You are on page 1of 8

Advanced propane dehydrogenation

Oxdehydrogenation-based on-purpose propane dehydrogenation can close the


propylene supply-demand gap

Max Heinritz-Adrian and Sascha Wenzel Uhde GmbH


Fekry Youssef Egyptian Propylene & Polypropylene Company

T
he world is currently witnessing a signifi- production is based on ethane feedstock yielding
cant change in the propylene market as it no propylene co-production
moves away from co-production and Several on-purpose propylene production tech-
towards more on-purpose production and the nologies such as propane dehydrogenation
supply of propylene. By applying the principle of (PDH) and metathesis have achieved technical
oxydehydrogenation, the feasibility of on-purpose maturity and acceptance, and significant devel-
propylene production by propane dehydrogena- opmental efforts have made themcompetitive
tion (PDH) is further improved. The first plant with co-production technologies in the market
based on this principle is now under realisation in Private-sector companies specialising in
Port Said, Egypt. specific value chains, such as the propylene value
chain, are growing in the marketplace
Propylene market The tight propylene market and high oil prices
Propylene is one of the most important interme- have continuously driven up prices for propylene
diate petrochemical products. Over many years, and propylene derivatives.
it has maintained a remarkable growth in As a result, while in 2003 more than 97% of
demand that is mainly attributed to the signifi- all propylene was produced from steam crackers
cant growth in polypropylene (PP) demand, and refineries, and only 3% by on-purpose tech-
which absorbs more than 60% of all the propyl- nologies, 10% of worldwide propylene production
ene produced worldwide. PP has grown into one by 2012 will be provided by on-purpose technol-
of the most widely applied plastic products, and ogies, mainly PDH and metathesis plants.
its broadness in application which is continu- Metathesis was first commercially applied to
ously expanding and its superior features in propylene production at the BASF/FINA Port
many applications continue to drive demand. Arthur Cracker, USA, and several plants have
Other important propylene derivatives include been built in Asia since then. PDH was commer-
acrylic acid, acrylonitrile, cumene/phenol and cialised in the 1990s, with six PDH plants having
propylene oxide. been started up (one is co-production) in
Although most propylene has been produced Thailand, Malaysia, Korea and Belgium, out of
as a co-product to ethylene in steam crackers which five apply UOP technology and one
and as a byproduct in refineries, we are currently (owned and operated by Borealis in Kallo near
witnessing a significant change in the propylene Antwerp, Belgium) employs Lummus technol-
supply chain with a move away from co-produc- ogy. In the first five years of the 21st century,
tion and towards on-purpose production. There two PDH projects started up; namely, the SPC
are several reasons for this change: plant in Al-Jubail, Saudi Arabia, employing
The growth in demand for propylene has Lummus technology and the Basell/Sonatrach
outpaced the growth in demand for ethylene for joint venture in Tarragona, Spain, employing
many years and will continue to do so UOP technology. A total of five PDH projects are
Due to significantly better feasibility, a large currently under realisation in Egypt, Saudi
share of new cracker projects for ethylene Arabia and Thailand. Development for most of

