Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
CA 02446889 2003-10-27
A METHOD FOR CONVERTING A LIQUID FEED MATERIAL INTO A VAPOR PHASE
PRODUCT
FIELD OF INVENTION
A method for converting a liquid feed material into a vapor phase product
using a
cross-flow fluid bed.
BACKGROUND OF THE INVENTION
A number of processes for upgrading heavy hydrocarbons involve spraying a
liquid feed onto a fluid bed. Fluid beds are well suited for providing the
heat and mass transfer
requirements associated with the thermal cracking reactions of the upgrading
process. At lower
gas velocities, the fluidized bed will have a free surface. In this operating
regime the bed exhibits
fluid behavior. As the gas velocity is increased, the fast fluidization or
dilute phase transport
regime is approached and the particles become entrained in the gas. In this
regime, there is no
identifiable free surface.
Examples of fluid bed processes which have been used for the upgrading of
heavy
hydrocarbons include riser coking, fluid coking and LR coking. All of these
processes are
directed at leveraging the inherent advantages of a fluid bed reactor,
including but not limited to
heat and mass transfer, to create a process that maximizes the value of the
products created from
a heavy hydrocarbon feedstock. Some of these technologies operate in the fast
fluidization or
dilute transport regime, while others make use of a low velocity bed of
fluidized solids which has
a free surface.
In general, with any heavy hydrocarbon upgrading process, one goal is to
promote
conditions that allow molecules contained in the feed to react to the point of
becoming a valuable
product then stop reacting. Ideally, the resulting products of the primary
cracking reaction are
withdrawn and quenched before any subsequent reactions can take place. A
second is to provide
all molecules contained in the feed sufficient time at reactor conditions so
that they can react
-1-
CA 02446889 2003-10-27
fully and all of the potential product has escaped into the vapor phase of the
reactor. All designs
require a compromise between these two goals.
For processes in which liquid hydrocarbon is sprayed onto hot fluid bed
particles
that provide the heat for reaction the ideal sequence of events would be as
follows:
1. a liquid feed droplet comes into contact with a solid fluid bed particle;
2. molecules in the feed droplet begin to thermally crack;
3. products of the thermal cracking reaction that evolve as vapor from the
surface of
the heat carrier (i.e., the fluid bed particles), are swept away by the
fluidizing gas,
and are quenched before subsequent reactions can take place; and
4. the residual liquid remaining on the heat carners from which no product can
be
derived is provided a residence time equal to that required to form dry solid.
Failure to cater to item four Leads to operational difficulties related to the
stickiness imparted by the residual reactant hydrocarbon. The current art for
spraying a heavy
liquid hydrocarbon onto a fluid bed represent attempts at achieving this
idealized mode of
operation that have not been entirely successful.
A schematic depicting a Fluid Coking reactor is shown in Figure 1. Fluid
Coking
is a low-velocity fluid bed process where hot solids enter the reactor in the
freeboard region
above the surface of the fluid bed, near the top of the reactor. Feed is
injected into the fluid bed
at several different elevations. Solids are generally well mixed within this
feed zone. >3efore
exiting the reactor, the solids pass through a stripping zone that is designed
to recover unreacted
liquid from the heat carriers. The stripping zone is designed to produce a
solids residence time
distribution (RTD) that is closer to plug flow. The additional residence time
provided to the
solids in the stripping section, together with stripping steam, allows the
recovery of additional
product from the surface of the fluidized particles. Solids exit the reactor
after passing through
this stripping zone. The solids RTD of the Fluid Coking process in the feed
zone, which is well
mixed, has an important impact on achieving the two primary goals of the
process as set out
above.
_2_
CA 02446889 2003-10-27
With a solid's RTD approaching that of a continuous stirred tank reactor
(CSTR),
it is necessary to tailor the designed solids holdup (as it relates to
minimizing the loss of
unconverted feed and production of "dry" solid particles) to the portion of
solids that exit the
reactor first. As a result, a fraction of the solid particles remain at
reactor conditions much
longer than is necessary to achieve the goals of the process. Solids with a
plug flow residence
time distribution would be less affected by this.
