Shaping the future of on-purpose propylene production
In the early 2000s, very few analysts predicted the innovation that would change the landscape in the energy, refining and petrochemical sectors over the next several decades.
In the early 2000s, very few analysts predicted the innovation that would change the landscape in the energy, refining and petrochemical sectors over the next several decades. Innovators successfully combined hydraulic fracturing and horizontal drilling to produce unprecedented volumes of shale gas and shale oil.
This innovation rapidly enabled petrochemical companies to begin producing ethylene from cheap ethane. Initially, the price of crude oil remained high, which meant that naphtha was rapidly phased out as the preferred feedstock, where feasible, in place of ethane. When the rate of naphtha feed was reduced, the coproduct propylene produced was also reduced.
This scenario created an imbalance in the supply and demand for propylene. As this imbalance occurred, the economics for producing propylene from propane began to look very attractive. Fortunately, proprietary propylene production processes had been adapted from existing technologies for the production of propylene.1,2 The licensors of these technologies were able to take advantage of market conditions, and sold many units to help bring the markets back into balance.
Now, another innovation is ready to change the landscape of the catalytic dehydrogenation marketplace. To understand the significance of this innovative technology, it is imperative to first understand the fundamentals and evolution of catalytic cracking process technology.
Fundamentals and evolution of cat cracking technology
Several basic fundamentals constrain potential process solutions when cracking or dehydrogenating hydrocarbons. Here, cracking and dehydrogenation are referred to collectively as “cracking.”
The first fundamental is that the reactions are endothermic, which means that heat must be added to the process to accomplish the reaction. Secondly, the chemical equilibrium of the dehydrogenation reaction favors olefins at low pressure and high temperature. The low-partial-pressure preference of the dehydrogenation reaction is at odds with the high-pressure cryogenic separation requirement for light hydrocarbons; therefore, the reactor pressure must be optimized with the entire process, as shown in Eq. 1:
C3H8 ↔ C3H6 + H2 (1)
The third fundamental is that the high temperature requirement of the reaction inherently results in the formation of coke on the catalyst, which depresses catalyst activity. The resulting loss of catalyst activity requires regeneration to remove coke.
Fourth, the thermal or gas phase reaction has a much lower selectivity to the desired product than the catalytic reaction, requiring the gas residence time at temperature to be minimized.
Each process that has been successful in the area of catalytic cracking of hydrocarbons has solved or managed each of these fundamental constraints. Table 1 illustrates four processes in this area.
To put heat into the reaction, both the proprietary fluid catalytic cracking (FCC) processa (“FCC process”) and the proprietary dehydrogenation process 1b (“DH Process 1”) primarily use the catalyst as a source of heat, while the proprietary dehydrogenation process 2c (“DH Process 2”) uses the superheated feed.1 The conventional styrene process uses a cofeed steam diluent.10 Processes that do not require the feed to be preheated over the thermal reaction temperature have an inherent advantage in product selectivity.
To drive the reaction equilibrium, the styrene process and DH Process 1 use vacuum.5,6,10 For the styrene process, creating vacuum can be done efficiently since the diluent steam and product liquids can be condensed prior to compression of the vent gas, which minimizes the compression duty. The FCC process is not as adversely impacted by higher pressure due to different chemistry; therefore, the process has tended to increase pressure over the life of the technology to reduce capital investment per ton of product. Finally, DH Process 2 tends to use more moderate pressures to minimize the energy and capital cost of the compression train. DH Process 2 does require a significant hydrogen (H2) recycle, which is one reason for the moderate pressure selection.8
The catalyst activity loss due to coke formation is mitigated primarily through burning the coke in the FCC process and DH processes 1 and 2.3,5,7,9 DH Process 2 does require a relatively complex regeneration, in that platinum must be redistributed with chlorine.6 In addition, sulfur (S) and H2 are intentionally injected to control coke formation.7,9 The styrene process uses a novel solution of reacting the coke off the catalyst in-situ, with steam and a catalyst promoter.10
Finally, the residence time of the gas at temperature must be managed to minimize selectivity loss, which is dictated by reactor type and heat input scheme. Both the conventional styrene process and DH Process 2 use radial adiabatic contacting reactors, which are known for low vapor residence times.7,10 DH Process 1 uses fixed adiabatic reactors that operate in a cyclic mode.2 Finally, the FCC process uses a circulating fluidized bed and rapid gas-solid separation to control gas residence time.3
All of these processes include designs and mitigations to manage the fundamental technical constraints, but some have evolved more than others, forming the basis for the next generation of catalytic dehydrogenation processes. Technology will ultimately converge to promote the best overall economic solution, as investors desiring higher returns drive companies to find additional ways to increase profits. The question is not if, but rather, when?
