Cracking is the conversion of relatively large hydrocarbon molecules into smaller molecules by breaking carbon–carbon bonds. In industrial practice, the term usually refers to processes that transform heavy petroleum fractions or light hydrocarbon feedstocks into transportation-fuel components, gaseous fuels, and raw materials for chemical manufacture. Cracking changes molecular composition, unlike distillation, which separates substances principally according to differences in volatility.
The principal industrial families are thermal cracking, catalytic cracking, and hydrocracking. Steam cracking, a high-temperature thermal process, is especially important for producing ethylene and other unsaturated chemical feedstocks. These processes differ in reaction mechanism, operating conditions, product distribution, and the way heat and hydrogen are supplied. Their development made it possible to obtain substantially more useful light products from petroleum than simple distillation could provide, while also establishing the material basis for much of the modern plastics and synthetic-chemicals industry.
Molecular chemistry and reaction selectivity
Cracking involves the cleavage of a carbon–carbon covalent bond, but an industrial cracking process is not a single chemical reaction. It is a network of bond-breaking, hydrogen-transfer, rearrangement, and condensation reactions. An illustrative overall equation is
This balanced equation represents the formation of an alkane and an alkene from a larger alkane; it does not specify a unique molecular pathway or imply that only these two products form. Real feedstocks contain numerous compounds, and even a pure hydrocarbon generally produces a mixture under cracking conditions.
Thermal cracking proceeds mainly through free-radical chemistry. At sufficiently high temperature, homolytic bond cleavage produces fragments with unpaired electrons. These radicals undergo hydrogen abstraction and beta scission: a bond in the beta position relative to the radical center breaks, producing an alkene and a new radical. Radical recombination and disproportionation provide termination routes. Because chain propagation can generate many different fragments, residence time and the speed of cooling strongly influence the final product distribution.
Acid-catalyzed cracking instead involves ionic pathways commonly described using carbenium-ion intermediates and carbonium-ion transition states or intermediates. A carbenium ion is a positively charged carbon species with three bonds at the charged center. Protonation of an alkene, hydride transfer, and reactions of alkanes on strong acid sites can initiate the reaction network. Subsequent beta scission, skeletal rearrangement, and hydrogen transfer explain the substantial production of branched hydrocarbons in catalytic cracking. The details depend on molecular structure, catalyst acidity, and confinement inside catalyst pores.
Hydrocracking combines cracking on acid sites with hydrogenation and dehydrogenation on a second catalytic function. Hydrogen supplies the atoms needed to saturate many of the fragments. For a simplified alkane split,
The equation illustrates why hydrocracking can produce predominantly saturated liquids rather than the alkene-rich products characteristic of severe thermal cracking. Actual hydrogen consumption also reflects aromatic-ring saturation and removal of sulfur and nitrogen.
Cracking is commonly endothermic overall, and high temperatures can favor the formation of more numerous, smaller molecules through the entropy contribution to Gibbs free energy. Nevertheless, equilibrium alone does not predict industrial yields. Competing reaction rates, catalyst accessibility, and rapid withdrawal of products are decisive. Excessive severity promotes secondary conversion of valuable liquids or alkenes into dry gas, aromatic compounds, and carbon-rich deposits known as coke. Selectivity therefore means controlling not merely how much feed reacts, but which reaction sequences occur before products leave the reactor.
Thermal cracking, visbreaking, and coking
Thermal cracking uses heat rather than an intentionally added cracking catalyst to drive molecular decomposition. Its operating conditions span a broad range because different processes pursue different objectives. Mild treatment can reduce the viscosity of a heavy residue; more severe treatment can maximize lighter liquids while deliberately rejecting some feed carbon as solid coke. These processes remain useful for materials whose high metal, asphaltene, or carbon-residue content would cause difficulties in more catalyst-dependent units.
Visbreaking is a relatively mild thermal treatment of atmospheric or vacuum residue. Typical reaction temperatures are approximately 430–500 °C, although furnace-coil and soaker designs use different combinations of temperature and residence time. A coil visbreaker supplies most of the reaction severity in the heated tube; a soaker design adds an insulated vessel that permits longer reaction at a lower furnace-outlet temperature. Breaking a limited proportion of large molecules lowers residue viscosity and can reduce the quantity of lighter blending oil needed to make a specified fuel oil.