Practical considerations associated with the design of fluid coking reactors
require
that significant bed depth be used in combination with increasing vessel
diameter to provide the
design solids holdup. Due to the increased bed height required to meet the
solids holdup
requirement, the vapor phase residence time is increased which may cause
product vapors to
react further, contrary to the first goal of the process
It is noted that although the stripping zone of the Fluid Coking reactor
attempts to
create a plug flow condition for the solids, there is a requirement that the
solid particles enter the
stripping zone of the reactor with little unconverted feed present. Failure to
meet this condition
results in fouling of the stripper internals by the wet solids and circulation
of solid particles is
impaired. The stripping zone therefore does little to mitigate the need for a
Fluid Coking reactor
to be designed with an average solids holdup that is much greater than what
would be required if
solid particles flowed through the entire reactor under plug flow conditions.
It is also noted that for all intents and purposes of this application, a
Flexi-Coking
reactor differs little from a Fluid Coking reactor.
A schematic depicting riser, or transfer line coking is given in Figure 2. The
fluid
bed for these processes is typically characterized as fast-fluidization or
dilute phase transport.
The solid carrier is contacted with feed at one end of the riser or pipe and
transported to the
opposite end of the pipe at a speed equal to the speed of the gas phase less
the slip velocity
between the solid and gas phases. Vapor and solid phase RTDs are closely
linked in this type of
reactor. Short vapor phase residence times can be achieved but because of the
link between
_3_
.. CA 02446889 2003-10-27
solids and gas phase RTDs, the time required to ensure the feed has completely
reacted is
compromised.
A schematic depicting an LR coking reactor is given in Figure 3. In LR Coking,
the force required to counter gravity and fluidize the solid particles is
imparted mechanically
through rotation of the auger-like internal of the reactor. The solids RTI~
approaches plug flow
and the absolute vapor phase residence time is relatively short. High capital
costs, and limited
throughput relative to other fluid bed processes, have practical design
implications that limit the
ratio of solids to feed that may be attained using LR coking processes. In LR
coking processes,
as in other fluid bed processes, reactor temperatures must typically be higher
since the solids-to-
oil ratios are lower. As a result, the high reactor temperatures of an LR
coking reactor relative to
other fluid bed processes and the high capital costs associated with LR coking
processes offset
the benefits of the relatively short absolute vapor phase residence times.
SU1VIMARY OF THE INVENTION
The present invention is directed at a method for converting a liquid feed
material
into a vapor phase product using a cross-flow fluidized bed. The present
invention is also
directed at an apparatus comprising a cross flow fluid bed reactor.
The liquid feed material may be any suitable material. In a preferred
embodiment, the liquid feed material is comprised of a heavy liquid
hydrocarbon. The vapor
phase product may be comprised of a single product, or substance, or may be
comprised of a
plurality of products or substances. The term "vapor phase product" as used
herein means that
the product is or behaves as a vapor phase under the conditions of the
conversion method,
although the product may ultimately be condensable to a liquid phase or even a
solid phase.
In one aspect the invention is a method for converting a liquid feed material
into a
vapor phase product comprising the following steps:
_q._
CA 02446889 2003-10-27
(a) providing a fluid bed comprising solid particles and a fluidizing medium,
wherein
the fluidizing medium is moving in a substantially vertical fluidizing
direction;
(b) moving the solid particles in a substantially horizontal solid transport
direction;
(c) contacting the liquid feed material with the solid particles in order to
convert the
liquid feed material into the vapor phase product;
(d) collecting the vapor phase product in a vapor collection apparatus; and
{e) collecting the solid particles in a solid collection apparatus.
In a second preferred aspect of the invention, a fluidizing medium such as a
gas is
introduced into a reactor to fluidize a bed of solid particles such that the
fluidizing medium is
moving in a substantially vertical fluidizing direction. The solid particles
are transported
substantially horizontally in a solid transport direction from a solids inlet
at one end of the
reactor to a solids outlet at the opposite end of the reactor, preferably but
not necessarily by the
force of gravity. As the solid particles move through the reactor they are
contacted by a liquid
feed material comprising a liquid hydrocarbon. The solid particles are at a
temperature which
facilitates the reaction of the liquid hydrocarbon to produce one or more
upgraded hydrocarbon
products as a vapor phase product. The vapor phase product is collected in a
vapor collection
apparatus, preferably with the fluidizing medium. The solid particles are
collected in a solid
collection apparatus and are preferably regenerated for re-use.
The selection and design of the solid particles, vapor collection system,
freeboard
and fluidizing mechanism may be made so that vapor phase residence time is
short relative to
competing technologies and so that the residence time distribution of the
solid particles
approaches plug flow conditions despite significant evolution of product
within the fluid bed.
The invention permits relatively high ratios of solids to liquid feed, which
aids in achieving
lower reactor temperatures.