The global gasoline industry is expansive, with more than 1.2 Btpy of gasoline produced.11 Within such a large market, economic drivers exist to improve technologies that can help reduce the cost of production. Technology has evolved rapidly9,10 to the most economical solution—a circulating fluid bed.
The styrene market is much smaller than the gasoline market, with 27.5 MMtpy of styrene produced in 2014.12 Therefore, the rate of innovation has been slower over the years.13 The propylene market is approximately 89 MMtpy, although only 5 MMtpy of this propylene is on-purpose propylene.12 Given that solutions to provide these products exist (steam cracking and FCC), little emphasis has been placed on improving returns through technological innovation.
Advanced FCDh process
A proprietary catalytic dehydrogenation reactor-regenerator system is shown in Fig. 1. The technology, a fluidized catalytic dehydrogenation processa (FCDh process), includes a circulating fluidized bed system with an alumina-supported catalyst. Propane is introduced into a turbulent or fast-fluidized reactor, followed by a plug-flow riser reactor that transports the catalyst to close-coupled cyclones for rapid product and catalyst separation. The catalyst is then stripped with a stripping agent to recover any product entrained with the catalyst.
Fig. 1. Proprietary catalytic dehydrogenation reactor-regenerator system. The reactor is shown in red, and the regenerator is shown in blue.
The spent catalyst is transported to a combustor, where it is reheated in a fluidized bed, using external heating fuel. The catalyst and flue gas then flow through a riser and a termination device. The remaining catalyst is separated from the flue gas in cyclones and returned to a bubbling bed. The catalyst is then rejuvenated with air to recover the activity in an oxygen soak zone.
Catalyst. The catalyst shown in Fig. 2 is produced using a commercially available alumina support and impregnated with the active components, which include gallium and low levels of platinum. The catalyst has been developed over decades and has been shown to meet the requirements of operating in a fluidized bed dehydrogenation process. For example, the catalyst can perform the dehydrogenation of paraffins and alkyl aromatic compounds, including ethane, propane, butane, isobutane, ethylbenzene and others.
Fig. 2. Proprietary catalytic dehydrogenation catalyst.
The catalyst also promotes the combustion of supplemental fuels, such as methane, hydrogen, ethane and others. As a catalyst operates in a circulating fluidized bed system, it is desirable that any deactivating function that may occur in the reactor, reactor stripper or combustor be reversible. As such, the catalyst is capable of cyclic reactivation after temporary activity loss.
In addition, the catalyst can tolerate reasonable levels of impurities that are commonly found in commercially available propane streams. Finally, the catalyst has the appropriate mechanical properties (particle size and density) to make it fluidizable and resistant to attrition.
Fig. 3. Proprietary reactor design.
Reactor design. The reactor design (Fig. 3) for the circulating fluidized bed process is an upflow reactor where the average velocity of both the catalyst and the gas are in the upward direction. The bottom portion of the reactor operates as a turbulent or fast fluidized reactor and is followed by a conventional riser reactor. Upon completion of the catalytic dehydrogenation reaction, the catalyst and product gases are rapidly separated in a cyclonic separation device that is commonly used in the FCC process. After the catalyst is removed, a stripping gas is injected in the annular space. The product gases are recovered prior to transporting the catalyst to the regenerator.
The advantage of this reactor and catalyst process is that they can achieve high conversions, such as 45% of the propane feed, at moderate pressures. In addition, there is a minimal impact from feed impurities, such as sulfur, which eliminates the need for costly front-end capital investment. Finally, the selectivity to propylene is maximized due to the low gas residence time at temperature requirements, and the low levels of coke that are formed on the catalyst during the reaction step.
Fig. 4. Proprietary regenerator design.