Visbreaking severity is limited partly by product stability. Heavy residues contain asphaltenes dispersed within a complex liquid medium. Conversion can change the solvency of that medium and produce less soluble material, causing sediment formation during storage or blending. Consequently, a larger apparent conversion is not always operationally useful. The process must preserve acceptable stability as well as viscosity, while controlling furnace fouling and gas formation.
Delayed coking applies greater severity and separates the heating step from most of the coke formation. Residual oil passes through a furnace, commonly reaching roughly 480–510 °C, and then enters a large coke drum. Cracking and condensation continue in the drum: vapors leave for fractionation, while petroleum coke accumulates. Multiple drums allow one vessel to receive feed while another is cooled, opened, and mechanically decoked, often using high-pressure water. “Delayed” refers to delaying substantial coke deposition until after the feed leaves the furnace, not to an absence of reaction in the heater.
Fluid coking and related fluidized processes handle coke as circulating particles rather than as a fixed mass in drums. In some configurations, part of the coke is burned to supply process heat; integrated gasification variants convert much of it into fuel gas. Coking produces gas, naphtha, gas oils, and coke, but its liquid products commonly require further treatment because they contain sulfur, nitrogen, and unsaturated compounds. Petroleum coke quality depends strongly on the feed. Low-contaminant coke can have specialized industrial uses, whereas high-sulfur or metal-rich coke has more restricted outlets. Thermal residue processing thus differs fundamentally from hydrocracking: it often raises the hydrogen-to-carbon ratio of the liquid products by rejecting carbon rather than by adding hydrogen.
Fluid catalytic cracking
Fluid catalytic cracking (FCC) converts heavy gas oils and, in suitable designs, some residual feed into gasoline-range material, liquefied petroleum gas, lighter gases, and cycle oils. Its defining feature is the continuous circulation of finely divided catalyst between a cracking reactor and a regenerator. The catalyst behaves as a fluidized solid when contacted with gas, allowing rapid mixing and efficient transfer of heat. FCC generally operates without a high-pressure external hydrogen supply.
In a typical modern riser reactor, preheated feed is injected into a stream of hot regenerated catalyst. The oil vaporizes and reacts at temperatures commonly around 500–550 °C, with contact times of only a few seconds. Rapid separation of catalyst from product vapors limits unwanted secondary reactions. Cyclones remove entrained solids, and a stripper uses steam to displace hydrocarbons retained on the spent catalyst before it enters the regenerator.
Coke deposited during cracking blocks active sites and reduces catalyst performance. In the regenerator, air burns off this coke, restoring much of the catalyst’s activity and heating the particles for another pass through the reactor. Regeneration temperatures are often about 650–750 °C, depending on design and operating conditions. The reactor’s heat demand and the regenerator’s heat release are therefore tightly coupled. Changes in feed quality, coke yield, or catalyst circulation affect the thermal balance of the entire unit.
Modern FCC catalysts normally contain zeolite Y or a modified form of it, dispersed within a porous matrix together with binder and other components. Zeolites are crystalline aluminosilicates whose frameworks provide molecular-scale pores and acid sites. The matrix helps accommodate and partially convert molecules too large to enter the zeolite pores directly. Rare-earth exchange and stabilization treatments alter activity and stability. Additives containing the medium-pore zeolite ZSM-5 can increase light-alkene production, particularly propylene, by selectively converting suitable gasoline-range molecules.
Product selectivity results from both catalyst properties and operating conditions. FCC gasoline contains branched hydrocarbons, alkenes, and aromatic compounds, while the liquefied petroleum gas fraction can be rich in propylene and butenes. Light cycle oil contains substantial aromatic material and is not equivalent to finished diesel fuel. Feed contaminants also matter: nickel can promote dehydrogenation and increase hydrogen and coke formation, while vanadium can damage zeolite structure under regeneration conditions. Nitrogen compounds can suppress acid-site activity.