-5-
CA 02446889 2003-10-27
For example, whereas there are significant constraints on the flow of solid
particles a single LR coking reactor can process, arid whereas significant
costs associated with
LR coking require that the process adopt relatively low solids-to-oil ratios,
the amount of solid
particles a cross flow coking reactor can process is virtually limitless. This
feature allows the
S invention to employ higher solids-to-oil ratios than may be employed with
some competing
processes, such as the LR coking processes.
Furthermore, whereas the LR coking process is forced to adopt a relatively
high
operating temperature to compensate for low solids-to-oil ratios, no similar
requirement exists
for the current invention. In the practice of the invention a relatively high
solids-to-oil ratio is
used with feed and product recovery zones that are staged such that solid
particle residence times
may be tightly controlled.
BRIEF DESCRIPTION OF DRAWINGS
Embodiments of the invention will now be described with reference to the
accompanying drawings, in which:
Figure 1 is a schematic drawing of a Fluid Coking reactor;
Figure 2 is a schematic drawing of a riser, or transfer line coking reactor;
Figure 3 is a schematic drawing of an LR coking reactor;
Figure 4 is a schematic drawing of a cross-flow reactor according to a
preferred
embodiment of the present invention;
Figure 5 is an alternate schematic drawing of a cross-flow reactor according
to a
preferred embodiment of the present invention depicting spraying of the liquid
feed material
within the fluid bed.
CA 02446889 2003-10-27
DETAILED DESCRIPTION OF THE INVENTION
In general, the present invention relates to a method or process for
converting a
liquid feed material into a vapor phase product. In a preferred embodiment,
the present invention
relates to a method for converting a heavy hydrocarbon feedstock material into
value added
reaction products. The method or process of the invention is herein referred
to as "cross flow
conversion" process or "XFC" process.
The central process unit in the XFC design is a cross-flow fluidized bed
reactor.
As in most fluidized bed processes, a fiuidizing medium, preferably a gas, is
introduced into the
bottom of the reactor base and exits at the top of the reactor so that the
fluidizing medium moves
in a substantially vertical fluidizing direction.
A significant difference between the XFC design and a conventional fluid bed
process is that the solid particles in the fluid bed move substantially
perpendicularly to the gas
phase in the fluid bed. Solid particles enter at one end, flow along the
length of the reactor under
the influence of gravity, and are removed at the opposite end. Since the
solids and gas flows are
generated by independent driving forces, the two are essentially independent.
This provides for a
significant increase in flexibility, which will be discussed in detail in the
description that follows.
This description teaches a method for designing an XFC reactor to produce a
solid
particle RTD that approaches plug flow, allowing for evolution of a vapor
phase product within
the fluid bed. This is an important feature of the present invention. Eenefits
accruing from this
solid particle RTD together with those inherent in the type of fluidized bed
reactor proposed can
be leveraged by a person skilled in the art to provide significant advantages
over the current art.
For example, it is well understood by those individuals skilled in the art how
to
manipulate operating and design conditions such as increased solids-to-feed
ratios and the ability
to deliver feed in a more controlled and uniform fashion to enhance
operability and yield at
typical reaction temperatures. The hydrodynamics of the XFC reactor have been
studied with
CA 02446889 2003-10-27
cold flow physical models, using dimensional analysis to establish a tie to
typical process
operating conditions.
1. XFC Reactor Vessel
The XFC reactor is divided into a number of zones, each having a different
function:
1. Solid feed zone
2. Liquid feed zone
3. Reaction zone
4. Solid withdrawal
zone
5. Gas distribution
zone
6. Freeboard zone
Figure 4 and Figure 5 both depict a schematic which demonstrates the different
zones of the XFC reactor.
The length of the reactor vessel is typically greater than its width. This
design
feature ensures the solids are well mixed across the width of the vessel, and
helps to maintain
plug flow characteristics in the moving solid phase. The impact of plug flow
on the
characteristics of the process is described below.
Gas is introduced as a fluidizing medium through a distributor located on the
bottom of the reactor vessel. The gas distributor can vary in complexity.
Bubble cap and
perforated plate designs have been tested, but any design capable of
adequately fluidizing the
solids is acceptable. The fluidization gas, along with any product vapor
generated by the
xeaction, will typically exit at the top of the reactor vessel.
_g_
p CA 02446889 2003-10-27
The height of the reactor vessel is designed to accommodate both the fluidized
bed contained in the vessel and the height required for solids disengagement
in the freeboard
region (see below).
