P. K. Niccum, Contributing Author, Houston, Texas (U.S.)
This article provides the > 80-yr evolution of fluid catalytic cracking (FCC), from process inception in World War II (WW II) to lower emissions of atmospheric pollutants from catalyst regeneration. The evolution of controls for minimizing FCC flue gas pollutants, including carbon monoxide (CO), sulfur oxides (SOx), nitrogen oxides (NOx) and particulate matter are discussed. Additionally, initiatives to reduce carbon dioxide (CO2) emissions from FCC unit (FCCU) operations are noted.
The history and evolution of the FCC process. FCC is a continuous process, utilizing a powdered catalyst in a reactor to heat and crack petroleum feedstocks, while coke is deposited on the catalyst. The process includes a regeneration vessel where coke is continuously burned off the catalyst, producing a flue gas stream composed mostly of nitrogen (N), water vapor and CO2.
The FCC process was developed and commercialized during WW II, where it played a critical role in the production of high-octane aviation fuels by producing FCC gasoline and olefins for alkylation. Following the war, FCCUs turned to the production of gasoline for domestic automobiles and C3–C4 olefin feedstocks for the petrochemicals industry. As oil refining has evolved over the last 80 yrs, the FCC process has evolved with it, meeting the challenges of cracking heavier, more contaminated feedstocks, while still accommodating increasingly stringent environmental regulations.1
The development of the FCC process was enabled by previous efforts to commercialize heavy oil catalytic cracking processes. Two of the most important milestones enabling the development of the FCC process are:
The commercial catalytic cracking of heavy oil on aluminum chloride (a Friedel-Crafts catalyst) was established in 1915 by A.M. McAfee of Gulf Refining in Port Arthur, Texas (U.S.).2 Literature and patents from the 1920s suggest that gasoil or kerosene was treated with < 5% anhydrous aluminum chloride in a vertical still provided with a stirrer.3 The product was of good quality, requiring only minimal treatment with water or a caustic soda solution and steam distillation for use as motor gasoline.3 A residue of asphaltic or coke-like material remained with the catalyst. While gasoline yield could be increased by 20%–30%, the high costs of making and recovering the catalyst prevented additional commercialization of the process.4,5
The demand for high-octane gasoline in the 1920s inspired industry to further explore the topic of catalytic cracking that had been discovered a decade earlier. Chief among these professionals was Frenchman Eugene Houdry, a mechanical engineer and avid participant in automobile racing who went on to lead the development of catalytic cracking.4 Houdry began studying the catalytic cracking of petroleum in 1927 after completing related work in France on the catalytic conversion of coal liquids and tars into gasoline. He was searching for a catalyst to start the cracking reactions, testing hundreds of prospects. By late 1938, he discovered that acid-activated natural clays (such as Fuller’s Earth) containing Montmorillonite provided the needed catalysis for petroleum.4,6 The most dramatic benefit of the earliest Houdry units was their contribution to the production of 100+ octane aviation gasoline just prior to the outbreak of WW II. The Houdry plants utilizing catalyst beads in fixed-bed reactors provided better gasoline for blending with scarce high-octane components, as well as byproducts that could be converted by other processes to make more high-octane blendstocks. Multiple reactors were used to periodically burn off coke, producing a flue gas stream that was vented to the atmosphere. By 1940, 14 Houdry plants were operating, with a total capacity of 140,000 bpd.4,7 The increased octane performance of the aviation gasoline blend, including other very high-octane components, meant that Allied planes were better than Axis planes by a factor of 15%–30% in engine power for takeoff and climbing, 25% in payload, 10% in maximum speed and 12% in operational altitude. In the first six months of 1940, at the time of the Battle of Britain, 1.1 MMbbl/mos of 100-octane aviation gasoline were shipped to Allied forces. Houdry plants produced 90% of this catalytically cracked gasoline during the first two years of the war.4,7 In the late 1930s, Standard Oil Development and The M.W. Kellogg company teamed up—which later included other oil companies and UOP—to develop a catalytic cracking process that would use a powdered catalyst to circumvent Houdry patents based on catalyst beads.6,8 This resulted in the first FCC process commercialized in May 1942. Spurred by the wartime need for gasoline, FCC was soon providing aviation gasoline to the Allies, and C4s for a new and rapidly expanding U.S. synthetic rubber industry. By the end of the war, 34 FCCUs (22 built by Kellogg) were operating, with a total capacity of > 500,000 bpd.8,9 While the FCC process continues, the tireless contributions of the developers of the Houdry fixed-bed plants and moving-bed Thermafor catalytic cracking plants cannot be understated as a valuable component in the war effort.
