery in Big Spring, Texas, was granted the status of a small
refner and was initially required to reduce sulfur (S) in refnerys
gasoline pool to less than 300 ppm between 20042009 (Interim
Case) and thereafter the refnery had to meet EPAs ultimate
requirement of less than 30 ppm S (Ultimate Case). Typically, the
refnerys PC would initially lead all process aspects of such a major
project such as determining the process design basis including,
feed analysis, selecting processing scheme and/or process licensor
and setting process scope. Early evaluations revealed that treating
FCC gasoline would be the most optimal investment solution for
the BSR. Of the fve diferent processing schemes available at the
time, the initial study narrowed down the list to three processes for
further study. Ten BSR acquired access to an idle 6,000 bpd (6
Mbpd) straight-run (SR) naphtha hydrotreater (NHT) complete
with a recycle compressor from an adjacent idle reformer.
Consequently, the refnery management asked the PC these
questions loaded with monumental challenges:
a) Can we relocate and revamp the acquired idle equipment
sized for only 6 Mbpd of SR naphtha to a 13.8 Mbpd unit treat-
ing FCC gasoline rich with 36 vol% olefns?
b) Can we decrease FCC gasoline sulfur from 3,000 ppm to
30 ppm with enviably limited octane loss?
c) Can we do all this with an intermediate operation (Interim
Case) with undercut FCC gasoline with 1,650 ppm S1,700 ppm
S and achieve 90% sulfur reduction, to difer capital expenditure
and thus utilize the advantage of being a small refner?
Te PC believed that it can all be done by working with a lot
of due diligence and fduciary responsibility and selecting a game-
changer FCC gasoline hydrotreating process as well as selective
hydrodesulfurization catalyst. Tis task was even more difcult
at BSR as:
Te refnery has the worlds shortest FCC main fractionator,
at only 61 ft in height with 15 trays and two packed-bed sections.
Tus, the FCC gasoline can have some heavy and tough-to-treat
sulfur compounds from the light cycle oil (LCO).
Te semi-regen reformer is the refnerys sole source for
hydrogen, where hydrogen purity varies from 88.6% at start of
run to 74% at the end of the run. When reformer is down, hydro-
gen purity from purchased liquid hydrogen is 99.9%.
FCC hydrodesulfurization principles. Te key to treating
FCC gasoline is in the ability to achieve the required sulfur reduc-
tion while maintaining octane levels. Octane loss results from
hydrosaturation of olefns in the feed during hydrodesulfurization
(HDS) of thiophenes and benzothiophenes in FCC gasoline in
several steps. Both reactions occur in parallel and are shown here:
Olefn + Hydrogen r Parafn
Example: 4-Methyl -2-pentene +H
2
r 2-Methyl-pentane
Tiophene + Hydrogen r Butane + H
2
S
Fig. 1 shows the olefns and sulfur distribution in BSRs FCC
gasoline, with the highest amount of olefns and lowest sulfur
occurring in the front end. Table 1 lists the octane numbers for
TABLE 1. Octane number of olefns and saturated
parafns
Octane numbers RON MON
1-Pentene C
5
olefn 91 77
2-Methyl-2-butene C
5
olefn 97 85
n-Pentane C
5
parafn 62 62
3-Methyl-2-pentene C
6
olefn 97 81
4-Methyl-2-pentene C
6
olefn 99 84
n-Hexane C
6
parafn 25 26
2,2,4-Trimethyl-1-pentene C
8
olefn > 100 86
2,2,4-Trimethyl-2 pentene C
8
olefn > 100 86
n-octane C
8
parafn Minus 19 Minus 15
REFINING DEVELOPMENTS SPECIALREPORT



