Operational flexibility is one of the key
parameters that leads to success and sustainability in refinery production.
Refineries perform many projects based on their needs. For example, preventing and/or
decreasing slop oil is one pathway to achieve sustainable operation. Co-processing
streams and finding alternatives for product streams are two more examples.
In Tupras’ Izmit refinery, co-processing coker
naphtha in the fluid catalytic cracking unit (FCCU) and a proprietary
hydrodesulfurization (HDS) unita presented a possible solution to prevent coker naphtha from
being sent to the slop storage tank. Trials were organized to monitor unit
parameters, yield and product quality effects. Three trials were executed to test
coker naphtha co-processing in the proprietary HDS unit, the FCCU, and in both the
FCC and HDS units simultaneously. While having the HDS unit’s stream
co-processed with FCC gasoline within the battery limit of the unit, coker
naphtha was injected into the bottom of the FCCU’s riser. The coker naphtha
stream was sent to the HDS and FCC units as R/D stream from the delayed coker
unit.
This article focuses on how co-processing coker
naphtha in the FCCU and proprietary HDS unit affects yields, operational
parameters and product quality.
Feed characteristics and unit
performance estimations.
Other than coker naphtha or olefinic
naphthas from the visbreaker and FCCU, naphtha sources can be light naphtha,
natural gas condensates, and hydrocracker and hydrotreater naphthas. Naphtha
cracking in a catalytic environment is selective to propylene at a propylene-to-ethylene
(P/E) ratio of 1:2.4. The P/E ratio provides an indication of propylene
selectivity. Higher severity is crucial for naphtha cracking (i.e., higher
reactor temperatures, catalyst-to-oil ratios, catalyst activity and increased
residence times).
Naphtha
feed can be injected at the lift zone or directly into the reactor bed. All of
these injection points have different effects. If the naphtha injection is done
from the feed nozzles, catalyst-to-oil ratios are increased, leading to reduced
heat and the need for higher severity. In many trials, delta coke and octane
loss were observed if the injection point was chosen as the feed nozzle.
If
the naphtha injection is done at a point below the current feed injection
point, this results in poor naphtha and catalyst mixing. The reactor residence
time can be inadequate, and regenerator catalyst temperatures and coking
problems can be potentially higher. If the injection point is in the reactor
bed, then lower cracking temperatures can be sufficient with increases in the
residence time. However, the method of injection becomes more prominent in this
case. Generally, by utilizing a second riser for the reaction zone, FCCU
revamps can be feasible to process more naphtha and to produce propylene yields
up to 25 wt% vs. propylene yields that can be maximized up to 10 wt% by the use
of ZSM-5.1
The primary active components in the naphtha
recycle are the olefin molecules, which can be cracked to produce propylene and
C4- products in the presence of ZSM-5. The rate of conversion of
aromatics, naphthenes and paraffins, and propylene selectivity is lower for
olefins. Therefore, the recycle stream composition—distillation and paraffins,
isoparaffins, olefins, napthenes, aromatics (PIONA)—has an effect on the
conversion and propylene selectivity. This is why the type of recycle stream is
important, especially when the impact of the recycle on the reactor’s
regenerator heat balance is considered. Recycling a full-range gasoline stream
to produce propylene is not recommended.
Compared
to recycling full-range gasoline, recycling light-cut naphtha (LCN) is
preferred because of its higher olefin concentration. Approximately 10% of the
LCN recycle is converted to propylene, and 40% is converted to C4-
material due to the presence of 25% propylene in the material. Nearly 60%
of the LCN recycle is not converted, and 25% of the LCN is expected to be
converted into dry gas. Clearly, a higher coke requirement is necessary to
process the recycle. The LCN research octane number (RON) and aromatics content
were also expected to increase.
TRIALS
Three
trials were performed to test:
Trial 1: Co-processing in the FCCU. Test
runs were conducted for the following four cases:
During
these trials, coker naphtha was injected at the bottom of the riser, in
addition to fresh feed. It was also possible to use the feed nozzles as
injection points.
There
was no decrease in the feed rate for the four cases. Trials were executed individually,
and samples were taken to be analyzed to conduct mass balance, while
operational conditions were kept at the Base Case. The riser’s acceleration
zone velocity was kept constant during the trials, and the lift steam rate was
adjusted accordingly.
By calculating Solomon yields, an increase in offgas and LPG yields was observed as the vol% of coker naphtha was increased in the feed rate (TABLE 1). By looking at the operational parameters, one significant change was observed in the reactor pressure when yields were increased for offgas and liquefied petroleum gas (LPG) (TABLE 2).
Trial 2: Co-processing in the
proprietary HDS unit.
Coker naphtha was processed in proprietary
HDS units by co-injecting it into FCC gasoline at the battery limit of the unit.