MAJOR MILESTONES IN FCC FLUE GAS EMISSIONS CONTROL
The combustion of coke in the FCC regenerator generates a variety of potential atmospheric pollutants, such as CO, SOx, NOx and particulate matter (PM), but these are controlled at or below regulatory mandated levels using a variety of technologies.10
FCC catalyst losses from early FCCUs were extremely high but improvements in FCC catalyst integrity, improved cyclone designs and electrostatic precipitators (ESPs) brought the atmospheric catalyst losses down significantly after WW II. The concentration of CO in FCC flue gas of older FCCUs was in a range of approximately 5 vol%–10 vol% (50,000 ppmv–100,000 ppmv). Flue gas was expelled into the atmosphere through tall stacks that diluted the flue gas with air. This controlled the ambient CO concentrations down to lower levels that were considered tolerable at the time.
Historic milestones in reducing FCCU regenerator flue gas pollutants following WW II include:
1940s: ESPs were applied on FCCUs during WW II to minimize losses of catalyst to the atmosphere. Later in the decade, multistage cyclones were used to lower catalyst losses.1 The desulfurization of FCC liquefied petroleum gas (LPG) and fuel gas products was practiced using amine scrubbing, and sulfur recovery units produced sulfur from hydrogen sulfide (H2S) for industrial use. Simultaneously, this reduced SOx emissions from burning LPG and fuel gas elsewhere.
1950s: The first FCCU CO boiler was commissioned at Sinclair Refining Co.’s refinery in Houston, Texas (U.S.) in 1953. The CO boiler was installed to burn CO and recover the heat of combustion. This also reduced flue gas CO emissions from 5 vol%–10 vol% to much lower levels.
Shell Oil commercialized the use of a turboexpander train in 1957 to recover electrical power from the letdown of FCCU flue gas pressure. This included the development of Shell’s Third-Stage Separator (TSS) to minimize expander blade erosion.
1960s: Additional Shell-licensed power recovery trains were installed.
1970s: FCCU wet gas scrubbers were introduced by Exxon in 1974, providing a robust means of reducing both SOx and catalyst dust emissions in flue gas to the atmosphere.11
Many states mandated the use of direct-flame afterburners (CO boilers) for FCCU CO clean-up. The energy crisis of the 1970s inspired refinery energy conservation projects that included many more FCCU CO boilers and more power recovery system installations.
Complete CO combustion in FCCU operations was introduced by AMOCO around 1975 in a process referred to as high-temperature regeneration.4 While this development targeted lower carbon on regenerated catalyst and lower coke yields for improved product yields and the elimination of CO boilers, it also significantly reduced NOx emissions compared to burning CO in very high-temperature CO boiler environments.
Further reductions in FCCU CO emissions were achieved when Mobil Oil introduced platinum-containing CO combustion promoter additives. This initially brought CO emissions in complete combustion operations into a range < 500 ppm and over time down into a range of 50 ppm–200 ppm.
1980s: Platinum-containing catalytic converters were being installed on new cars, helping to eventually bringing the atmospheric CO levels in cities down to much lower levels.
The desulfurization of FCC feedstocks for lower emissions and better yields began to accelerate.
FCCU SOx reducing catalyst additives were also introduced during this decade. The reduction in SOx emissions was achieved based on chemistry that converts SOx in the regenerator into H2S released in the FCCU reactor/stripper where it would later be scrubbed from the hydrocarbon products.
1990s: Driven by increasingly stringent U.S. Government environmental regulations, more FCC flue gas scrubbers and more efficient ESPs further reduced particulate emissions to the atmosphere.