65
FCC) was used. It required several startup issues to be resolved
and incorporated into the fnal startup procedures. Additionally,
BSR provided detailed training to operations, technical support
and maintenance outlining the fnalized procedures. Color-coded
process fow diagrams for each step with associated operating
parameters were used in training. Te diagrams as part of the
training contributed to the successful start-up.
Results. Post startup audit and an outside review have revealed
that this unit: 1) meets the BSR gasoline pool sulfur specifca-
tions of 30 ppm S and 2) has the best performance amongst
other similar functional competitors units, achieving very low
octane losses in a single-stage unit when processing feed with
high olefn and high sulfur, nominally at 2,1002,400 ppm S,
as shown in Fig. 7.
Te refnery has experienced enviable octane losses as low as
0.30.5. Te refnery PC recently developed an excellent cor -
relation for predicting octane losses as a function of feedrate and
% HDS. Tis helps BSR manage octane losses in the range of
0.70.8 at normal feedrates with 2,300 ppm S and 97.2 % HDS.
Most other typical FCC gasoline hydroprocesses treat feed with
less than 1,300 ppm S and while % HDS is typically less severe,
at less than 96.1%, and still experience octane losses commonly
in the range of 1.41.5 or higher. On this basis, BSR has reached
top of the class in FCC gasoline hydrotreating. Higher S feeds at
BSR is directly due to processing of higher sulfur West Texas sour
crude, providing BSR another great economic advantage over
refneries processing sweet crudes.
Te issue related to high HDS reactor pressure drops has been
efectively resolved with installing feed flters, SHU and macro-
porous media to the HDS reactor grading system, as evidenced by
pressure drop charts for both SHU and HDS reactor, as shown in
0
500
1,000
1,500
2,000
2,500
3,000
3,500
4,000
0 50 100 150 200 250 300 350 400 450 500 550 600
Days since 1st startup
F
e
e
d

s
u
l
f
u
r
,

w
p
p
m
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4,500 4.5
4.0
(
R
+
M
)
/
2
,

l
o
s
s
Feed sulfur (R+M)/2 loss
Feed sulfur and octane loss for Ultimate Case. FIG. 7
TABLE 3. Revised equipment sizes for greater fexibility
Base case Alternative cases
Purity case average Low High
Makeup H
2
purity, % 80.2 74.0 88.6
Recycle gas, MW 8.4 10.56 3.07
Reciprocating compressor, acfm 1,320 1,320
H
2
fow, lb/hr 17,758 22,257 6,678
SHU reactor steam preheater, ft
2
260 603
HDS reactor efuent air cooler, MMBtu/hr 22 42.3
H
2
heater, MMBtu/hr 10.7 13.34
Select 169 at www.Hydrocarbon Processing.com/ RS Select 170 at www.Hydrocarbon Processing.com/ RS
REFINING DEVELOPMENTS
Another unit re-design included a continuous wash-water
injection system due to the extra bay at the reactor efuent air-fn
condensers, which were susceptible to chlorides in the makeup
hydrogen. It also provided the option for a future water-wash
column to minimize amine carryover.
Startup of ultimate operation. In 2009, the BSR started
the revamped Ultimate Case. Te successful startup was contrib-
uted to several key factors:
1) Te technology licensor and BSR inspectors performed a
detailed conformance check of new vessels and trays. Te SHU
and HDS reactor internals were a focal point to ensure proper
installation and levelness.
2) Safe loading of pre-sulfded, pre-activated catalyst, that
does not require in-situ sulfding or activation step, was supervised
by catalysts provider/BSR verifying correct layers and loading
densities.
3) Combined eforts in writing detailed start-up procedures
and complete technical assistance during startup.
4) Around the clock technical support by technology licen-
sor and BSR technical engineers.
Modifed startup procedures were necessary as BSR did not
have the typical feedstock (low olefnic naphtha) required for
startup. A more difcult feedstock (the normal feedstock from
Example of coke buildup on catalyst and the
agglomeration from unstable dienes in feed.
FIG. 5
0
500
1,000
1,500
2,000
2,500
3,000
0 150 300 450 600 7509001,0501,2001,3501,5001,6501,800
Days since 1st startup
S
u
l
f
u
r
,

w
p
p
m
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3,500 3.5
(
R
+
M
)
/
2

l
o
s
s
Feed sulfur (R+M)/2 loss
Feed sulfur and octane loss during the Interim operating
case while meeting 150-ppm S in gasoline.
FIG. 6
Select 168 at www.Hydrocarbon Processing.com/ RS
REFINING DEVELOPMENTS SPECIALREPORT
HYDROCARBON PROCESSING SEPTEMBER 2011

