Refiners will face a tough challenge over the coming years
to cope with growing diesel demand at the expense of gasoline,
while conforming to the progressively tighter diesel fuel
specifications to meet environmental regulations. The more
stringent specifications are likely to hurt the overall diesel
yield due to possible rejection of heavy ends from the present
Future diesel will have a narrower density band, lower cloud
point, lower T95 (D86) specification and less sulfur.
Therefore, aligning the existing refineries with progressively tighter diesel
specifications at a sustained diesel yield will be quite
challenging. Some recent compliance project experience is reported here;
the goal is to identify the most cost-effective and appropriate
modifications to an existing plant to achieve future diesel
specifications at a sustained production cost and rate. Project constraints prompted
applying engineering solutions rather than installing more
expensive process technology options that would have
required reconfiguring the existing refinery at high cost.
Finished diesel product in a refinery is a blend of various
distillate-grade streams, as shown in Fig. 1. A blend
optimization is performed with refinery blending simulation
software, utilizing the individual blend streams composition
and associated proportions in the pool for maximum diesel
production at the required market specifications. In a typical
refinery configuration, these blend constituents are produced
from the units, as illustrated in Fig. 2, and include:
Crude distillation unit (CDU).
This is the main source of straight-run (high-cetane) saturated
distillate-grade material to the pool.
Hydrocracker unit. The main source
for cracked but saturated distillate-grade material in a
refinery designed for higher diesel production.
Fluid catalytic cracking unit
(FCCU). This unit produces cracked unsaturated (poor
quality diesel) distillate material as a secondary product. The
intent of the unit is to produce mainly high-quality gasoline
Coker unit. This unit also
produces cracked unsaturated poor-quality distillate-grade
material as a secondary product.
1. Typical product diagram for a
A typical refinery contains either an FCCU or a hydrocracker
unit, depending on whether higher gasoline or diesel production
is targeted. A review of Figs. 1 and 2 identifies the CDU and
vacuum distillation unit (VDU) as potential sources for
improving the diesel production by improving fractionation
performance. In particular, for older refineries, these may
have already been revamped to achieve greater throughput rates.
This improved fractionation may enable the operator to
compensate fully for the possible loss in diesel production due
to tighter future specifications.
2. Process flow diagram of
It is quite challenging to upgrade the atmospheric
distillation column to achieve higher diesel recovery while
maintaining the stream cloud-point specification. In principle,
such upgrades tend to lift more diesel-grade material by
elevating the feed flash-zone temperature or stripping harder
in the CDU bottom section. These require the refiner to
implement several de-bottlenecking options, either individually
Maximizing utilization of crude-feed heater duty
and tactically upgrading the crude preheat train to maximize
the heat recovery to increase the crude column feed flash-zone
Rectifying CDU bottom stripping steam rate to
maximize diesel-grade material stripping from the atmospheric
residue. This may require upgrading the stripping section and
atmospheric gasoil (AGO) wash section (section between feed
flash zone and AGO pumparound) to avoid any possible fouling or
Maximizing diesel recovery by minimizing diesel
fraction overlap with AGO, kerosine and fractions.
Fig. 3 shows a typical configuration of an atmospheric crude
column bottom section. Increasing the CDU feed flash-zone
temperature for higher diesel lift also results in heavy
constituents lift, responsible for diesel cloud-point
specification. A typical overloaded existing crude column
cannot prevent these heavy tails entering into the diesel draw
stream. A similar limitation occurs when the column bottom
stripping steam rate is increased to enhance diesel recovery.
Further, these operational changes increase the columns
pressure drop due to the increased vapor traffic and suppress
the upgrades effectiveness to some extent.
3. Configuration of the atmospheric crude
column bottom section.
Therefore, these operational changes should be accompanied
by some mechanical modifications in a typical crude column
bottom section, as suggested in Fig. 4, to obtain the desired
4. Modified atmospheric crude column bottom
section to increase diesel-product yield (modifications
shown in red).
Controlled and optimum AGO wash rate will prevent heavy
tails carryover into AGO and diesel sections that would
otherwise occur at the increased feed flash-zone temperature
and at the increased stripping steam rate. An optimum AGO wash
rate minimizes valuable AGO loss from the wash section.
Replacing trays with packed beds will overcome the impact of
increased vapor loading on the column pressure drop. The
proposed modifications of Fig. 4 can provide several advantages
over the existing conventional atmospheric crude column:
A controlled AGO wash rate ensures uninterrupted
wetting of wash-section packing needed for trapping the heavy
tails, responsible for off-spec diesel product on cloud
A controlled AGO wash configuration provides
flexibility to optimize the stream flow to minimize valuable
AGO loss from the wash section. In general, the optimum AGO
wash rate is achieved by ensuring that the rate of liquid from
the wash section is approximately 33.5 mass% of the total
A guaranteed performance of the wash section
enables optimizing stripping steam rate in the column bottom
for improving diesel and AGO recovery from the column bottom
stream while maintaining diesel cloud-point specification.
Depending on COT, 10 lb to 20 lb stripping steam per bbl of the
bottom product provides the optimum results.
Replacing trays with packing increases the
number of theoretical stages in the AGO and diesel sections,
which helps to minimize the distillation overlap between AGO and
Able to control diesel under-reflux flow to the
AGO section to control slippage of valuable diesel material in
the AGO fraction.
Increased diesel and AGO recovery improves the
crude preheat-train performance (extra heat available for
exchange) and also reduces feedrate to the VDU. The reduced VDU
feedrate lowers the total ambient heat lost via the vacuum
gasoil (VGO) air coolers, thus improving the overall heat
Reduced crude column bottom rate generates spare
duty in the VDU feed heater and the vacuum section for
minimizing the vacuum residue rate.