Lower NOx CO promoters were introduced.11,12
More improvements were made to reduce NOx emissions in the FCCU flue gas from both catalytic and non-catalytic processes to reduce NOx to nitrogen gas. Examples of these development include selective catalytic reduction (SCR) processes, selective non-catalytic reduction (SNCR) by Exxon and regenerator design changes.2
2000+: The wider use of FCC feedstock desulfurization and FCC naphtha product desulfurization resulted in a large reduction in SOx emissions to the atmosphere. The reduced SOx emissions included benefits of both lower SOx in the FCC flue gas and lower sulfur levels in FCC products.
FCCUs often do not get full recognition for their broader role in the desulfurization of refinery products. Consider that about 40% of the sulfur in FCCU feed is directly converted to H2S in the FCC process, where it is easily captured in amine scrubbers and sulfur recovery plants.
REGULATORY CONTROLS
In the U.S. over the past 40 yrs, there have been many regulatory drivers impacting FCCU flue gas quality controls. Prime examples have included the continuing application of New Source Performance Standards (NSPS), the implementation of Hazardous Air Pollutant (HAP) controls via what is known as MACT II regulations, and the U.S. Environmental Protection Agency (EPA) enforcement actions and its Consent Decrees. At the same time, FCCUs operating outside of the U.S. have also been under pressure to reduce emissions, sometimes to levels even lower than required in the U.S.
NSPS. NSPS for FCCUs were established for the control of PM, CO and sulfur dioxide (SO2) emissions.5,6 These standards were initially applied to FCCUs constructed after January 17, 1984, as well as units that triggered applicability with any of the following occurrences:
Major FCCU modifications (reconstruction) wherein cumulative investments over a 2-yr period exceed 50% of the capital cost of facility replacement.
Changes in equipment or operation, which increase the rate to the atmosphere of any pollutant to which a standard applies.
NSPS do not set explicit limits on NOx emissions from FCC regenerators. However, site and situation specific NOx limits may be established at the time the FCCU is permitted or modified.
Maximum Achievable Control Technology (MACT) standards. These standards were issued for FCCUs in August 1995.7
MACT II. In September 1998, the EPA proposed National Emission Standards for Hazardous Air Pollutants (NESHAP) to cover these remaining three types of refining process units.8 This NESHAP, commonly referred to as MACT II regulations, would establish the allowable pollution levels for FCCU regenerator PM and CO emissions. As proposed, MACT II PM and CO limits would be the same as the newest NSPS requirements but would apply to FCCUs previously grandfathered with respect to NSPS.
EPA Consent Decrees. In the early 2000s, the EPA entered into binding Consent Decrees9 with several major U.S. refiners to significantly reduce the amount of SO2 as well as NOx emissions from their FCC regenerators. Since the SO2 emissions limitations sought by the regulatory authorities were significantly lower than NSPS levels, their implementation on existing sources via these consent decree projects were expected to ultimately portend revisions to NSPS limits.10
Comments on a few options FOR CONTROLLING FCCU FLUE GAS EMISSIONS
Highlights of a few equipment and other process strategies applied to reducing FCCU flue gas emissions are described in the sections below.2
Regenerator cyclones. Cyclones are the primary defense againt catalyst loss from the regenerator and have continually evolved to better performance and reliability.
Electrostatic precipitators. Electrostatic precipitators (ESPs) have been used for the reduction of FCC particulate emissions since the 1940s. Modern ESPs can be designed to reduce particulate emissions to very low levels. FIG. 1 depicts an ESP in a typical FCC application. ESPs consist of one or more gas tight chambers containing rows of collection plates and voltage discharge electrodes that apply electrical charges to the particles in a waste gas stream to collect them before they reach the stack.
ESP operation consists of three basic steps: particle charging, particle collection and particle removal. Each step must be executed properly to effectively remove particulate to acceptable levels.
The process of charging a particle is accomplished by establishing a non-uniform electric field between the discharge electrodes and the collection plates. This non-uniform field is established by applying high voltage to the discharge electrodes—this generates electrons that flow from the discharge electrodes to the collection plates, resulting in neutral gas molecules that are charged when struck by the high velocity/high energy electrons. The flow of negatively charged gas ions and electrons is generally referred to as corona current flow, depicted in FIG. 2. As the flue gas travels through the resulting corona, suspended particles in the flue gas become charged by the negative ions, which are attracted to the surface of the particles.