Catalyst is withdrawn and replaced continuously or periodically to maintain an acceptable equilibrium inventory. This distinguishes commercial FCC operation from an idealized experiment with fresh catalyst: the circulating solid contains particles of different ages, contaminant loadings, and hydrothermal histories. Unit performance therefore depends on catalyst management as well as on reactor design.
Hydrocracking and the role of hydrogen
Hydrocracking converts hydrocarbons in the presence of hydrogen over catalysts that combine acid and hydrogenation functions. It is especially important for producing jet-fuel and diesel-range components from vacuum gas oils, cycle oils, and other heavy feeds. Depending on feed and catalyst, representative conditions are approximately 300–450 °C and 70–200 bar, although commercial processes extend beyond these ranges. High hydrogen partial pressure suppresses coke formation and promotes saturated products.
A bifunctional catalyst enables a sequence of reactions that neither catalytic function would perform as effectively alone. Metal or metal-sulfide sites dehydrogenate saturated hydrocarbons to more reactive intermediates. Acid sites then promote rearrangement and carbon–carbon bond cleavage, after which hydrogenation stabilizes the products. Industrial catalysts may use sulfided nickel–molybdenum or nickel–tungsten components, or noble metals for suitably clean feeds. Their supports can include amorphous silica–alumina, zeolites, and combinations selected for the desired product range.
Feed preparation is often essential. Hydrotreating removes sulfur and nitrogen compounds and can saturate aromatic rings before extensive cracking. Sulfur is converted largely to hydrogen sulfide and nitrogen to ammonia. Basic nitrogen compounds can strongly inhibit the acid function of a hydrocracking catalyst, while metals and heavy deposits may cause more persistent damage. Guard beds, staged catalyst systems, and separate pretreatment reactors help manage these effects. Although hydrotreating and hydrocracking are related, hydrotreating primarily changes contaminants and saturation, whereas hydrocracking deliberately reduces molecular size.
Hydrocrackers may use one-stage or two-stage arrangements and may operate once through or recycle unconverted heavy oil. Fractionation removes products in the desired boiling ranges, while heavier material can return for additional conversion. Recycling raises overall conversion but also changes hydrogen demand, equipment loading, and the exposure of refractory compounds to the catalyst. The choice between maximizing naphtha, middle distillates, or unconverted oil depends on catalyst selectivity, operating severity, and the wider refinery configuration.
Temperature control is particularly important because hydrogenation reactions release heat even though carbon–carbon cracking itself often requires heat. Commercial reactors commonly contain several catalyst beds with intermediate cooling by injection of hydrogen-rich quench gas. Hydrogen consumption depends on feed aromaticity, contaminant content, and the depth of conversion, not simply on the number of broken carbon–carbon bonds.
Residue hydrocracking presents additional challenges because very large molecules, metals, and asphaltenes can plug pores or deactivate catalysts. Ebullated-bed and slurry systems improve contact and permit catalyst management under these demanding conditions. Their operating principles differ from conventional fixed beds, but the underlying objective remains the same: produce lighter hydrocarbons while retaining more feed carbon in useful products through hydrogen addition.
Steam cracking and petrochemical production
Steam cracking is a thermal process designed primarily to produce light alkenes, especially ethylene and propylene, rather than finished transportation fuels. Feeds range from ethane and propane to naphtha and heavier liquid fractions. Hydrocarbon and dilution steam pass through externally heated furnace coils, where very high temperatures and short residence times promote free-radical decomposition. Typical coil-outlet temperatures are approximately 780–900 °C, with residence times commonly measured in tenths of a second.
The steam is principally a diluent, not the stoichiometric reagent responsible for splitting the hydrocarbon. It lowers hydrocarbon partial pressure and helps limit some secondary reactions and coke deposition. This distinguishes steam cracking from steam reforming, which reacts hydrocarbons with water to produce hydrogen-rich synthesis gas. For ethane, the simplest net reaction is
The actual radical network also produces methane and other compounds. With liquid feeds, the mixture becomes much more complex because many molecular structures are present at the furnace inlet.