To provide effective contact between liquid feed and solid heat earner and to
take
advantage of high solids-to-feed ratios it will generally make sense to
provide an amount of solid
particles substantially in excess of what is required for the given feed zone.
By staging several
of the units depicted in Figure 4 in series, the bulk of the solid particles
will be contacted in a
more uniform fashion. To increase capacity, the width of the reactor can be
increased, an option
not available in many commercial configurations.
2. Solids Particle Characteristics
The solid particles in the XFC reactor provide the surface area upon which the
conversion reaction occurs. In addition, the solid particles provide a heat
source or sink for the
reaction, depending upon whether the reaction is endothermic or exothermic.
They may also
possess catalytic activity, although this is not a requirement. The most
critical attribute is that
the particles fluidize well. Based on the Geldart classification (Kunni D. and
Levenspiel, O.
Fluidization Engineering Zed. Butterworth-Heinemann 1991), only the following
two types of
particles are suitable for the XFC reactor:
1. Geldart A type particle: Aeratable particles or materials having small mean
particles size (<40 microns) or low particle density (<1400 kg/m3). Fluidized
Cracking Catalyst is an example of particles of this type.
2. Geldart B type particles: Most particles of size 40 microns to 500 microns
and
density 1400 kg/m3 to 4000 kg/m3. Sand is an example of this type of particle.
These two particle types characterize the typical particles used in industrial
fluidized beds. When fluidized, they provide the positive characteristics that
are most often
associated with fluidized bed reactors: uniform temperature, high rates of
heat and mass transfer,
and high specific surface area. In addition, Geldart A and B particles will be
fluid enough to
allow for smooth horizontal flow.
-9-
CA 02446889 2003-10-27
All remaining particles fall into either the Geldart C (cohesive powders) or
Geldart D (large coarse particles) classifications and are not typically
suitable for use in the XFC
process unless they make up a relatively small fraction (<10%) of the
particles, with the majority
being either Geldart A or Geldart B.
The other factors to consider when choosing the solid particle material are
the
heat storage and transfer characteristics, attrition rates and cost.
3. Bed Characteristics
The fluidized bed will preferably be operated in the bubbling bed regime or,
in the
case of Geldart A particles, may be operated in the smooth fluidization regime
below the
bubbling fluidization velocity but above the minimum fluidization velocity.
In the bubbling bed regime the fluidized bed resembles a boiling liquid with
bubbles forming at the gas distributor, rising through the bed quickly then
bursting at the surface
of the bed. For descriptive purposes the fluid bed can be thought to have two
phasesn
1. Emulsion phase containing both solids and gas
2. Bubble phase containing primarily only gas
Gas exits the bed almost exclusively in the bubbles. Gas in the emulsion phase
must therefore first enter the bubbles in order to exit the fluid bed. The
transfer of gas between
the bubbles and the emulsion can occur by diffusion in the bed, or by mixing
in the turbulent
region in the vicinity of the gas distributor.
4. Freeboard Region
The freeboard is the solids lean region of the reactor vessel above the
surface of
the fluidized bed. Solids are ejected from the fluidized bed by the action of
bubbles bursting at its
- 10-
CA 02446889 2003-10-27
surface. The freeboard region is required for the solid particles to disengage
from the gas so that
they are not carried out of the reactor vessel.
The optimum freeboard height is that which allows all of the solid particles
with
terminal velocities less than the superficial gas velocity to disengage.
Extending the freeboard
above this height will not reduce the solids carryover and will only add to
the cost of the vessel
and to the residence time of the gas.
Even for a very large freeboard region, solid particles will be earned out of
the
reactor because they have been entrained in the gas or because of large
eruptions of bubbles at
the surface of the bed which can potentially eject solids to the roof of the
reactor vessel. If the
downstream gas processing units can not tolerate the presence of solids then a
unit must be
installed to separate the solids from the gas stream. Proven technologies,
such as cyclones, will
be sufficient for this purpose.
Reducing the height of the freeboard region will reduce the residence time of
the
gas phase, which will in turn limit the severity of the gas phase reactions.
However, inadequate
height of the freeboard region can result in an excessive amount of solids
carryover, requiring
larger solids handling units outside the reactor to separate the solids from
the gas.
The optimum freeboard height will be dependent on the type of particles, the
fluidization velocity and the effects of the feed on the cohesive forces
between particles. The
residence time distribution of the gas in this zone has been shown to be
substantially plug flow.