A measure of how readily a particle takes a charge is referred to as the particle resistivity. A highly resistive particle is difficult to charge. The resistivity of the catalyst plays a key factor in collection efficiency. Some FCC catalysts display high resistivity, making it difficult to place a charge on them. If a particle is resistive to receiving an adequate charge, a greater electric field must be generated to capture this particle. If a sufficient field cannot be generated, the resistive particle will simply pass through the ESP.
The particle collection process begins the moment the particle absorbs a charge sufficient enough to be attracted by the collection plates. The electric field generated by the discharge electrodes causes the charged particles to migrate towards the grounded collecting plates where they accumulate in a layer, gradually losing their charge. The factors that affect the particle charging and collection process include particle size, particle resistivity, electric field, and the temperature and composition of the flue gas.13,14
In addition, certain gas molecules, which are found in FCCU flue gas, are easier to charge than others, as shown in FIG. 3. Molecules such as NOx, SOx, ammonia (NH3) and water readily absorb an electrical charge. Ammonia and/or water are often injected into FCC flue gas streams upstream of the ESP to increase removal efficiency.15
The removal of collected particles is accomplished in two phases. The first phase involves removing the catalyst from the plates. Upon proper rapping, a solid sheet of catalyst falls by gravity into hoppers located beneath the ESP. The second phase in particle removal is to remove the catalyst from these hoppers. Several methods to remove catalyst from hoppers are offered, including gravity drop out systems, screw conveyor systems and pneumatic/vacuum transfer systems.
ESPs can be very efficient but require more operating attention to keep the efficiency high while also having a risk of exploding if the FCC flue gas stream experiences a high dose of combustible gases.
Third-stage separators. Third-stage separators (TSSs) provide another stage of cyclonic separation downstream of the regenerator in addition to the two stages of cyclones typically included for catalyst containment within FCC regenerator vessels. FCCU flue gas TSSs were developed by Shell Oil in the 1950s to protect flue gas power recovery expanders from erosion by catalyst fines in the flue gas exiting the regenerator. Over the years, TSS dust concentrations at the outlet of the TSS have been measured as low as 10 mg/Nm3–20 mg/Nm3. The overall system, including catalyst not captured in an underflow separator, has provided an FCC stack flue gas dust content as low as 30 mg/Nm3, with 50 mg/Nm3 being nominally considered good performance for a system with an underflow separator.16,17 Troubleshooting TSS particle capture efficiency issues can be an expensive and time-consuming process.
Flue gas scrubbing. On a long-term basis, FCCU flue gas scrubbing systems have demonstrated the ability to remove particulates to mandated levels. In addition, flue gas scrubbing systems have demonstrated SO2 removal efficiencies in as high as 90%. Operating experience has shown that day-to-day changes in flue gas rate, composition, solids loading, temperature, etc. can readily be handled with small changes in the flue gas scrubber operating conditions.
An appropriately designed flue gas scrubbing process meets refinery flue gas particulate and SOx emissions limits.18
FIG. 4 shows a schematic of an ExxonMobil wet gas scrubber first commercialized in 1974.19
Other licensed FCC flue gas scrubbers operate on similar principles. The flue gas enters the scrubbers where intensive contact between the gas and liquid removes both the particulates and SOx. Particulate capture occurs by inertial impaction of the liquid droplets with particles in the gas stream. SOx removal occurs by a reaction with a well-known sulfite buffer in the liquid. Thus, the system provides a single step removal of both SOx and particulate pollutants.
The clean gas is separated from the “dirty” liquid in the separator drum. The cleaned gas then exits to the atmosphere through a stack. The scrubbing liquid is regenerated by direct addition of a sodium-based chemical to the scrubber liquor and recycled back to the scrubbers. Water is lost through evaporation and liquid purge is also made up. A liquid stream may be purged from the scrubber disengaging drum to maintain an equilibrium concentration of solids and dissolved salts (products of SOx removal) within the system. The purge stream can be further treated in a purge treatment unit (PTU) to reduce its chemical oxygen demand (COD) and total suspended solids content to environmentally acceptable levels. Solids are often captured in a simple roll-off bin.