www.digitalrefining.com/article/1000632 PTQ Q1 2008 1


these projects had already begun in the 1990s. tures would lead to a significant increase in side
However, only after the long-expected tightening reactions and therefore a decrease in the selec-
of the propylene market materialised did they tivity of the main reaction for propylene
finally begin to move forward. production. Furthermore, the most significant
Although additional propylene production side reactions are not limited by thermodynamic
capacity will come on stream in the next few equilibrium. As a result, the higher retention
years, propylene prices are expected to remain times required to achieve higher per-pass conver-
above $800/ton.1 The EPPC PDH/PP project at sions also have a strongly negative impact on
Port Said, Egypt, is the first project in the world propylene selectivity. At the same time, partial
to apply the Uhde STAR process with pressure is kept low by applying absolute pres-
oxydehydrogenation. sures slightly above atmospheric pressure or
even under vacuum, or by using a diluent to
PDH chemistry reduce the hydrocarbon partial pressure. As a
Paraffin dehydrogenation reaction chemistry, result of these considerations, all the technical
although quite simple from a stochiometric point processes for PDH have limited conversion in
of view, is very complex due to its strongly endo- the range of 3255%, with selectivities to propyl-
thermic nature and significant conversion ene in the range of 8791 mole%.
limitations caused by thermodynamic equilibrium. As the significant amount of coke laydown
Side reactions include the cracking of hydrocar- leads to a fairly quick deactivation of the catalyst
bons and hydro-genolysis, as well as (ie, in therange of minutes to a few days), differ-
oligomerisation, cyclisation, hydrogenation of ent reactor systems are applied to account for
olefins, deep dehydrogenation and eventually the the regeneration needs. Continuous systems
formation of coke and tar laydown on the catalyst, applied are moving-bed systems and fluidised
which requires frequent catalyst regeneration (ie, bed systems. Discontinuous (fixed-bed) systems
burning off of coke and tar laydown on the catalyst applied are heated fixed-bed systems (as in
with oxygen or air). Uhdes STAR process) and adiabatic fixed-bed
Two main catalytic systems have been identi- systems.
fied and commercialised for light paraffin
dehydrogenation, the first being supported chro- Key factors in PDH design
mia catalysts (ie, chromia on alumina support), Product work-up sections for all the PDH tech-
which are doped with alkali metals to add alka- nologies are very similar, as the side products,
linity and suppress unwanted side reactions, and and therefore the separation tasks, are also simi-
the second supported platinum or platinum-tin lar. When comparing different PDH technologies,
catalysts, with different support materials. These several key components need to be evaluated to
materials include alumina or zinc-/calcium- find the differences and advantages of one tech-
aluminate, and potentially further modifiers such nology over the other. Table 1 lists certain data
as alkali metals, again to reduce the acidity of generally available in the industry for commer-
the support and suppress side reactions. While cial PDH technology.
chromia catalysts were already used commer- Reaction system Due to the endothermic nature
cially for butane dehydrogenation in the 1940s of the reaction, the fairly high temperature (to
(UOP), platinum catalysts have only been used achieve acceptable conversion) and catalyst
since the 1960s, when they were first applied in regeneration requirements, the reaction systems
UOPs proprietary Pacol process. Today, both for each process are quite complex and therefore
catalyst systems are used in commercially availa- relatively high in investment. An effective design
ble PDH processes (ie, chromia catalyst in of the reaction system applied with respect to
Lummus CATOFIN and platinum-tin catalysts in certain design criteria is essential. Key parame-
UOP Oleflex and Uhdes STAR process). ters are an optimised temperature profile to
With the dehydrogenation reaction being achieve high space-time yields (ie, high conver-
favoured by high temperatures and low partial sion and selectivity at high space velocity), an
pressures, all commercially available technolo- optimised operation/regeneration system that
gies keep the process temperature at or below minimises the effort for the regeneration of the
approximately 650C, because higher tempera- catalyst and an efficient heat supply to the endo-

2 PTQ Q1 2008 www.digitalrefining.com/article/1000632


PDH process technology parameters
UOP Oleflex Uhde STAR process Lummus CATOFIN
Catalyst Pt-Sn on alumina Pt-Sn on zinc/calcium aluminate Chromia on alumina
Reaction system Fluidised bed system consisting of Fixed-bed system consisting of Fixed-bed system consisting of
several radial-flow adiabatic reactors in an externally fired tubular reactor a number of axial-flow adiabatic
series and an external regenerator (CCR) and an adiabatic reactor in series reactors connected in series

Main source for Preheating of the feed and External firing of the tubular reactor Heating the catalyst during
heat supply intermediate products and selective hydrogen combustion catalyst regeneration
before each reactor in the adiabatic reactor

Operating mode Continuous (510 days Discontinuous (7 hours normal Discontinuous (620
catalyst cycle time) operation, 1-hour regeneration) minutes reaction cycle)
Pressure, bara 2 5 0.30.5
Temperature, C 580650 570590 560650
LHSV, [m_ liq. feed/(m_ catalyst * h)] 4 6 <1
Per-pass conversion propane, % 35 40 4865
Selectivity propene, mol-% 8991 89 88