Regenerator design. The regenerator design, as shown in Fig. 4, consists of a lower combustor section and is followed by a catalyst removal section at the top and a catalyst reactivation section in the center. Due to the low coke generation, supplemental fuel and air are injected into the combustor, which generates the heat for the endothermic dehydrogenation process. The combustor is a bubbling bed that operates with an average gas and catalyst velocity in the upward direction. The catalyst and flue gas are transported through a riser, where the catalyst is partially removed from the flue gas. The flue gas is then processed in a cyclone system, which removes any remaining catalyst. The catalyst is reactivated by a simple air treatment prior to being transported back to the reaction system.
Advantages of the proprietary regenerator process include the catalytic combustion of a fuel, which minimizes any NOx formation vs. competing alternatives, which require a flame for combustion. In addition, the catalyst is continuously regenerated, which eliminates the need for large high-temperature valves operating at a high frequency. These types of valves have been prone to failure in batch regenerator processes. In addition, there is no need for complex regeneration steps, such as chlorine introduction.
Comparison to existing technologies
As with any new technology, it is imperative to benchmark the new process with existing technologies to ensure that a sufficient improvement in performance can be obtained. Table 2 summarizes the key differences between the FCDh process and DH processes 1 and 2.
In regard to reactor technology, DH processes 1 and 2 require multiple reactors to process the required amount of propane to achieve acceptable propane dehydrogenation (PDH) conversion levels and economies of scale. In DH Process 2, four reactors are used to achieve the target conversion, as multiple heat-up steps are required.7 DH Process 1 requires multiple reactors, depending on capacity. Each reactor spends a significant time operating in the regeneration step.2 The proprietary FCDh process utilizes a single reactor and operates in reaction mode continuously.
Heat introduction into the reactor is extremely important, as it impacts the selectivity of the process, as well as the capital intensity. DH Process 1 provides the heat by a combination of hot catalyst generated in the decoking cycle and hot propane.2,5,14 The cyclic heat introduction by catalyst results in a high temperature initially during the start of a reaction cycle, and then a low temperature just prior to regeneration. This cycle results in a non-optimum temperature for the reaction, thereby reducing the overall selectivity of the process. In addition, by feeding the propane at temperatures above the thermal cracking temperatures, nonselective products are formed.
Alternatively, DH Process 2 preheats the propane and gas products before each reactor to temperatures above the thermal cracking temperature of propane.1,7 This process results in non-selective reactions due to the residence time in these sections, as well as to increased capital costs and NOx emissions from the fired heaters. The FCDh process introduces propane below the thermal cracking temperature and utilizes the catalyst to rapidly heat up the propane to reaction temperature. The high surface area of the catalyst and rapid heat transfer effectively transfer heat to the propane vapor.
Each process utilizes a different catalyst, which impacts the economics and handling requirements. For example, DH Process 1 uses a large quantity of relatively inexpensive chromium catalyst.2 The catalyst poses environmental handling challenges that must be managed. Alternatively, DH Process 2 utilizes a catalyst that contains approximately 3,000 ppm of platinum supported on alumina.15 The Dow process uses a more sustainable catalyst that is significantly less expensive than the DH Process 2 catalyst, as the platinum levels are substantially lower.
The regeneration of all of the processes involves burning any coke buildup on the catalyst, in either a continuous or batch process. DH Process 2 uses a complex cycle of coke combustion, drying and platinum redispersion with chlorine and oxygen, followed by H2 reduction.16,17 Both the FCDh process and the regeneration in DH Process 1 are relatively simple and straightforward.
The residence time of the catalyst in the reactor varies widely between the three processes, and generally impacts the ability of the catalyst to withstand feed impurities and deactivation. For example, the reactor in DH Process 1 undergoes regeneration every 7 min–15 min, which enables any deactivation that may have occurred from feed impurities to be removed each cycle.2 Alternatively, DH Process 2 has a catalyst residence time of 7,200 min–14,400 min (5 d–10 d), which makes the catalyst much more susceptible to feed impurities and may require additional capital to mitigate.18 The proprietary FCDh process has a catalyst residence time in the reactor of less than 2 min, which makes the process more forgiving to feed impurities and potential impacts to the catalyst.