The proposed upgrade of the studied tower is able to
increase the diesel yield approximately 1.5 vol% to 2 vol% of
the total crude feedrate. The proposed upgrades have a
potential payback of three years.
Fig. 5 shows a typical dry VDU. A wet VDU (stripping steam
in the column bottom) is almost taken over by dry VDU once high
vacuum generation and low pressure drop internals became
feasible. No stripping steam in the column bottom makes the dry
VDU column smaller in diameter and also requires a smaller
vacuum section as compared to the wet VDU.
5. Typical scheme of a dry
The usual intent of a VDU in refinery is to recover the VGO
fraction from the CDU bottom stream by separating the fuel-oil
grade from the heavy tails. The recovered VGO fractions are
processed further in downstream conversion units into a more
valuable diesel or gasoline products. VGO is coarsely separated
into heavy VGO (HVGO) and light VGO (LVGO) in the column upper
section to provide required quality wash oil (HVGO) to HVGO
wash section and also to minimize the column top section
diameter. Thereafter, LVGO and HVGO combined streams are
usually sent to either the hydrocracker or FCC units for
A typical dry VDU has approximately eight theoretical stages
in total, almost equally distributed in all four sections,
i.e., the slop-oil wash section, HVGO wash section, HVGO
section and LVGO section. The slop-oil wash section is an
optional feature considered in some VDUs; it is used to further
ensure that the heavy metals are retained in the slop oil,
thereby eliminating contamination of the VGO stream.
6. Scheme of a dry VDU with modifications
to improve fractionation performance (modifications shown
Traditionally, VDUs are not designed for sharp separation
between HVGO and LVGO. This can be clearly seen in a
performance data plotted for a typical existing VDU in Fig. 7.
This VDU was receiving lighter than the usual feed due to
inadequate lift in the CDU feed flash zone, caused by the poor
performance of the crude preheat train and a reduced stripping
steam rate in the column bottom to maintain the diesel
side-draw stream cloud-point specification. The reduced
stripping rate had resulted in fouling/coking in the CDU
stripping section, thereby causing further deterioration of the
Poor performance in the CDU could be compensated by
upgrading the downstream VDU for enhanced diesel recovery. For
example, an existing VDU modification can be justified if the
upstream CDU is upgraded for a higher diesel recovery.
Cost-effective modifications of an existing CDU can reduce
diesel loss in the bottom stream significantly but not
completely. Furthermore, a CDU fractionation efficiency cannot
be stretched beyond a certain limit. Therefore, it has been
observed that even a reasonably well-performing existing CDU in
post-upgrade operation can slip diesel-grade material into the
bottoms in sufficient quantity to justify the downstream VDU
upgrades to recover the diesel. This upgrade option will not
only compensate for the diesel-product loss due to the tighter
specifications, but also may improve the refinerys
overall economics by increasing diesel recovery.
7. Performance of the existing
A study was conducted for upgrading a conventional dry VDU for
recovering diesel-grade material from a poorly performing CDU
bottom stream. The proposed modifications shown in Fig. 6
shifted VDU fractionation performance from Figs. 7 and 8. The
improved separation recovered allowed approximately 80% of the
diesel slipping from the CDU. The proposed modifications were
able to recover almost 40 vol% of VDU feed as diesel-grade
material for hydrotreating into a finished diesel product. In
many conventional refinery configurations, this diesel fraction
would have been recovered as VGO and would have been routed to
either the hydrocracker or FCCU for conversion. The proposed
upgrade was able to enhance the overall diesel production by 3
vol%5 vol% of the CDU total feed in the diesel-based
refinery (refinery with hydrocracker unit) and that could be
even greater in a refinery operating with an FCCU. The
estimated payback period for the proposed modifications is two
8. Performance of the existing VDU after
modification to increase diesel make.
The VDU upgrades proposed in Fig. 6 (shown in red) can be
Tripling HVGO section theoretical stages for
sharp separation (reduced overlap) between HVGO and LVGO. In
this scenario, LVGO was essentially the diesel-grade
Hot LVGO under-reflux (P1 new) to HVGO top
section to maximize the diesel recovery in the LVGO section
while maintaining the stream cloud-point specification.
Doubling of the LVGO section theoretical stages
reduces LVGO pumparound (P1) rate by achieving a close
temperature approach between the vapor inlet and the liquid
leaving the LVGO section.
This study shows that future diesel specifications can be
met at sustained production rates, provided that refiners
reorient their distillation operations to improve the diesel
separation in CDU and VDU. This concept will allow refiners to
improve diesel production while also meeting future
specification at the nominal upgrade CAPEX. A similar distillation approach in the
existing residual FCC or hydrocracker fractionation sections
may further improve overall refinery diesel production.
Special thanks to Graeme Cox, Uhde Shedden chief technology and sustainability officer; Patrick
Cadenhouse-Beaty, Uhde Shedden manager of refining technology; and Richard
OBeirne, Uhde Shedden chief process engineer refining, for reviewing and
providing their valuable feedback to shape this article into
its present form.
1 Singh, D. and S. Van
Wagensveld, Redesign crude preheater train for
efficiency, Hydrocarbon Processing, May 2007,
||Dinesh Singh is a principal engineer
with Uhde Shedden (Australia) Pty Ltd. He has over 18
years of worldwide process engineering experience in
petroleum refining, offshore and onshore
oil and gas projects. Mr. Singh holds a BS
degree in chemical engineering from the Indian Institute of Technology, Roorkee, India. He is a recognized
chartered engineer with IChemE UK.