Table 1

thermic dehydrogenation reaction to allow full optimised to achieve a recovery of more than
utilisation of the catalyst activity. 99.5% of the propylene produced in the reaction
Raw gas compression For the separation and section, while keeping it simple indesign.
production of the propylene product the raw Catalyst and chemicals The factors for the cost of
product consisting of propane, propylene, hydro- catalyst and chemicals include the specific cost
gen and side products the gas needs to be of the catalyst, catalyst inventory (as determined
compressed to higher pressures (in the range of by the space-time yield), catalyst make-up (if
1535 bar). Since the dehydrogenation reaction any), catalyst lifetime and chemicals
is commercially realised at fairly low pressures, consumption.
as previously described, the raw gas compression The catalyst for the process has demonstrated
requires significant investment and energy costs. a lifetime of up to seven years in commercial
As the STAR process operates in the presence of application. Together with the optimised reac-
steam, the absolute pressure of the reaction tion concept incorporating oxydehydro-genation,
section can be increased while keeping the it provides for the highest space-time yields (and
partial pressure of hydrocarbons low (and thus lowest catalyst inventory) in the industry. The
conversion high). As a result, the process has a addition of steam to the system already converts
relatively higher suction pressure, resulting in part of the coke and tar laydown mainly to CO2
lower cost (Capex and Opex) for raw gas during the operation and thus allows longer
compression. operation cycles, as well as shorter and simpler
Propane consumption The propane consumption regeneration. Regeneration takes place only in
of a PDH plant typically makes up more than the presence of steam and air, and no chemicals
80% of all operating costs and more than 60% of addition is required for catalyst reactivation (eg,
the cost of production. Propane consumption is chlorinating) or for coke suppression (eg,
determined by the product selectivity achieved in sulphiding). Therefore, no caustic wash down-
the reaction system and the recovery of propyl- stream of the reaction section for the removal of
ene product (as well as unconverted propane) in chlorine is required, and no environmental
the separation section. An optimisation of Capex concerns that may arise from the use of chlorine
vs Opex needs to be considered in terms of high in the process have to be considered. Chemical
conversion vs high selectivity in the reaction consumption in the downstream separation
section, as well as high recovery/high sophistica- section is very limited, with significantly lower
tion vs low recovery/low sophistication in the catalyst and chemical consumption costs.
separation section. An evaluation and optimisa- Energy efficiency Due to the endothermic nature
tion of the process design with respect to these of the dehydro-genation reaction, PDH processes
issues has resulted in the product recovery being are quite energy intensive and a careful design

www.digitalrefining.com/article/1000632 PTQ Q1 2008 3


HP steam

Combustion
air
Fuel gas
Fuel gas

Reformer
Olefin
Gas product
Feed Fractionation
O2/air separation
preheater

Oxyreactor
Heat
recovery Raw gas
compression
Paraffin feed

Boiler feed water

Process condensate

Process steam

Figure 1 STAR process block-flow diagram

with respect to energy efficiency is essential for the second plant, a 40 000 mtpa unit for PBA at
their viability. In the STAR process, the heat Ensenada, Argentina, was commissioned in
from the hot reactor product is effectively recov- 1994.
ered for the production of HP steam. Preheating In 1999, Uhde GmbH acquired the process
of the boiler feed water and the feed/steam technology and catalyst, including all associated
mixture supplies heat for the downstream sepa- patents and rights from Phillips Petroleum.
ration section, which in addition to the reduced Directly after acquisition of the technology, a
energy consumption for raw gas compression, as pilot plant was installed at Uhdes R&D centre in
previously described, provides for high energy Ennigerloh, Germany, with the aim of further
efficiency. enhancing the technology by adding an oxydehy-
drogenation section downstream of the main
Process evolution reactor. The pilot plant was started up in early
The STAR process was originally developed by 2001 and the developmental work took place
Phillips Petroleum (now ConocoPhillips). between 2001 and 2003. In 2006, Uhde signed a
Development began in the 1960s with two contract for the lump-sum turnkey supply of a
primary objectives: the aromatisation of C6+ cuts PDH/PP plant complex based on the STAR proc-
and the dehydrogenation of light paraffins for ess for the oxydehydrogenation of propane and
the production of light olefins. The main devel- the Basell Spheripol process for propylene
opment took place in the 1980s, for isobutane polymerisation for Egyptian Propylene & PP
dehydrogenation as well as for propane dehydro- Company, to be built at Port Said, Egypt, with a
genation. Commercialisation of the technology capacity of 350 000 mtpa. The plant complex is
took place in the early 1990s, when two commer- scheduled to start up in 2010.
cial process plants for the dehydrogenation of
isobutane to isobutene were built and commis- Process description
sioned, with MTBE being the downstream The acronym STAR stands for steam active
product of both plants. The first plant, a 100 000 reforming, describing one of the main process
mtpa plant for Coastal Chem Inc at Cheyenne, features whereby the reaction takes place in the
Wyoming, USA, was commissioned in 1992 and presence of steam. Propane feed, together with