Feed composition strongly determines the product slate. Ethane gives a relatively high ethylene yield and comparatively few heavier coproducts. Propane and butanes produce broader mixtures, while naphtha cracking yields ethylene, propylene, a C4 fraction containing butadiene, and aromatic-rich pyrolysis gasoline. These differences make feed selection a chemical and commercial decision: two crackers producing the same quantity of ethylene may supply very different quantities of other chemicals. Yield comparisons must specify whether percentages refer to feed mass, carbon, or another basis.
The furnace effluent must be cooled rapidly to preserve the desired products. Transfer-line exchangers recover heat, commonly by generating high-pressure steam, before further cooling and separation. Without rapid quenching, valuable alkenes would continue reacting to form smaller gases, heavier compounds, or coke. The downstream plant then compresses, treats, dries, and separates the cracked gas. Low-temperature fractionation and specialized separation sequences recover polymer-grade ethylene and propylene. Selective hydrogenation removes trace acetylene or related unsaturated impurities that would interfere with subsequent processing.
Coke accumulates inside furnace coils, reducing heat transfer and increasing pressure drop. Furnaces therefore require periodic decoking, usually with carefully controlled steam and air treatment. Coil alloys, surface condition, feed composition, and operating severity all affect run length. Raising severity can increase conversion of the feed but may lower the yield of a particular desired coproduct or accelerate fouling. A steam cracker is consequently optimized as an integrated reaction-and-separation system, not merely as a high-temperature furnace. Its products feed industries making polyethylene, ethylene oxide, ethylene glycol, polypropylene, synthetic rubber, and numerous other materials.
Feed characterization, product recovery, and refinery integration
A cracking feed is usually a petroleum fraction or a blended stream rather than a single chemical. Describing it requires more than an average boiling point. Important measurements include density, boiling-range distribution, sulfur and nitrogen content, metal concentration, aromaticity, and carbon-residue tendency. Heavy fractions may also contain asphaltenes and entrained solids. These properties determine vaporization behavior, hydrogen requirement, coke formation, catalyst life, and the suitability of particular reactor technologies.
Feed names can obscure substantial compositional differences. Vacuum gas oil is recovered by distilling atmospheric residue under reduced pressure, but its quality depends on the original crude oil and the cut points used. Straight-run gas oil differs from gas oil already exposed to thermal or catalytic conversion. Coker gas oil, for example, may contain more unstable unsaturated species and require additional treatment. Similarly, a light, paraffinic naphtha and an aromatic-rich naphtha of comparable boiling range behave differently in a steam cracker.
After reaction, separation converts a complex product mixture into usable streams. In FCC service, the main fractionator separates products into gas, gasoline-range liquid, cycle oils, and heavy bottoms. Gas recovery systems separate fuel gas and liquefied petroleum gas, while further fractionation can recover propylene-rich material. Hydrocracker effluent normally enters high-pressure separation equipment so that hydrogen-rich gas can be recovered and recycled. Liquid products then undergo stabilization and fractionation. Acid gases require dedicated treatment rather than being treated as ordinary hydrocarbon coproducts.
Analytical methods connect molecular composition to operating decisions. Gas chromatography measures light gases and many liquid components; simulated distillation estimates boiling-range distributions. Mass spectrometry and nuclear magnetic resonance provide additional information about hydrocarbon classes and molecular structure, particularly in complex research samples. Sulfur, nitrogen, and metals are often measured with dedicated elemental techniques. No single measurement fully predicts cracking behavior, so industrial feed evaluation combines laboratory assays, pilot testing, and plant experience.
Refinery integration determines which cracking route is most useful. FCC produces gasoline components and light alkenes without a large external hydrogen requirement, but also generates coke and aromatic cycle oils. Hydrocracking can produce highly saturated middle-distillate components, but requires hydrogen and high-pressure equipment. Coking accepts very heavy feed while producing a solid carbon-rich stream. These are complementary capabilities rather than interchangeable versions of one machine.