5. Fluidization Velocity
The solid particles are fluidized by the gas that enters through the gas
distributor
at the base of the vessel. The velocity of the fluidizing gas can be between
the minimum
fluidization velocity and the turbulent fluidization velocity of the solids.
If the gas velocity is
below the minimum fluidization velocity of the particles, then the bed will
not fluidize and the
solids will not flow across the bed. At fluidization velocities larger than
the turbulent tluidization
-11-
CA 02446889 2003-10-27
velocity, the carryover of solids will be too great for a solids handling
system of a reasonable
size.
The range of gas velocities that would function in the bed for Geldart B and
Geldart A particles is from approximately 0.01 m/s to 1 m/s. there the liquid
feed is viscous, a
safety margin should be added to the operating fluidization velocity to manage
agglomeration of
the wet particles.
The fluidization velocity has an impact on many characteristics of the
reactor. As
the velocity is increased the gas phase residence time will decreases, but the
concentration and
height of solid particles in the freeboard region will increase. Solid mixing
also increases within
the bed when the fluidization velocity is increased. This reduces the plug-
flow nature of the
solids flow, but increases the resistance of the bed to defluidization, which
is of concern when
processing a viscous liquid feed. All of these factors must be considered when
choosing the
fluidization velocity.
6. Solids Throu~h~ut
Solids particles will be fed into one end of the reactor and withdrawn at the
opposite end. The solid particles preferably flow in a substantially
horizontal direction. The
fluidized solids behave hydrodynamically like a continuous fluid and can be
made to flow across
the bed under the influence of gravity. This flow could be simply induced by
the difference in
bed depth caused by feeding the solids into one end, or by tilting the reactor
vessel in the
direction of flow. Tilting the reactor has the advantage of maintaining a more
uniform bed depth
and allows for greater solids flowrates. In either case, the depth of the bed
may be maintained
through the use of a weir at the solids withdrawal end.
The solids flux through the reactor will likely be the primary factor in
determining
the capacity of the reactor to accept liquid feed. This will be the case when
the heat or surface
area requirements of the reaction are limiting. If required, it is possible to
increase the mass flow
of solid particles through the reactor at a constant flux by increasing the
cross-section of the bed.
-12-
CA 02446889 2003-10-27
7. Liquid Feed Delivery
The liquid feed material is sprayed onto the bed of fluidized solids using
conventional nozzles. The zone of the bed used to accept the atomized feed
will be jusl; after the
solid feed zone.
The feed system should maximize the distribution of liquid feed over the
solids
particles that pass through the feed zone. The optimum situation would be to
have every droplet
of feed hit and engulf a different solid particle. This would maximize the
surface area over which
the reaction occurs which reduces any mass transfer limitations.
To maximize the spread of liquid feed and minimize mass transfer limitations,
the
droplet size should preferably be less than or equal to the solid particle
size, which will allow the
I S droplet to form a thin film over the solid particle. This will be limited
by the wetting properties
of the solid and the liquid. If the feed droplets are too large they can
potentially cause the
agglomeration of the bed solids, and if they are too small they can be
entrained in the rising gas.
The feed nozzles are preferable oriented so that the feed material is sprayed
in a
spraying direction which is substantially perpendicular to the solid transport
direction. The feed
nozzles can be oriented vertically, pointing downward through the surface of
the bed.
Alternatively, the nozzles can be oriented horizontally, through the walls of
the vessel or in
through the base of the vessel. There is an advantage to spraying the feed
into the bottom of the
fluidized bed since this is an area of good mixing between the bed solids and
the gas. For any
nozzle orientation, the aim is to penetrate the bed with feed without
impacting the bottom or
sides of the vessel.
8. Bed D~th
A shallow bed has the advantages of reduced gas phase residence time,
increased
gas solid contacting, reduced axial solids mixing and a reduced concentration
of solids in the
-13-
CA 02446889 2003-10-27
freeboard region. All of these effects will be advantageous for most of the
reaction systems that
will operate in the cross-flow reactor.
There are two operational issues that will determine the minimum depth of the
fluidized bed in the reactor:
1. The required solids throughput
2. The liquid feed jet penetration
The maximum solids throughput is dependent upon the maximum horizontal
solids velocity and the bed cross-section perpendicular to the flow. ~Nhile
operating with a
smaller bed depth can have many advantages (see below), reducing the bed depth
will reduce the
solid particle capacity of the reactor.