4 PTQ Q1 2008 www.digitalrefining.com/article/1000632


propane recycle and steam diluent, is heated up catalyst consists of platinum based on a basic
through heat exchange with the reactor effluent calcium and zinc aluminate, which is highly
and a fired preheater, then admitted to the main stable in the presence of steam and oxygen at
reactor, the process reformer, an externally fired high temperature. The catalyst is robust and
tubular reactor, which will be discussed in more stable in operation, and has demonstrated an
detail. Most of the propane conversion takes expected lifetime of more than five years in
place in the reformer. commercial application. In fact, one of the
The intermediate product from the reformer is commercial plants has used one load of catalyst
cooled by steam and condensate injection, thereby for almost seven years without interruption. As
adjusting the inlet temperature and steam-to- the catalyst tolerates oxygen-containing compo-
hydrocarbon ratio for the second reactor, the nents as well as some levels of olefins and C4+ in
oxyreactor. In the oxyreactor, oxygen is injected the feed, typically no upstream guard beds need
into the process gas shortly above the catalyst to be considered in the process.
bed. The oxygen selectively converts the hydrogen The catalyst combines a high activity and
from the intermediate product in the top part of selectivity for the dehydrogenation reaction, as
the catalyst bed (selective hydrogen combustion well as for the selective hydrogen combustion in
SHC), which shifts the thermodynamic equilib- the oxy-dehydrogenation section. Therefore, the
rium of the dehydrogenation reaction to higher same catalyst is used in the reformer and the
equilibrium conversion and at the same time oxydehydrogenation section. No start-of-run
provides the necessary heat for the further dehy- (SOR) and end-of-run (EOR) need to be
drogenation reaction taking place in the same considered.
catalyst bed (Figure 1). The operational cycle of the reactor system is
Heat from the hot process gas is utilised for the seven hours of normal operation, followed by
production of steam in the effluent boiler, feed one hour of regeneration. Thus, a surplus of only
preheating and supplying heat for the downstream one-seventh (14.3%) is required to compensate
separation section. The process gas is further for the capacity loss due to regeneration
cooled down and process condensate is separated. requirements.
The process gas containing propylene, propane and
hydrogen, as well as methane, ethane, ethylene and Reactor design and oxydehydrogenation
CO2 is compressed in the raw gas compression and
The reactor design in the STAR process is simi-
sent to the gas-treating section, which comprises
lar to Uhdes reaction section design for syngas
an MDEA wash for the removal of CO2 and a cold production with its ammonia technology. The
box (LTRS) for the recovery of conversion of propane to
propylene and propane from propylene is first achieved in a
light ends. The separated liquid steam reformer-type reactor.
C3 fraction is sent to fractiona- This reactor design has been
tion, which consists of a used in over 60 plants built by
de-ethaniser column for the Uhde for various applications
removal of C2- and a C3- splitter (eg, ammonia and hydrogen).
column for the production of Additional conversion then
high-purity propylene (ie, poly- takes place in the oxyreactor,
mer-grade or chemical-grade, the design of which is based
as required). Unconverted on the secondary reformer
propane from the C3- splitter used in ammonia plants. Uhde
bottoms is, together with fresh has built more than 40 of
feed, sent to a depropaniser these reactors. The oxyreactor
column for the removal of any operates at an absolute exit
C4+ prior to sending it back to pressure of 5.3 bar absolute
the reaction sections. with steam as the diluent, thus
maintaining a low partial
PDH catalyst performance Figure 2 Reactor system of STAR process pressure for hydrocarbons
The commercialised STAR for oxydehydrogenation of light paraffins to ensure high per-pass