I


61
the need for feed flters, feed pretreatment
with SHU and arsenic guard as a part of the
grading system for the HDS reactor. Table 2
highlights the design feed characteristics for
the Interim and Ultimate Cases.
BSR full-range FCC gasoline has a lon-
ger end-point tail than normal due to its
very short FCC main column. Tis material
was being undercut for the Interim Case
operation. When compared to typical FCC
naphtha feedstocks, the BSR feed proves to
be one of the most difcult with high sul-
fur and olefn content. Te concentration
of dienes, as measured by MAV analysis, is
exceptionally high and resulted in frequent
pressure drop buildup events during the
Interim Case.
Despite the difcult feedstock processed
even during the Interim Case, the results
met BSR product sulfur specifcation with
excellent octane retention. Fig. 6 highlights
the feed sulfur and (R+M)/2 octane loss
during the Interim Case while meeting the
150 ppm S gasoline pool specifcation. Te
higher than design feed sulfur during the Interim Case was the
result of processing higher end-point material, a step closer to the
planned future ultimate case full-range feed. During this period,
there were refnery hydrogen limitations. To conserve hydrogen
in the diesel hydrotreater, LCO make was reduced by increasing
the Interim Case gasoline end point.
New thinking for the ultimate operation. Te ultra-
low-sulfur gasoline (ULSG) requirement of 30-ppm sulfur in the
gasoline pool was required by BSR starting after 2009. To meet
the regulation, the Interim operation was now set to be revamped
to the Ultimate operation. Not only was it necessary for the prod-
uct sulfur to meet requirements but also 1) excellent octane reten-
tion to meet refnery economics and 2) a continuous catalyst cycle
to meet the four-year FCC turnaround schedule. Also during the
Interim operation, the BSR crude capacity increased thus rais-
ing the FCC gasoline rate. Tis required a new study to assess
the impact from a higher feedrate to the HDS section, from the
original Ultimate Case value of 8 Mbpd to 10.8 Mbpd.
A common industry practice is to design the units reactor and
heat transfer equipment including the heater(s) based on a) both
reactors being at the start of the run (SOR) and/or both reactors
being at the end of the run (EOR), in tandem, based on a four-year
run length and b) the average hydrogen purity at 80.2% for BSR.
But during mid-2008 when restarting work for the Ultimate
Case to increase operational fexibility and economic advantage,
the refnerys PC asked that other scenarios be considered in the
design and equipment to cover:
a) Staggered reactor operation, with SHU reactor being at
SOR while HDS reactor continues to run its course and vice
versa, which de-couples the reactors
b) Unit fexibility to cover the expected 74%88.6% hydro-
gen purity as the semi-regen reformer cycle progresses.
Tis revised basis increased sizes for the HDS reactor and
the units heat exchange equipment, as well as the sizes of the
hydrogen heater and reactor efuent air-fn condenser, as shown
in Table 3.
Also the refnerys PC requested adding a macroporous trap-
ping media for scales as a part of the HDS reactor grading system
and using wedges and pins in place of traditional nuts and bolts
for reactor internals, for easier installation and removal. Addition-
ally, due to the arsenic measured on the catalyst during the Interim
operation, a layer of arsenic trap was installed on top of the main
HDS catalyst bed.
TABLE 2. Design feed characteristics for Interim and
Ultimate Cases
Feed properties Undercut full range, FCCN Full range, FCCN
Sulfur, wppm 1,6501,700 3,0694,132
Olefns, vol% 36.0 35.037.0
MAV, mg/g 1522 1221
D-86, vol% F F
10% 110120 114
90% 310355
95% 330375 415420
FBP 380430 450475
Ultimate
case feed
F/E heat
exchangers
SHU reactor
Splitter
LCN
product
F/E heat
exchanger s
H
2
heater
HDS reactor
Recycle
compressor
Product separator
Efuent cooler
Makeup H
2
Amine
contactor
HCN produc t
Stripper
Reboiler
New equipment/pipin g-Phase 2
Revamp equipmen t-Phase 2
Existing equipment/pipin g-Phase 1
Wash water
Liquid quenc h
Interim
case feed
A/B/C D
Purge
Final process design for B SR FCC gasoline revamp. FIG. 3
0
0 150 300 450 600 750 900
Days since 1st startup
1,0501,2001,3501,5001,6501,800
R
e
a
c
t
o
r