The pressure of the processes dictates the maximum potential conversion due to chemical equilibrium. For example, the DH Process 1 reactor operates at a low pressure of approximately 7 psia to drive higher conversions.5,6 This increases compressor operating cost and capital. Alternatively, both the proprietary FCDh process and DH Process 2 operate between approximately 16 psia and 25 psia, which is slightly above atmospheric pressure.8 Due to the long catalyst residence times, DH Process 2 does require recycle H2 to suppress coke make, which also suppresses the maximum possible conversion.7,9
Finally, the reactor conversion and selectivity ultimately dictate the economics of the compression and separation sections of a commercial plant. After a thorough literature review, the conversion and selectivity for both DH Process 1 and DH Process 2 are difficult to find reported together. DH Process 1 has published propane conversion numbers ranging from 48%–53%, with a separate selectivity performance of 86 mol% at an unknown conversion.4
DH Process 2 has published propane conversions ranging from 30%–40%.6,20 DH Process 2 selectivity has been reported at 89 mol% at an unknown conversion, 81 mol%–85 mol% at 30% conversion and 84 mol% at 40% conversion in a patent.19,20 The FCDh process is expected to operate at 43%–53% conversion, with propane to propylene selectivities of 92 mol%–96 mol%, which is a significant improvement over existing commercial technologies. The performance of the units is expected to change over time and as a function of plant scale, which provides some challenges when comparing conversion and selectivity performance.
The FCDh process delivers significant capital and operating cost advantages due to its inherent simplicity and multifunctional nature. The capital outlay is expected to be at least 20% less and up to 50% less than other existing commercial technologies at world scale, based on internal studies. In addition, by performing the reaction at moderate pressure without H2 recycle, the process is expected to have 20% lower energy consumption. The low gas residence time at temperature design enables the process to achieve a 5% lower propane unit ratio than current processes. In addition, through catalytic combustion of process gas, the NOx emissions will be lower than alternative technologies. In conclusion, the technical advantages of FCDh process technology offer significant economic advantages vs. current technologies.
Potential applications for the FCDh process are numerous, as shown in Fig. 5. For example, in PDH, the reactor technology can plug into steam crackers that are new or bolted on to an existing cracker. The reactors can also be used to build a new PDH unit or to retrofit an existing one.
Fig. 5. FCDh process platform reactor technology with multiple application options.
Additionally, the reactors can be used to build new or retrofit existing ethylbenzene to styrene dehydrogenation facilities, butane to butene, or isobutane to isobutene plants. C4 dehydrogenation units can be integrated into refineries to upgrade butane or isobutane for use in the alkylation section to produce alkylate.
In the case of PDH plant integration with existing steam crackers, the FCDh process is claimed to be the only technology that can integrate economically with a cracker, due to its comparable operating pressure, high propane conversion, lack of an H2 recycle requirement and simple reactor section design. Due to the simple reactor-regenerator configuration, small plants still maintain attractive capital intensity when integrated with a steam cracker, unlike competing technologies.
To show the advantage of an integrated FCDh reactor system and ethane cracker vs. an ethane-propane (EP) cracker, two examples are presented. The first example is an EP cracker (Fig. 6). In this example, if a producer wanted to make 1.5 MMtpy of ethylene and 500 Mtpy of propylene, an EP cracker with approximately eight furnaces could be envisioned. Two furnaces would feed ethane, and five would feed propane, including the recycle streams. One furnace would be in decoking mode. The feedstock would be 250 Mtpy of ethane and 3 MMtpy of propane. This cracker would produce 1.25 MMtpy of other products.
Fig. 6. EP cracker with capacity of 2 MMtpy.
In the second example, an integrated FCDh reactor system with ethane furnaces is presented (Fig. 7). In this example, seven furnaces are used to crack ethane and one FCDh reactor is used to process propane, which, in turn, produces about 1.5 MMtpy of ethylene and 500 Mtpy of propylene. In this example, 1.95 MMtpy of ethane feedstock and 550 Mtpy of propane feedstock are required. Only 500 Mtpy of byproducts are produced, representing an approximate 750-Mtpy reduction in the amount of byproducts produced.
Fig. 7. Integrated FCDh reactor system with ethane cracker.
The total predicted economic advantage of the integrated FCDh reactor and ethane cracker vs. an EP cracker that produces essentially the same product mix is $234 MM/yr of additional value. This additional value comes from the products vs. the raw materials, based on the average predicted price from 2016–2020.21
The FCDh reaction system is considered a breakthrough technology that provides significant advantage in the production of propylene from propane. The authors believe the technology represents the future of how on-purpose propylene will be produced.