Flue gas NOx controls. In general, NOx are generated either from thermal oxidation of nitrogen in the combustion air, which is known as thermal NOx, or by oxidation of organically bound nitrogen found in a fuel known as fuel NOx. In the FCC process, fuel NOx is produced in the regenerator as a result of burning nitrogen contained in coke that originates from the FCC feed.20 Very little thermal NOx is generated in the FCC process due to the relatively low operating temperatures in the fluid bed. The NOx species present in the regenerator are mostly in the form of NO and NO2 with a higher proportion of NO. The factors that affect NOx generation in the FCCU regenerator include flue gas oxygen content, carbon on regenerated catalyst, regenerator design, combustion/particle temperature, concentration of nitrogen in coke and FCC additives such as CO promoters and SOx additives.
Current methods for controlling the NOx from FCC regenerator flue gas can be grouped into the following two classifications:
Post regeneration technologies such as SCR and SNCR.
Source control technologies such as catalyst additives, feed hydrotreating and counter-current regeneration which lower the amount of NOx produced in the FCC regenerator.
SCR. SCR technology is commercially proven for reducing NOx in FCC regenerator flue gas and involves the reaction of NH3 with NOx in the presence of oxygen and catalyst. The catalyst, depicted in FIG. 5, may be vanadium pentoxide/titanium dioxide-based.21 Other catalysts based on precious metals (platinum or palladium) or zeolites can also be used as SCR catalyst. SCRs can operate in the temperature range between 300°F and 1,100°F (149°C and 593°C)21,22 depending on the catalyst [preferably 600°F–750°F (316°C–399°C) for vanadium pentoxide/titanium dioxide catalyst] and achieve > 90% NOx removal efficiency. A NH3/NOx molar ratio of 1 or slightly higher is commonly used in SCR systems.
The reactions between NOx and NH3 on the SCR catalyst are shown in Eqs. 1 and 2:
4 NH3 + 4 NO + O2 → 4 N2 + 6 H2O (1)
4 NH3 + 2 NO2 + O2 → 3 N2 + 6 H2O (2)
The first reaction is the conversion of NO to nitrogen and the second reaction is the conversion of NO2 to nitrogen. One mole of NH3 is required to convert one mole of NO, whereas two moles of NH3 are required to convert one mole of NO2. This means that as the NO2 concentration in the flue gas increases, the amount of NH3 required will increase.
There is usually sufficient oxygen in the flue gas without the need to supply additional oxygen. Another important reaction is the oxidation of SO2 to sulfur trioxide (SO3) (Eq. 3):
2SO2 + O2 → 2SO3 (3)
This reaction is reversible. The SO2 conversion to SO3 is a function of temperature and the SCR catalyst formulation [vanadium pentoxide (V2O5) content]. Regardless of the catalyst formulation, the SO2 conversion increases with temperature in the range of interest.
Unreacted NH3 leaving the SCR reacts with SO3 to form ammonium sulfate and bisulfate that deposit on downstream equipment. The key reactions for the formation of ammonium bisulfate and ammonium sulfate are shown in Eqs. 4 and 5 and data describing their formation as a function of temperature are presented in FIG. 6.23
NH3 + SO3 + H2O → NH4HSO4 (4)
2NH3 + SO3 + H2O → (NH4)2SO4 (5)
Ammonium sulfates deposit on surfaces below 450°F (232°C) and increase particulate emissions.24 Ammonium sulfate is a dry particulate matter that contributes to plume formation. Ammonium bisulfate is highly acidic and sticky substance that deposits on downstream equipment, such as convection coils and air heaters or economizers, resulting in pluggage and deterioration of equipment performance.24 Keeping ammonia slip low and monitoring down-stream flue gas temperature can minimize deposit formation.
The SCR is usually installed downstream of the waste heat boiler either before or after the electrostatic precipitator. In either case, the waste heat boiler must be modified by removing the economizer tubes or by providing hot gas bypass around it to maintain the desired flue gas temperature to the SCR.
NH3 slip is a term used to describe the amount of NH3 escaping unreacted from the reaction zone in the flue gas. The most important parameters considered for the design of the SCR are the interdependence between NOx reduction, NH3 slip and the catalyst volume. The required volume of catalyst increases with the design NOx removal efficiency, and for a given volume of catalyst. Careful consideration must also be given to design catalyst life and overpressure protection for the SCR.