P
,

p
s
i
20
40
60
80
Pressure drop due to buildup in the HDS reactor due to
lack of pretreating feed.
FIG. 4
REFINING DEVELOPMENTS SPECIALREPORT
60

I

SEPTEMBER 2011 HydrocarbonProcessing.com
olefns vs. resulting saturated parafns. Fractionation upstream
of the HDS section is an attractive frst step to concentrate the
olefn-rich light-cat gasoline (LCG) as a product and the sulfur-
rich heavy-cat gasoline (HCG) for hydrodesulfurization (HDS).
BSR focused on several essential characteristics and challenges
in selecting a successful process including:
Minimize octane loss. Gasoline is hydrodesulfurized selec-
tively and collateral damage that can occur through olefn satu-
ration is minimized; accordingly, the scheme achieves the total
lower octane loss.
Minimize hydrogen consumption per barrel of feed was
another important consideration for BSR. Olefn and aromatic
preservation is essential; otherwise, a large amount of hydrogen
would be used in saturating these compounds as compared to
desulfurizing them.
Retain excellent gasoline yield with no Rvp increases.
Tis is vital for maximizing product. Tis is attainable with mild
operating conditions that avoid cracking reactions,
Maintain catalyst cycle length inline with the FCC turn-
around schedule to avoid untimely blending issues due to of-
spec FCC gasoline.
Conserve total capital investment to cover both the
Interim and Ultimate Case operations.
Detailed evaluation showed that for BSR, the selected gasoline
hydrotreating processing scheme could meet all of the essential
characteristics for both the Interim and Ultimate requirements.
Fig. 2 outlines the basic process fow diagram.
Selective hydrogenation principles. In the selected
scheme, for the Ultimate Case, the feed would be pretreated in a
selective hydrogenation unit (SHU) to convert lighter mercaptans
and light sulfdes to heavier sulfur species and also to saturate
unstable dienes with no octane loss and minimal hydrogen con-
sumption. Dienes, unless removed through saturation, would
thermally decompose and agglomerate into a coke crust; thereby
accelerating pressure drop buildup in the downstream HDS reac-
tor. Tis would then shorten the units run length.
Pretreated feed would then be fractionated in a splitter to
remove about 29 vol% to 33 vol% of the feed as onspec LCG with
less than 30 ppm sulfur and rich in high-octane olefns.
In most cases, the balance of the feed stream, HCG, would be
hydrodesulfurized to reduce sulfur to below 30 ppm. LCG can be
blended back with HCG. Otherwise, if a separate storage sphere is
available, then the LCG can be segregated for blending fexibility.
BSR chose the former option for LCG. Selectivity of the HDS
catalyst to minimize octane saturation while treating heavier sulfur
compounds in HCG would determine the total octane loss.
Challenges of the I nterim Case. With the idle 6,000 bpd-
SR naphtha hydrotreater available as part of the FCC gasoline
hydrotreater revamp, the frst of many project challenges were pre-
sented. In combination with a minimal investment requirement
for the Interim Case, the challenges increased signifcantly. A joint
efort between BSR and licensor to develop a scheme was initi-
ated to not only minimize investment but to meet the required
HDS level with acceptable octane loss for both the Interim and
Ultimate Cases.
Roadmaps. BSR developed roadmaps for both Interim and
Ultimate Cases so that the least amount of equipment would be
wasteful during the transfer from the Interim to Ultimate process-
ing schemes. Te licensor and BSR worked closely to arrive at the
fnal Interim and Ultimate cases that encompassed the project
challenges and requirements. For the Interim Case, a simpler ini-
tial fow scheme was developed to meet the immediate processing
requirements, while simultaneously considering future require-
ments for the Ultimate Case. Despite the challenges presented, the
design basis for each case was studied, and the technology licensor
provided BSR with the fnal process design package. Both cases
are shown in Fig. 3.
Lessons learned contributed to success. Te Interim Opera-
tion during January 2004 to September 2009 was with full-range
gasoline feed to the HDS reactor without pretreatment by the
SHU. Tis operating mode provided an opportunity to study
features needed for optimal Ultimate Operation. Fig. 4 shows
that the pressure drop buildup in the HDS reactor during Interim
Operation determined the units run length. Te high pressure
drop would require frequent outages to skim the top-bed cata-
lyst or a complete catalyst changeout. Tis was attributed to the
absence of SHU pretreating and the protection it ofers to the
HDS reactor. Te importance of installing an SHU reactor in
the Ultimate Case was further strengthened. With a 30-wppm S
gasoline pool requirement for the Ultimate Case, frequent unit
downtime would jeopardize refnery economics/blending.
Analysis of crusts from the reactor revealed high coke buildup
from thermal decomposition and agglomeration of unstable
dienes in the feed, as shown in Fig. 5. Also, the catalyst deactiva-
tion rate was high during the Interim Operation. Analyses done
on the spent catalysts revealed signifcant arsenic contamination
which was linked to the feed. Te lessons learned confrmed
0
10
20
30
40
50
60
70
80
100 125 150 175 200 225 250 275 300 325 350
Cut end point, F
C
u
m
u
l
a
t
i
v
e