The liquid feed nozzles should deliver the liquid feed material to the fluid
bed
without creating liquid droplets of a size that will be entrained in the
upward rnovin g gas. To
accomplish this, adequate momentum is imparted to the feed droplets to allow
some penetration
of the liquid into the fluidized bed. The bed should be deep enough relative
to the momentum
imparted to the feed droplets so that the feed material does not impact on the
base of the gas
distributor. This limit on bed depth can be avoided if the liquid feed
material is sprayed
horizontally into the fluidized bed. This will then place the constraint on
the minimum bed width
in order to avoid the feed material impacting upon the sides of the reactor
vessel. Through
adequate design, the feed delivery system can be engineered to provide the
required
performance.
9. Temperature
The temperature of the reactor will be dependent upon the requirements of the
reaction. The temperature drop across the reactor will be dependent on the
heat requirements of
the reaction and the heat capacity and mass flow of solids.
-14-
CA 02446889 2003-10-27
10. Pressure
Slight positive pressure (0.5-10 psig) is desirable in that there is expense
involved
with providing fluidizing gas which, at constant superficial velocity, will
decrease as the pressure
is reduced. In addition, the rectangular shape of the reactor is less suited
to pressure containment
than cylindrical designs which again make low operating pressures desirable.
Downstream gas
processing requirements will likely set the Iower boundary for system
pressure.
11. Process advantages
Via) Approach to PIu~ Flow of Solid Phase Residence Time Distribution
The assumption of plug flow is an ideal case where every solid particlf; has
the
same axial velocity. The solid particles move along the length of the reactor
in uniform plugs
that are well mixed in the radial direction. Since every particle has the same
horizontal velocity,
there can be no mixing along the length of the reactor. In the current process
of the invention,
the solids RTD approaches the plug flow ideal since the bulk rate of solids
flow along the length
of the bed is much larger than the rate of solids mixing in the same
direction. In engineering
terms, this is equivalent to stating that the Peclet (Pe) number is relatively
large.
The plug flow characteristics of the solid phase takes greater advantage of
the
reactor volume than a fluidized bed that is well mixed. This is because the
residence time
distribution of the solid particles is much narrower than in a fiuidized bed
reactor that is well
mixed. This allows for many advantages all of which are related to the narrow
residence time
distribution:
1. Greater capacity in a smaller reactor, thus reducing capital costs;
2. Larger ratio of dry solids to feed material, allowing more feed material to
be
added to the bed without risking agglomeration of the bed solids, a condition
known as "bogging". In many cases, this allows the reactor to be operated at
-15-
CA 02446889 2003-10-27
lower temperatures, as more severe conditions are not required to address
bogging
issues';
3. Reduced loss of liquid feed material by preventing solid particles from
short
circuiting through reactor vessel. In well mixed reactors, short circuiting of
wetted solids force operation at higher reaction severity;
4. Reduced incidence of solid particles with excessive residence times
(seventies).
(b) Flow Characteristics of the Gas Phase in the Freeboard Re ion
The gas exiting the free surface of the fhuidized bed has also been shown to
exhibit substantially plug flow characteristics, with relatively little mixing
in the direction of
flow. As a result, the time that the gas spends under reactor conditions is
minimi:~ed, and
subsequent reactions that can downgrade the vapor phase product are minimized.
A further benefit to tr.e plug flow nature of the gas is that the product gas
streams
generated at the various locations along the length of the reactor can bc~
collected independently.
This allows for tailoring of the gas recovery system for different gas flow-
rates and allows the
downstream gas processing units to 1=re tailored for different gas
compositions.
(c) Substantial I~ecoupling of Gas and Solid Phase Residence Times
The cross-flow design allows the residence times of the solid phase and the
gas
phase to be adjusted independently. The solid phase residence time is set by
the solid particle
bulk horizontal velocity and the reactor length. The gas phase residence time
is ccmtrolled
primarily by the bed depth and th~,e fhuidization velocity. This allows for
the independent
optimization of the gas phase and solids phase, and hence independent control
of the reaction
severities associated with the gas and liquid phases, offering a significant
advantage over
technologies based on dilute transport.
-I6-
CA 02446889 2003-10-27
~d, High Rate of Vertical Mixing in the Bed
The high rate of vertical solids mixing in the fluidized bed increases the
efficiency
with which the feed material is distributed throughout the bed solids. This
attribute is dramatic
when compared to other technologies that incorporate a moving bed of non-
fluidized particles.