SNCR. SNCR is used to reduce the FCCU flue gas NOx concentrations by reacting the FCCU flue gas with NH3 at temperatures much higher than needed in SCR systems.25
The process is accomplished by burning CO or fuel gas in the flue gas steam downstream of the FCCU regenerator cyclone system such as may be associated with an in-line burner or CO boiler.26 The NOx removal efficiency increases with NH3 slip, as shown in FIG. 7.23 See FIG. 8 for the effect of temperature on NOx reduction.
FCCU regenerator designs for low NOx emissions. The FCC regenerator design also plays an important role in NOx emissions because the percentage of nitrogen in coke converted to NOx varies widely with regenerator design and operation. Regenerator designs that can control afterburning with less reliance on high excess oxygen levels and higher use of CO promoters will tend to have lower NOx emissions. Also, designs where coke-rich incoming spent catalyst distributed counter-currently near the top of the regenerator fluid bed promote the reduction of NOx to nitrogen in the fluid bed according to the reaction mechanism in Eq. 6:
2C + 2NO → 2CO + N2 (6)
In this process, it was reported that 5% of nitrogen in coke on catalyst was converted to NOx compared to 10%–20% in some other regenerator designs.2
CATALYST ADDITIVES FOR REDUCING FCCU FLUE GAS EMISSIONS
A number of environmentally focused FCC catalyst additives have been introduced over the years which include additives targeting CO, SOx and NOx. These additives have physical properties similar to FCC catalysts and are normally separately added into the circulating FCCU catalyst inventory.
Platinum-based CO combustion promoters were introduced after the commercialization of high-temperature regeneration. These additives promoted the burning of CO in the dense phase bed, which minimized CO combustion in the regenerator dilute phase, cyclones and flue gas system. This avoids damaging high temperatures that would damage equipment while reducing CO emissions to the atmosphere. It was observed that the platinum-based CO combustion promoters significantly increased FCC NOx emissions.27
SOx reducing catalyst additives were developed to minimize SOx emissions. These additives functioned by capturing sulfur on the SOx additive in the FCCU regenerator and then releasing the sulfur in the FCCU reactor stripper as H2S.
Takeaways regarding FCCU regenerator flue gas pollution controls. As evident from the preceding comments, tradeoffs are required in terms of optimizing regenerator flue gas NOx and CO emissions together with regenerator bed temperature, afterburning in the regenerator dilute phase and flue gas excess O2.28
With respect to the combination of both hardware and catalyst options, since the inception of the FCC process until the present there have been continual and dramatic reductions of pollutants in the regenerator flue gas components CO, SOx, NOx and PM.
The proper choice of technology to comply with environmental requirements is greatly influenced by the specifics of the application and the overall goals of the facility.
What might be a great option for one facility may not work for another. TABLE 1 summarizes the relative attributes of the FCC regenerator flue gas control technologies discussed in this paper and provides insight into the pace of improvements in pollution control.
Part 2 of this article will appear in the November issue, and will discuss the impact of rising atmospheric CO2 concentrations on plant growth and food production, as well as the impact on society from the widespread use of fossil fuels. Examples announced this year in U.S. government policy will be included, showing reduced U.S. government investment in carbon capture and other climate-/CO2-related projects. HP
Literature Cited available upon request.
Following graduation with a chemical engineering degree in 1980 from California State Polytechnic University (U.S.), Phillip Niccum worked for Texaco for 9 yrs. He performed process design calculations and technical services for Texaco-owned and licensed FCCUs worldwide for 4 yrs, and then coordinated and developed FCCU technology, catalyst selection and operations in six Texaco U.S. refineries for 5 yrs.
At M.W. Kellogg Co./KBR, beginning in 1989, he served for 26 yrs executing FCC projects, technology development and licensing, including 8 yrs as FCC Chief Technology Engineer and 4 yrs as Director of FCC Technology. More recently, Niccum served as FCC Technology Adviser at McDermott/Lummus in support of FCC projects and technology development. Throughout his career, he has been granted 17 U.S. Patents and authored numerous papers and publications for major industry conferences and trade journals.