www.digitalrefining.com/article/1000632 PTQ Q1 2008 5


conversion. Therefore, an effi- supplied by the selective
cient heat recovery from the hot exothermic combustion of
product gas is possible, and the hydrogen, leading to a more
suction pressure for raw gas uniform heat supply
compression is 3.5 bar absolute. compared to an external
This relatively high suction firing.
pressure lowers compression Cooling of the intermedi-
costs (Figure 2). ate product from the
In principle, oxidative dehy- reformer (upstream of the
dro-genation is a commercially Figure 3 Effluent boiler oxyreactor) by the injection
proven concept, installed in of steam and condensate
seven commercial plants in the 1970s and early allows for the adjustment of the oxyreactor
1980s, with a total capacity of more than one entrance temperature and also increases the
million tons per annum (tpa) butadiene produc- steam-to-hydrocarbon rate, thus reducing coke
tion based on the so-called Oxo-D process by laydown on the catalyst. Therefore, the de-acti-
Petrotex. vation of the catalyst in the oxyreactor (back part
With oxydehydrogenation, part of the conver- of the overall reaction system) is similar to the
sion is moved to the oxyreactor. The applied reformer, and so the oxyreactor can be operated
shift of the endothermic equilibrium by selective on the same operational cycle as the reformer.
hydrogen combustion at the top segment of the After steam/ condensate injection, the interme-
oxyreactor catalyst bed allows for more efficient diate product enters the oxyreactor at the bottom
utilisation of catalyst activity throughout the and passes into one centrally arranged tube,
entire reaction section. As a result, the space- which conducts the gas to the top. This elimi-
time yield and, therefore, the overall catalyst nates troublesome external hot piping to the
requirement are reduced significantly. This head of the reactor. Process gas flow discharged
allows for smaller reactors, thereby reducing from the central tube is reversed at the top of
investment cost. Applying the oxydehydrogena- the oxyreactor and directed over a gas distribu-
tion step also increases the propylene yield. In tor for even distribution. Air or oxygen is
addition, a portion of the heat ofreaction for the admitted to the system and uniformly distributed
endothermic dehydrogenation is internally directly over the catalyst surface via a proprietary
oxygen-distribution
system. In other words,
process gas and oxygen
are rapidly mixed and
directly contacted with
the catalyst bed to
achieve high selectivities
for the conversion of
hydrogen with oxygen.

Effluent boiler
Extensive experience in
ammonia technology
was put to use when
designing the heat-
recovery system. Similar
to the fired-tube boiler
used for HP steam
generation in ammonia
plants, the effluent
Figure 4 3D model of the EPP petrochemical complex in Port Said, Egypt, to start up in boiler used for dehydro-
2009 genation plants consists

