Ascent Petrochem Holdings Co., Limited
Articles
Articles

Articles

Furnace Run Length Penalties During C5 Raffinate Cocracking with Naphtha

On a naphtha-fed liquid steam cracker, the run length between steam-air decoking cycles is not a fixed feed property but a response to radiant coil coke accumulation, transfer line exchanger fouling, and tube metal temperature limits. The coke layer formed on the inner surface of centrifugal cast HP40Nb or microalloy 35Cr-45Ni-Nb radiant tubes raises the tube metal temperature under constant heat flux; a tube skin temperature rise from **950°C** to **1080°C** at the same process gas outlet condition typically indicates an insulating coke layer thickness in the range of **0.5 mm to 1.5 mm**. When hydrotreated C5 raffinate is blended into straight-run naphtha at **5 wt% to 15 wt%**, the blended feed no longer behaves as a single liquid with uniform decomposition kinetics. The raffinate contributes C5 paraffins, mono-olefins, residual cyclopentene, and very low sulfur if it has passed through first-stage pyrolysis gasoline hydrotreating. This combination increases the olefin-to-paraffin ratio while reducing the sulfur compounds available for catalytic passivation of the coil surface. The run length penalty is therefore not merely proportional to the raffinate flow; it is a non-linear response to changes in radical initiation, coking precursor formation, tube surface chemistry, and quench-boiler tar condensation. Quantitative site-specific correlations must be developed because published data for this specific configuration is limited, although the general mechanisms are established in industrial cracked-gas service.

What Operational Penalties Emerge When C5 Raffinate Displaces Straight-Run Naphtha?

The primary measurable penalty is a reduction in the number of calendar days between decoking operations. Industrial operating systems define end-of-run by either radiant coil tube metal temperature reaching **1090°C to 1120°C**, radiant coil pressure drop exceeding **2.5 bar to 3.0 bar**, or transfer line exchanger outlet temperature rising above **450°C to 500°C** for high-pressure steam generation. A blend containing **10 wt%** C5 raffinate may shift the limiting criterion from tube metal temperature to transfer line exchanger fouling because the lighter raffinate increases cracking conversion at the same coil outlet temperature and simultaneously produces unsaturated C5 radicals that condense into heavy tar in the quench zone. The magnitude of the run length penalty is expressed as ΔRL = (1 − RL_blend/RL_base) × **100%**, where RL_blend and RL_base are measured in days under identical coil outlet temperature, steam-to-hydrocarbon ratio, and feed rate. Under constant severity, operational reports from naphtha crackers with C5 recycle indicate that replacing **10 wt%** straight-run naphtha with hydrotreated C5 raffinate can reduce run length by **10% to 30%** when no compensating adjustment is made. The penalty is larger when the blend increases the reactive olefin concentration beyond **25 wt% to 30 wt%** of feed or when total sulfur falls below **50 mg/kg**. The exact value depends on the base naphtha paraffin, olefin, naphthene, and aromatic distribution, the coil outlet temperature profile, the dilution steam ratio, and the transfer line exchanger design.

Because the run length penalty is measured on production-scale equipment, the choice of radiant coil geometry and transfer line exchanger design modifies the sensitivity to raffinate addition. Split-coil furnaces with shorter residence times and lower hydrocarbon partial pressures often tolerate higher olefin feeds than long-residence-time reactors because the peak olefin yield is reached before C5 intermediates can be converted into polynuclear aromatics. The run length in a furnace with radiant coil mass flux of **200 kg/(m²·s) to 350 kg/(m²·s)** may respond differently from a low-mass-flux coil because the wall film temperature and radical termination reactions are different. Operators measure the response by tracking daily tube skin temperature standard deviation, radiant pressure drop, and transfer line exchanger outlet temperature. An increase in the standard deviation of tube skin temperatures above **20°C to 30°C** is frequently associated with accelerated local coking and an impending run length penalty.

Radiant Coil Coking Mechanisms With Reactive C5 Mono-Olefins

Coke in radiant coils forms by three overlapping mechanisms: thermal cracking of gas-phase hydrocarbons at high temperature, catalytic coking on exposed iron and nickel sites, and condensation of aromatic precursors on the coke surface. Mono-olefins in C5 raffinate, including 1-pentene, 2-methyl-2-butene, and cyclopentene, generate allylic and tertiary radicals that undergo β-scission to produce ethylene, propylene, butadiene, and isoprene. The diolefins formed in this radical chain are not stable at radiant coil temperatures. They react by Diels-Alder cycloaddition with unsaturated C5 and C6 species, yielding cyclic C10 dimers that dehydrogenate to naphthalene and polycyclic aromatic hydrocarbons. The radical addition products deposit on the tube wall as condensed coke or are carried into the transfer line exchanger where thermal degradation continues. The apparent activation energy for coke formation in naphtha cracking has been reported in the range of **180 kJ/mol to 230 kJ/mol**, and the rate is strongly dependent on hydrocarbon partial pressure. The presence of reactive C5 olefins increases the concentration of conjugated diolefin intermediates at the same bulk gas temperature, shifting coke precursor formation further upstream in the coil. Unlike catalytic coking, which is surface-limited and can be passivated by sulfur, the thermal and condensation mechanisms continue as long as the precursor concentration is high.