Conversion products are generally intermediates, not automatically market-ready fuels. FCC gasoline may need sulfur removal; hydrocracked naphtha may need further processing to become a suitable gasoline component; cracked middle distillates may need blending or additional treatment. A refinery’s final product properties emerge from the combination of conversion, purification, separation, and blending.
Historical development
Cracking developed in response to a mismatch between the products obtained by crude-oil distillation and the products increasingly demanded by transport and industry. Early petroleum refining emphasized illuminating oils. The expansion of gasoline-powered vehicles increased demand for lighter fractions that many crude oils could not supply in sufficient quantity through distillation alone. Heating heavy oil to obtain lighter material was known before modern industrial cracking, but reliable large-scale operation required control of pressure, residence time, fouling, and equipment strength.
An important early milestone was the thermal-cracking process patented in 1891 by Vladimir Shukhov and Sergei Gavrilov in the Russian Empire. In the United States, William Merriam Burton and his colleagues developed a pressure thermal-cracking process that entered commercial service at Standard Oil of Indiana in 1913. Burton’s process increased gasoline production from heavier fractions and became widely influential. Later thermal systems, including Dubbs processes, improved continuous operation and heat handling. Commercial adoption depended as much on workable equipment and operating cycles as on recognition that hydrocarbons could decompose when heated.
Catalytic cracking advanced through the work of Eugène Houdry and collaborators. A commercial Houdry unit began operation at Sun Oil’s Marcus Hook refinery in 1937. Early Houdry systems used fixed catalyst beds that alternated between reaction and regeneration, because carbon deposits rapidly reduced catalyst activity. Catalysis improved gasoline yield and quality relative to contemporary thermal processes, but cyclic fixed beds imposed limitations on continuous throughput and heat management.
Fluid catalytic cracking resolved many of these limitations by circulating powdered catalyst continuously between reaction and regeneration zones. The first commercial FCC unit began operation at Standard Oil of New Jersey’s Baton Rouge refinery in 1942. The technology emerged from a substantial collaborative development effort involving oil companies, engineering organizations, and catalyst specialists. During the Second World War, catalytic petroleum conversion contributed to the supply of high-performance fuel components and chemicals needed for synthetic rubber.
The introduction of zeolite-containing cracking catalysts in the early 1960s marked another major transition. Their strong acidity and crystalline pore structures improved activity and selectivity compared with the amorphous silica–alumina catalysts previously used. Subsequent development refined zeolite stability, matrix properties, catalyst accessibility, and resistance to contaminants. FCC reactor designs also moved toward shorter contact times and faster product–catalyst separation.
Modern high-pressure hydrocracking expanded commercially from the late 1950s, building on earlier hydrogenation research and improvements in catalysts and pressure-vessel technology. Steam cracking grew alongside demand for ethylene-based chemicals and polymers. These developments did not eliminate older thermal methods: delayed coking and visbreaking retained important roles because heavy residues require different solutions from clean gas oils or light petrochemical feeds.
Kinetics, reactor design, and process control
The design of a cracking reactor requires both chemical kinetics and transport analysis. Reaction rates increase strongly with temperature, commonly represented over a limited range by the Arrhenius expression,
where is a rate constant, a pre-exponential factor, the apparent activation energy, the gas constant, and absolute temperature. Industrial feedstocks contain too many reacting species for a single rate constant to describe the process. The equation is therefore applied to individual model reactions or to empirically defined groups of compounds.
“Lumped” models combine molecules into categories such as feed gas oil, gasoline, light gas, and coke, then describe conversion between those groups. More detailed models distinguish boiling intervals, hydrocarbon families, or individual molecular structures. Their usefulness depends on the available analytical data and the operating range over which parameters were fitted. A model calibrated for one feed and catalyst may predict poorly after major changes in aromaticity, contaminants, or catalyst age. Mechanistic detail does not automatically eliminate uncertainty in feed characterization.
Reaction and transport occur on comparable timescales in many cracking systems. In FCC, liquid droplets must heat, vaporize, and contact catalyst before the desired short reaction time expires. Within catalyst particles, diffusion determines how quickly molecules reach active sites and products escape. Large molecules may react mainly in the matrix or near pore entrances. In hydrocracking, hydrogen must pass from the gas phase into the liquid and then reach catalytic surfaces. In steam cracking, heat transfer through the furnace tube wall constrains the temperature profile and permissible throughput.