s
u
l
f
u
r
10
15
20
25
30
35
40
45
50
55
60
O
l
e

n
s
,

v
o
l

%
Sulfur
Olens
Cumulative sulfur and olefns distribution vs. cut-end point. FIG. 1
LCN to pool, TAME or alky unit
ULSG
66

REFINING DEVELOPMENTS
Fig. 8 and Table 4. Te HDS reactor is almost close to start of run
temperature after one-half years of operation. Te arsenic con-
tamination of HDS reactor seems to be efectively resolved too.
Successful project. BSR FCC gasoline HDS unit is in a
position to provide the refnery excellent economic advantage
and leverage. It has demonstrated that it will not constrain refn-
ery operations while processing lower-cost sour crude oils that
in turn results in feeds with higher sulfur. Tis can be classed as
an extraordinary achievement, especially for the worlds shortest
FCC main fractionator and restrictions imposed by repurposing
an idle 6 Mbpd NHT and reformer compressor. Intelligent factors
contributing to top of class performance are:
(1) Superior processing scheme, based on saturation of unsta-
ble dienes in a selective hydrogenation unit and separation of
the front-end FCC Gasoline as LCG before HCG is treated in
reactor with selective HDS catalyst. Tis scheme would always
assure process success in terms of superior octane retention and
four-year unit run length.
(2) Early roadmaps prepared for both Interim and Ultimate
Cases ensure minimal wastage of investment.
(3) Implementing lessons learned from the Interim Case into
Ultimate Case design resolved issues related to high reactor pres-
sure drops, catalyst activity, catalyst stability and catalyst arsenic
contamination.
(4) Excellent capability of the refnerys PC to guide the licen-
sor and also for setting right design basis and process direction and
infusing new thinking for a more robust unit. HP
Kirit Sanghavi is senior refnery process engineering consultant at Alons Big
Spring Refnery. He is responsible for the largest capital projects at this refnery. Previ -
ously, Mr. Sanghavi worked at Esso Chemical and Imperial Oil in Canada for 15 years
and for four Engineering Companies in the US, UK and Canada during his career. He
earned a bachelors degree in chemical engineering from London University.
Jeff Schmidt is a senior technical service engineer for Axens North America, Inc.
He has been with the company for the past fve years and is responsible for start-up
and technical support for Axens licensed units. Previous to Axens, he worked at UOP
for fve years. Mr. Schmidt holds a BS degree in mechanical engineering from the
University of Wisconsin-Madison.
SHU DP HDS DP
0 50 100 150 200 250 300 350 400 450 500 550 600
Days since 1st startup
0
4
8
12
16
20
P
r
e
s
s
u
r
e

d
r
o
p
,

p
s
i
Pressure drop across the HDS and SHU. FIG. 8
TABLE 4. Ultimate operation HDS reactor performance
Time 10/23/2009 1/31/2011 5/30/2011
Day onstream 14 460 575
HCG feed, bpd 7,964 7,890 7,925
HCG HDS, % 99.0 98.1 98.5
Normalized P, psi 4.3 6.6 6.9
Deactivation rate < 0.5F/month
Select 171 at www.Hydrocarbon Processing.com/ RS Article copyright 2011 by Gulf Publishing Company. All rights reserved. Printed in the US.
Not to be distributed in electronic or printed form, or posted on a website, without express written permission of copyright holder.

You might also like