Tube metal temperature is the immediate feedback variable on a production furnace. Optical pyrometers and retractable thermocouples measure tube skin temperatures at multiple elevations, and the temperature difference between the tube wall and the bulk process gas is used to estimate coke thermal resistance. If the clean tube wall temperature at a given elevation is **930°C** and the end-of-run limit is **1080°C**, a coke-induced temperature rise of **150°C** provides the operational window. The coke layer thermal conductivity is typically an order of magnitude lower than the alloy substrate, so a thin layer of **0.2 mm** can produce a temperature rise of **30°C to 60°C** at radiant heat flux of **80 kW/m² to 120 kW/m²**. Adding C5 raffinate tends to increase the axial non-uniformity of this temperature rise because the raffinate decomposes earlier in the coil and increases local heat demand. The resulting uneven coke deposition raises the maximum tube skin temperature faster than a uniform coke layer would, which is one of the principal causes of shortened furnace run length.

When Feed Sulfur Falls Below the Catalytic Passivation Threshold

Hydrotreated C5 raffinate contains very low sulfur, often below **10 mg/kg**, while straight-run naphtha may contain **200 mg/kg to 800 mg/kg** of sulfur. Sulfur compounds such as mercaptans, sulfides, and thiophenes decompose in the radiant coil and form a stable metal sulfide layer on the tube surface. This sulfide layer blocks the iron and nickel active sites responsible for catalytic coke formation, but the passivation effect is dynamic and depends on maintaining a minimum sulfur partial pressure in the process gas. When a low-sulfur C5 raffinate stream is blended into naphtha, the total feed sulfur concentration may fall below **50 mg/kg to 100 mg/kg**. At these levels, catalytic coking on exposed metal surfaces can become the dominant coke path, even if the thermal coking rate is unchanged. The resulting coke is often filamentous and tightly bound to the tube wall, causing a faster tube metal temperature rise than amorphous thermal coke. In many liquid crackers, dimethyl disulfide injection is used to restore feed sulfur concentration to a target range of **100 mg/kg to 200 mg/kg**. The injection point is usually downstream of the feed preheater and upstream of the convection section to ensure complete mixing without excessive pre-cracking. The dimethyl disulfide flow is controlled by a mass flow controller and trimmed by total sulfur analysis using **ASTM D4294** or on-line XRF.

Continuous dimethyl disulfide injection is preferred over intermittent spiking because the sulfide layer is stable only under reducing process gas conditions. If dimethyl disulfide is injected at a rate equivalent to **50 mg/kg to 150 mg/kg** of sulfur in the combined feed, the decomposition products passivate the coil surface without contributing substantially to SOx emissions in the decoking effluent. The optimum sulfur level is feed-dependent. Excessive sulfur can increase sulfidation and reduce creep ductility of the tube alloy after long-term exposure, while insufficient sulfur accelerates catalytic coke formation. The upper limit for continuous sulfur addition is often set at **200 mg/kg** to avoid excessive sulfidation of HP40Nb tube material at metal temperatures above **1000°C**. This sulfur balance becomes more difficult when the C5 raffinate fraction increases because the low-sulfur raffinate dilutes the natural sulfur present in naphtha and reduces the passivation effect of the base feed.

The TLE Fouls Faster When C5 Diolefin-Derived Tars Condense in the Quench Zone

Transfer line exchangers in naphtha crackers are vertical shell-and-tube heat exchangers with process gas on the tube side and high-pressure steam on the shell side. The process gas entering the transfer line exchanger is rapidly cooled from **800°C to 900°C** to **350°C to 450°C** within **0.02 s to 0.05 s**. This quench is necessary to stop secondary reactions that degrade olefin selectivity, but it also creates a steep temperature gradient that condenses high-molecular-weight aromatic precursors into a viscous tar layer on the tube wall. C5 raffinate cocracking increases the concentration of C5 diolefins and cyclopentadiene-derived species in the cracked gas, and these species are known tar precursors. The foulant layer reduces the overall heat transfer coefficient and causes the transfer line exchanger outlet temperature to rise. Since high-pressure steam temperature is set by the steam drum pressure, the outlet temperature increase works directly against the quench function. A transfer line exchanger outlet temperature increase from **420°C** to **480°C** may reduce the steam superheat and increase the risk of thermal degradation of the cracked gas. The transfer line exchanger run length is therefore not identical to the radiant coil run length, and in some cocracking campaigns the transfer line exchanger becomes the limiting equipment item before the radiant tubes reach maximum metal temperature.