The high rate of vertical mixing in the fluidized bed allows for a deeper bed
as opposed to a non-
fluidized bed which must be made shallow. Furthermore, the: superior feed
distribution has a
positive impact on product formation in cases were mass transfer through the
reacting liquid
phase is in issue, as the liquid films t:hicknesses are kept to a minimum.
(e) Amenability to Small Scale Field Application
The XFC process is rwell suited for scale down. Therefore, it can be used in
field
applications to process relatively small volumes of feed material, on the
order of 1000 barrels per
day.
(f) Ability to Design Solids Holdu~Without Affecting Vapor Please Severity
When a viscous liquid feed material is sprayed into a fluidized bed, there is
the
danger that the particles will agglomerate, and the bed will defluidize. This
condition is referred
to as "bogging". The tendency for bogging can be addressed '~y increasing the
amount of dry
solid particles onto which the feed is introduced. This parameter is fixed
during the design
phase. With a well-mixed reactor, increased solid particle circulation
necessitates increased
reactor residence time to accommodate additional product short circuiting. Due
to mechanical
and other practical constraints, it is not possible to increase reactor volume
without increasing
the reactor height. As a result, increasing solid particle circulation
increases the gas phase
residence time, and hence product losses through the increase in reactor
severity. T:he XFC
reactor does not suffer from these design issues, as solid particle throughput
can be increased at
constant bed height by increasing the width of the reactor.
- 17-
CA 02446889 2003-10-27
Once a reactor is in operation, avoiding bogging in a conventional well-mixed
Fluid Coker is usually accomplished by increasing severity, which ensures that
the liquid quickly
reacts to completion. This operating strategy can be undesirable for many
reactions, as the
products may be degraded to less valuable chemicals under the increased
severity.
fig) Discrete Zones with Ability to T:~ilor Fluidizing Gas to These Zones
Because feed zones and product recovery zones are distinct, fluidizing gas
rates
and fluid bed properties can be tailored to the requirements of the specific
zones. For example,
to manage bogging, more fluidizing; gas can be used in the region where the
fluid bed accepts
feed.
~h) Fluid Bed Volume Near gas entry point is a Significant Fraction of Total
Volume
It has been shown that exchange of product vapour from emulsion to bubble
phase is much better near the gas entry point a fluid bed reactor than
anywhere else in the bed.
This region is often termed the "grid zone". Given that the proportion of the
fluid bed of the XFC
reactor in this region is much higher than current art this provides an
advantage to the quick
evolution of product vapour.
12. Application Case 1-Application to the Uneradin~ of Heavy Oil
The following description describes the specific application of the XFC
process to
the upgrading of heavy oil, such as t~.thabasca bitumen. In this application,
four reactor units are
used in series (i.e. the solid particles will flow from one reactor into the
next). Each unit has
been designed to process feed at a rate of 250 bbl/day making the total
capacity 1000 bbl/day.
The design specifications and operating conditions for this application are
listed in Table 1. The
specifications listed in Table 1 are for a single reactor unit, and are based
on an extensive
piloting exercise designed for this purpose. The following is a brief
explanation of the rationale
behind the numbers in Table 1.
-I8-
CA 02446889 2003-10-27
Table 1
Operating Condition and Design Ranges for a Single Reactor Unit
Specific to Bitumen where the Goal is to Maxirruze Liquid Yield.
Broadest Preferred Optimal
Vessel Length3m - 6m 4m - Sm 4.2rn
Vessel Height2m - 7m 3rn - 4m 3.7Sm
Vessel WidthO.Sm - 2m 0.75m -1.Sm 0.75m
Solid Particle50 microns - 400 120 microns - 240 microns
250
Size Coke microns microns
coated sand)
Bed Depth 0.5 m - 2m O.Sm - 2m 0.75m
Freeboard lm - Sm 2m -3m 2.Sm
Height
Reaction 2.Sm - 3.Sm 2.Sm - 3.5m 3.2m
Zone
Feed Zone 0.5 m - lm O.Sm - 1 m 1 m
Superficial0.1 m/s - 0.7 0.2 mls - 0.6 0.45 m/s
mls m/s
Gas Velocity
Solids Bulk0.02 m/s - 0.2 0.05 m/s - 0.15 0.1 m/s
m/s m/s
Horizontal
Velocity
Mean Reactor440C - 540C 460C-500C 484C
Temperature
The solids particles selected for this application are sand particles with a
mean
particle size between 150 and 250 microns. During operation a coke layer will
form on the sand
particles. The average coke layer will be between 10 and 40 microns thick.