The process’ simple design provides a capital-efficient and energy-efficient method to produce propylene from propane. In addition, it provides for lower raw materials usage and simplicity of operation vs. other options.
This technology is thought to have the potential to change how hydrocarbons are produced in the chemical industry. Dow would be willing to license the technology for applications in which catalytic dehydrogenation is the preferred solution. HP
a The Fluidized Catalytic Dehydrogenation (FCDh) process is a trademark of The Dow Chemical Co.
b The CATOFIN process is a registered trademark of CBI Lummus.
c The OLEFLEX process is a registered trademark of UOP.
- Chaiyavech, P., “Commercialization of the world’s first Oleflex Unit,” The Journal of the Royal Institute of Thailand, Vol. 27, No. 3, July–September 2002.
- Oviol, Bruns, Fridman, Merriam and Urbancic, “Mind the tap,” Hydrocarbon Engineering, September 2012.
- Wilson, J., Fluid Catalytic Cracking: Technology and Operation, PennWell Books, Tulsa, Oklahoma, 1997.
- Lummus Technology, “CATOFIN dehydrogenation,” Lummus Technology, a CB&I Company, 2009, online: http://www.cbi.com/images/uploads/tech_sheets/Dehydrogenation.pdf
- Seo, S.-T., W. Won, K. S. Lee, C. Jung, and S. Lee, “Repetitive control of CATOFIN process,” Korean Journal of Chemical Engineering, Vol. 24, No. 6, November 2007.
- Weckhuysen, Sattler, Ruiz-Martinez and Santillan-Jimenez, “Catalytic dehydrogenation of light alkanes on metals and metal oxides,” ACS Publications, August 2014.
- UOP, Oleflex Process Brochure, “UOP Oleflex process for propylene production.”
- Kogan, Herskowitz, Woerde, and van den Oosterkamp, “Article 16A: Select dehydrogenation of propane using an improved dehydrogenation catalyst,” 10th Ethylene Producers Conference, American Institute of Chemical Engineers, December 1998.
- Heyse, Johnson and Mulaskey, “Dehydrogenation processes, equipment and catalyst loads,” US Patent 5,406,014, Chevron, April 11, 1995.
- Chen, S.-S., “Styrene,” Raytheon Engineers and Constructors, Kirk-Othmer Encyclopedia of Chemical Technology, John Wily and Sons Inc., Hoboken, New Jersey, December 4, 2000.
- US Energy Information Administration, “World crude oil production by year,” 2014.
- US Energy Information Administration, “Global annual production of styrene and propylene monomer,” 2015.
- Mitchell, J. E., “Styrene: Its polymers, copolymers and derivatives” in “Chapter 2: Manufacture of Styrene Monomer,” Hafner Publishing Co., 1952
- Gartside, R. J., “Catalytic hydrocarbon dehydrogenation system with prereaction,” US Patent 6,392,113 B1, ABB Lummus, May 21, 2002.
- UOP, “Oleflex DeH-16 catalyst description,” online: http://www.uop.com/objects/DeH16.pdf
- Cottrell, P. R. and M. E. Fettis, “Process for the dehydrogenation of hydrocarbons,” US 5,087,792 A, UOP, February 11, 1992.
- Sechrist, P. A., D. W. Robinson and W. D. Schlueter, “Method for controlling moisture in a catalyst regeneration process,” US 6,123,833 A, UOP, September 26, 2000.
- Meyers, R. A., Handbook of Petrochemicals Production Processes, McGraw-Hill Handbooks, New York, New York, 2005.
- Glover, B. K., J. A. Zarraga and M. A. Schultz, “Fluidized bed reactor with back-mixing for dehydrogenation of light paraffins,” US patent application, US 2008/0161624 A1, UOP.
- Houdek, J., et. al, “On-purpose propylene technology developments,” UOP, ARTC 8th Annual Meeting, Kuala Lumpur, Malaysia, April 29, 2005.
- IHS, “Chemical price and economics,” January 5, 2016.
- Chimney and Refractory International SRL, “APPC PDH/PP Project—8 Catofin Reactors—Jubail Saudi Arabia Engineering, Design, and Installation of Refractory Lining,” CRI International presentation, online: http://www.chimneysandrefractoriesinternational.com/attachment/APPC-SAMSUNG%20PDH-PP%20CATOFIN%20REACTORS%20PROJECT.pdf
From the Archive