6 PTQ Q1 2008 www.digitalrefining.com/article/1000632


of a refractory lined
inlet chamber. The
Egyptian Ethylene
steam drum is mounted natural gas import
on top of the process
gas cooler, and is
supported by downcom- Gas
ers and risers. Erection fractionaton
costs are low due to PDH Propylene PP
PP
(Uhde STAR
shop assembly, and Propane (base spherpol)
Process)
access for maintenance
H2
and inspection is easy.
H2 rich gas
Operation is based on a Offsites and
utilities
reliable, free convection EPP petrochemical
flow, where there are no complex, Port Said
heated dead ends on the
water side where debris
can settle (Figure 3).
Figure 5 Block-flow diagram of the EPP petrochemical complex in Port Said, Egypt
Supplying
polypropylene recycled to the PDH plant, as the necessary sepa-
In late 2006, Uhde was awarded the lump-sum ration equipment is available
turnkey contract for a greenfield petrochemical Hydrogen required in the polymerisation reac-
complex by Egyptian Propylene and tors of the PP plant can be produced from a
Polypropylene Company (EPP). Location of the hydrogen-rich gas stream drawn off in the PDH
complex is Port Said, approximately 170 km plant using pressure swing adsorption (PSA)
north east of Cairo, close to the mouth of the Process oxygen required for the oxydehydro-
Suez Canal. The expected startup is in 2009. EPP genation section in the PDH plant is produced in
is investing some $680 million in this new petro- an air-separation unit that also produces all the
chemical complex, which consists of a PDH plant, service nitrogen required for the overall complex
PP plant, and offsites and utilities plant, as (eg, for pneumatic transport of the pellets in the
shown in Figures 4 and 5. PP plant)
The PDH plant is designed to produce 350 000 Central refrigeration supplies cryogenic service
mtpa of polymer-grade propylene based on the to the PDH plant and chilled water to the PP
enhanced STAR process technology with oxydehy- plant
drogenation. Feedstock to the PDH plant will be The steam system consisting of HP, MP and
propane from Egyptian natural gas sources. The LP headers is designed and optimised for maxi-
propylene produced will be converted to polypro- mum flexibility, while integrating the steam
pylene in a 350 000 mtpa PP plant applying requirements of both process plants at different
Basells Spheripol process. The PP plant will pressure levels efficiently.
produce a wide range of high-quality PP plastic Low-level heat in the PDH plant is utilised to
pellets. EPP will supply the PP to both the domestic preheat all the boiler feed water required for the
and export markets. overall complex
When planning the complex, special attention Flaring and water usage is minimised.
was given to the coupling of the PDH and PP Besides the technology selection and the opti-
plant not just with regard to the utility supply, misation performed in the design, EPP has
but also integration of these process plants. Some managed to build certain key competitive advan-
examples of the plant integration are as follows: tages in developing and contracting the complex,
Typically, a bleed stream produced in the PP including location, feed supply and low energy
plant consisting of unconverted propylene with and labour costs, and contracting concept.
propane as an impurity had to be reprocessed Port Said is strategically well located at the
inside the PP plant. When combining a PDH mouth of the Suez Canal connecting the
plant with a PP plant, such streams can be easily Mediterranean and the Red Sea. Therefore, the

www.digitalrefining.com/article/1000632 PTQ Q1 2008 7


project is perfectly connected and can provide enhance the feasibility of on-purpose propylene
product to many different regions in the world at production by applying the oxydehydrogenation
a competitive cost. Propane feed is supplied from principle. Advantages of this technology are a
Egyptian natural gas, mainly from the neigh- lower reactor size and lower compression costs.
bouring UGDC gas fractionation plant. This The PDH plant in EPPs new petrochemical
further increases the Egyptian value chain and complex in Egypt, expected to start up in 2009,
gives the project a long-term and secure feed- will be the first application for the advanced
stock supply. Furthermore, the plant location oxydehydrogenation technology.
provides direct access to fresh water and
energy supply. The cost for energy and labour in References
Egypt is at an attractively low level in regional 1CMAI, Propylene Markets and Tech Conference, Bangkok, 2007.
comparison. The plant was designed and built
Max Heinritz-Adrian is global marketing manager for Uhde GmbH
under a lump-sum turnkey EPC contract. This
gasification technology in Werdohl, Germany.
provides for full single-point responsibility with
Email: max.heinritz-adrian@thysenkrupp.com
a true wraparound guarantee.
Sascha Wenzel is senior process engineer for Uhde GmbH in
Werdohl, Germany. Email: sascha.wenzel@thyssenkrupp.com
Summary and outlook Fekry Youssef is chief process engineer for Egyptian Propylene
Propylene demand is expected to grow annually, & Polypropylene Companys Port Said petrochemical complex in
with a 5.4% increase in the period from 2007 to Egypt.
2012.1 Steam cracking as the biggest conven-
tional source of propylene will no longer be able
to supply enough propylene to the market as
cracker feedstock moves from naphtha towards
ethane, thus producing no propylene byproduct. Links
The resulting supply-demand gap can only be
More articles from: ThyssenKrupp Uhde
overcome by on-purpose propylene production
technologies like PDH and metathesis. In recent More articles from the following categories:
years, the share of propylene produced by on- Design, Engineering & Construction
purpose technologies has increased significantly. Petrochemicals Propylene Maximisation
The process discussed in this article can further

8 PTQ Q1 2008 www.digitalrefining.com/article/1000632

You might also like