This will increase
both the mean and variance of the particle size distribution. Small particles
will also be formed
by the attrition of larger particles.
- 19-
4 CA 02446889 2003-10-27
The freeboard height is 2.Sm. This is larger than the optimal freeboard
height, but
will ensure that solids carryover is kept to a minimum.
The fluidized bed depth for the bitumen feed will be between 0.75 and 1.25 m.
This brings the total required vessel height to 3-4 m. The bed depth was set
by considering the
gas phase residence time while still maintaining a sufficient reactor cross
section for solids
throughput.
Due to the viscous nature of the feedstock a minimum fluidization velocity of
0.2
mls is provided to maintain proper fluidization.
In this application, the horizontal solid particle velocity through the
reactor will
be between 0.05 and 0.15 m/s. This velocity is based on the heat and surface
area requirements
of the system. It may be necessary to tilt the bed in the direction of solids
flow to achieve the
required bulk horizontal velocity.
To produce the desired reaction, the minimum operating temperature of the
reactor will be about 460°C. Solid feed temperatures to the reactor
will be 490°C. A heat balance
for the overall system indicates that the temperature drop across each reactor
unit will be about
12°C making the mean reactor temperature 4~4°C. This application
of the XFC operates at
moderate pressures (5 psig).
A partial oxidation gasifier may be used to provide heat to the reactor. This
technology is readily available from a number of vendors. The solids particles
will be heated in
this unit before they are returned to the main reactor unit. The gasifier will
use the coke that is
formed in the reactor as a fuel source, and the gas formed in the combustor
will be used to
fluidize the main reactor.
Most of the evolved vapor phase product, which may comprise more than one
substance or product, will be generated in the emulsion phase of the fluid
bed. Due to the rapid
vertical mixing of solid particles the products will be formed at all heights
within the fluid bed.
-20-
. CA 02446889 2003-10-27
Due to the fluid mechanics associated with the bed, the gas contained. in the
emulsion phase of
the fluid bed will generally flow downwards, in opposition to the upward flow
of gas in the
bubble phase. Vapor phase products will be transferred from the emulsion phase
to the bubble
phase mainly through the mixing of the gas from these two phases in the grid
zone of the reactor.
Reducing the height of the bed increases the portion of the bed occupied by
the grid zone, and
also reduces the time it takes for the evolved products to reach the bottom of
the bed and be
mixed into the escaping bubbles. The bubbles can then rise to be collected in
the vapor
collection apparatus at the top of the reactor vessel.
The invention provides quantifiable economic advantages over competing fluid
bed technologies that have well mixed solids and confounded gas and solid
phase residence
times. dVhere the desire is to maximize the yield of condensable overhead
vapors, three main
advantages are noted:
I S l . Reduced reactor operating temperatures. As mentioned above under point
(f~, in a
well mixed reactor, lower reactor temperatures increase the risk of bogging.
This
concern is managed in the XFC design through increased solids throughput,
enabling lower operating temperatures, and increased yields;
2. Decreased losses of unreacted product. A well mixed reactor will also
result in
the loss of unreacted feed, due to the residence time distribution of the
solid
particles. This does not occur in the XFC unit; and
3. Decreased over-cracking of gaseous products. The shallower bed allows for a
reduced gas phase residence time so that valuable products are not downgraded
due to high seventies.
Issues 1 and 2 can be addressed for a process incorporating a well mixed fluid
bed
reactor by making the vessel significantly larger. However, this ~~ill
significantly increase
capital costs. In the current example, a well mixed reactor would require 16.5
times the solid
particle holdup in order to ensure 95% of the solid particles are retained for
a sufficient amount
of time for the reaction to go to completion. Furthermore, this apparent
remedy will only serve
-21
., CA 02446889 2003-10-27
to exacerbate the problem outlined in 3, offsetting any incremental benefit
associated with the
increased reactor size.
As a conservative estimate of incremental yields, the XFC is expected to have
the
capacity to increase the yield of condensable products by 2-3%, on an absolute
basis. Tlhis value
is very significant in the industry, where yield increments ors the order of
0.1% are seen as
significant, and have formed the basis for major capital expenditures. As a
result, the XFC
process has the additional benefit of making better use of the natural
resource. This is significant
given environmental concerns that have received worldwide attention and
general endorsement.
-22-