Fouling resistance in a transfer line exchanger is commonly represented as a fouling factor added to the clean overall heat transfer coefficient. The clean heat transfer coefficient in a transfer line exchanger is typically in the range of **800 W/(m²·K) to 1200 W/(m²·K)** for high-pressure steam generation; a fouling factor of **0.0002 m²·K/W** corresponds to a design allowance, while **0.0008 m²·K/W to 0.0015 m²·K/W** indicates severe tar accumulation. When the calculated fouling resistance exceeds the design allowance, the steam generation rate falls and the process gas outlet temperature rises. Operators then reduce feed rate or increase transfer line exchanger blowdown frequency. For C5 raffinate blends with high olefin content, transfer line exchanger fouling often correlates with the concentration of cyclopentadiene and isoprene in the feed rather than total olefins. This is why hydrotreating to remove conjugated diolefins is often applied before C5 raffinate is returned to the cracking furnace.

What Feed Characterization Methods Identify C5 Raffinate Blend Fouling Potential?

Because the run length penalty is linked to olefin, diolefin, and sulfur concentrations, the blend recipe must be supported by periodic laboratory analysis. The analytical methods that are most relevant are **ASTM D86** for distillation, **ASTM D4052** for density, **ASTM D1159** for bromine number, **ASTM D4294** for total sulfur, and **UOP 326** for maleic anhydride value of conjugated diolefins. The table below summarises the typical interpretation of these measurements for a C5 raffinate stream. The values are indicative ranges; published data for specific configurations is limited, and each site should maintain its own feed-quality correlation to furnace run length.

Measurement Method Indicative range for C5 raffinate Fouling relevance
Density at **15°C** ASTM D4052 0.62 g/cm³ to 0.70 g/cm³ Low density indicates high paraffin and olefin content and complete vaporization before the radiant coil.
Final boiling point ASTM D86 50°C to 75°C High final boiling point may indicate heavy C6+ contamination and tar precursor carryover.
Bromine number ASTM D1159 30 g Br/100g to 150 g Br/100g for raw raffinate; <5 g Br/100g after hydrotreating Bromine number correlates with mono-olefin concentration and potential coke precursor formation.
Conjugated diolefins UOP 326 <0.5 wt% after hydrotreating; 0.5 wt% to 5 wt% in raw raffinate Conjugated diolefins dimerize and contribute to transfer line exchanger tar.
Total sulfur ASTM D4294 <10 mg/kg after hydrotreating; 50 mg/kg to 300 mg/kg for raw raffinate Low sulfur may necessitate dimethyl disulfide injection for passivation.

Operating Envelope Adjustments for Cocracking Campaigns

When C5 raffinate is introduced into the feed pool, the furnace can be operated at the same severity and accept a run length penalty, or the operating envelope can be adjusted to partially recover run length. The most common production-scale adjustments are coil outlet temperature reduction, dilution steam ratio increase, feed rate derate, and sulfur supplementation. Reducing coil outlet temperature by **5°C to 10°C** lowers the rate of coke formation because coke formation has a high apparent activation energy relative to olefin yield. However, this also reduces ethylene yield by approximately **0.5 wt% to 1.0 wt%** depending on the base naphtha composition. Increasing dilution steam ratio from **0.35 kg/kg** to **0.45 kg/kg** reduces hydrocarbon partial pressure and suppresses bimolecular condensation to heavy aromatics. The penalty is additional fuel consumption for steam generation and possible capacity limitation in the dilution steam system. A feed rate derate of **5% to 10%** reduces radiant heat flux and residence time, but decreases furnace throughput. Many operators use a combination of all three adjustments rather than a single large change. The table below lists the operational levers that are commonly applied.

Lever Typical adjustment Mechanism Operational limitation
Coil outlet temperature Reduce by 5°C to 10°C Lower radical concentration and coking precursor formation Ethylene yield loss
Dilution steam ratio Increase by 0.05 kg/kg to 0.10 kg/kg Lower hydrocarbon partial pressure; reduced condensation coking Higher energy consumption; transfer line exchanger heat recovery changes
Feed rate Reduce 5% to 10% Lower heat flux and shorter residence time Reduced production
Sulfur injection Maintain 100 mg/kg to 200 mg/kg as dimethyl disulfide Passivates catalytic coking sites Risk of sulfidation; SOx emission
C5 raffinate blend ratio Limit to 10 wt% to 15 wt% Reduces olefin loading and sulfur dilution Recycle disposal constraints

In a production furnace equipped with a transfer line exchanger producing 125 bar steam and a radiant coil outlet temperature limit of 850°C, the introduction of 10 wt% hydrotreated C5 raffinate requires daily monitoring of tube metal temperature and transfer line exchanger outlet temperature. If the tube metal temperature approaches 1080°C before 30 days of operation, the site reduces the C5 raffinate blend to 5 wt% or increases dimethyl disulfide injection to 150 mg/kg sulfur equivalent. If transfer line exchanger fouling is the limiting factor, the feed rate is derated by 5% and the dilution steam ratio is raised by 0.05 kg/kg until the transfer line exchanger outlet temperature returns to 420°C. These constraints define the run length penalty for that specific furnace and cannot be transferred to another cracker without recalibrating against local radiant flux, coil geometry, and quench performance.

Related Articles