There was no unit feed rate increase during the trials. The following three
cases were conducted:
The
product sulfur target was kept the same during the three trials. Due to the low
octane value of coker naphtha vs. FCC gasoline, the combined feed (FCC gasoline
+ coker naphtha) had lower RON and motor octane number (MON) values vs. the
Base Case.
Because
coker naphtha has a higher sulfur content than FCC gasoline, reactor delta
temperature (∆T) and reactor inlet temperature (RIT) values were
increased during the trials, as shown in FIG. 1. Hydrogen consumption was increased
without causing any limitations, and there was no issue in ∆ pressure (∆P).
∆P increased in parallel with an increase in the reactor feed rate.
Up
to three unit value increases in ∆RON were observed, while the MON value
decreased one unit value after co-processing coker naphtha. The combined feed’s
RON value decreased by approximately two unit values.
Comparing
the Base Case with Cases 1 and 2, the product’s RON value decreased approximately
four unit values, and the product’s MON value decreased only one unit value. The
most important parameters are reactor ∆T values and product RON/MON
qualities. These values are the limiting values to co-process coker naphtha in the
proprietary HDS unit. Having close communication with the planning department,
it has always been advantageous to release the product sulfur target as much as
possible to witness a positive impact on RON, MON and ∆T values. That
way, the amount co-processed in the unit may increase. For long-term
co-processing, it is critical to monitor RIT values closely to observe any metal
deactivation on the catalyst caused by the coker naphtha.
Trial 3: Co-processing in the FCCU and proprietary HDS unit
simultaneously. There is only one rundown pipe
from the coker unit to the proprietary HDS unit. Flows to the units were
adjusted via a control valve. The trial was conducted by adjusting the amount
co-processed in the FCCU, while co-processing a fixed amount of coker naphtha
in the proprietary HDS unit. The results were:
One
drawback was observed during this trial. Because two units were fed by one
rundown stream, it was sometimes difficult to keep the amounts the same due to the
pressure values at the units’ limits.
Takeaway. Refiners
must blend more different components in the units’ feeds every day. The variety
of feeds that can be processed in the plant’s units are a significant driving
force to gain additional profit. FCCUs have a large tolerance for processing
different kinds of feeds (i.e., lube oil extracts, high-value gasoil,
atmospheric residue, vacuum residue, hydrocracker bottoms or different kinds of
naphthas simultaneously).
Although
processing light naphthas can result in propylene increases of up to 10 wt% of
the total volume processed, this is not the case for whole catalytic naphthas (WCNs),
such as coker naphtha. Therefore, the trials’ results showed that there was a
greater increase in dry gas yields vs. propylene yields in the FCCUs. The other
drawback was a decrease in ∆ octane numbers of the proprietary HDS units—up to
4 in RON and 1 in MON—and a decrease in the catalysts’ lifecycle due to
silicone accumulation. However, being able to co-process this stream as rundown
in the proprietary HDS unit and FCCU simultaneously decreased fouling problems
that may have occurred because of gum formation in tanks and slop formation. HP
NOTE
a
The Tupras’ Izmit refinery’s HDS unit licenses Axens’ Prime G+ technology
REFERENCE
Begüm Öztürk Kantar is a Process Superintendent, responsible for the FCC,
DC, Prime G+ and Merox units at the Tupras Izmit refinery. Her previous
experience has included positions as a Process Chief Engineer and Process Engineer.
She earned a BS degree in chemical and biological engineering from Koç University. In
addition, she is currently working toward an MS degree in software engineering at
Boğaziçi University.
Miray Genç is a Senior Process
Specialist, responsible for troubleshooting, process, operations, catalyst,
additive changes, revamp and design projects for the FCC, DC, VSB, Prime G+ and
Merox units at Tupras’ refineries. Previously, she worked as a Process
Superintendent and as a chemical engineer at the Turkish Council of Research
and Development/Middle East Technical University (METU) collaboration. She earned
a BS degree in chemical engineering, along with an MS degree in polymer science
and technology, from METU.
Ozan Tunca is a Process Engineer,
responsible for monitoring, troubleshooting and optimizing Prime G+ and FCC
units in Tupras’ Izmit refinery. Previously, he worked as a process safety
engineer in the oil and gas and petrochemical industries. He earned a BS degree
in chemical engineering from Yildiz Technical University.
Zeynep Azizoğlu is a Process Chief
Engineer, responsible for troubleshooting, monitoring and processing of the KTU
and LPG Merox units at the Tupras’ Izmit refinery. She previously worked at TU
Delft as a guest researcher. She earned a BS degree in chemical engineering and
materials engineering from Istanbul Technical University (ITU), and an MS
degree in materials engineering and nanotechnology from Politecnico di Milano.
Mehmet Fatih Önen is an Operations Supervisor, responsible for the daily operations at the Tupras’ Izmit refinery, including maintenance of units; optimization of production; and troubleshooting and monitoring of FCC and LPG Merox units, caustic neutralization, and CD and VD units. He earned BS and MS degrees in chemical engineering from ITU.