Residence-time distribution matters as much as nominal residence time. Backmixing or poorly separated catalyst can expose some products to excessive reaction even when average conditions appear correct. Rapid disengagement in FCC and rapid quenching in steam cracking address this problem in different ways. Computational fluid dynamics is used to study injector behavior, gas–solid flow, mixing, and temperature nonuniformity, but such calculations require validation against experiments and plant measurements.
Performance metrics must also be defined carefully. Conversion may mean disappearance of a feed component, material boiling above a specified temperature, or a chosen group of heavy fractions. Yield can be reported by mass, volume, or moles; these are not interchangeable. Selectivity usually relates a desired product yield to the quantity converted. Liquid-volume yields can exceed the original feed volume because products have lower density, without violating conservation of mass.
Process control coordinates feed rate, reactor temperature, catalyst circulation, hydrogen flow, recycle, and product separation. In FCC, reactor and regenerator heat balances constrain one another. In hydrocracking, bed temperatures and hydrogen availability are central variables. In steam cracking, coil-outlet conditions must be considered alongside tube-metal temperature, pressure drop, and downstream separation capacity.
Energy use, emissions, and emerging feedstocks
Cracking processes are major energy users because they heat large material flows, perform demanding separations, and, in some cases, compress gases to high pressure. Their environmental burdens differ by process. Steam-cracking furnaces require intense heat input; FCC regenerators release carbon dioxide when coke is burned; hydrocracking requires hydrogen whose production can carry substantial emissions. Comparisons therefore depend on the system boundary, feedstock, product slate, and the source of heat, electricity, and hydrogen.
FCC regeneration also creates emissions-control challenges beyond carbon dioxide. Feed sulfur can contribute sulfur oxides, feed nitrogen and combustion conditions influence nitrogen oxides, and catalyst particles can appear in flue-gas particulate matter. Cyclones, additional particulate controls, sulfur-control additives, and flue-gas treatment address different parts of this problem. Carbon monoxide may be oxidized within the regenerator system or in a separate boiler, depending on the design. Hydrocracking and thermal units also produce sour gas and sour water requiring dedicated treatment.
Hydrogen creates an important trade-off. Adding it can preserve more feed carbon in liquid products and produce low-sulfur, saturated fuel components, but conventional hydrogen production commonly releases carbon dioxide. Hydrogen made using lower-emission electricity or combined with carbon capture changes the balance, though the outcome depends on actual production and energy conditions. Similarly, electrifying steam-cracker furnaces can reduce direct combustion emissions without eliminating emissions associated with electricity generation or subsequent use of the products.
Interest in alternative feedstocks includes waste-plastic-derived oils and renewable hydrocarbon streams. Pyrolysis of polyolefin-rich waste can produce hydrocarbon mixtures that may be further cracked after treatment. Such oils are not automatically equivalent to conventional naphtha: chlorine, oxygenates, nitrogen compounds, metals, and unstable components can cause corrosion, fouling, catalyst damage, or product-quality problems. Feed sorting, pretreatment, and specification control are therefore central technical constraints. Cracking mixed waste also differs from selectively recovering an original monomer through depolymerization.
Renewable feeds introduce other reaction requirements. Lipids and biomass-derived liquids can contain oxygen that must leave as water, carbon monoxide, or carbon dioxide during processing. The resulting hydrogen demand and carbon efficiency depend on the pathway used. Biological origin alone does not determine compatibility with an FCC unit, hydrocracker, or steam cracker.
Research addresses lower-coking surfaces, more selective catalysts, improved heat recovery, electrified heating, and better handling of difficult feeds. Environmental accounting remains sensitive to coproduct allocation and to the fate of the products. Carbon retained temporarily in a polymer is treated differently from carbon emitted promptly when a fuel is burned, while recycling claims depend on material losses and the share of recovered feed that actually returns to chemical production.