In the two-stage vapor-phase oxidation of propylene to acrylic acid, the intermediate acrolein stream is the point where the heat removal capacity of the reactor and the thermal stability of the catalyst intersect. Propylene is first converted over a bismuth molybdate-based catalyst to acrolein, and the acrolein-containing effluent is subsequently oxidized over a molybdenum-vanadium-based catalyst to acrylic acid. The selective oxidation of propylene to acrolein releases approximately 340 kJ·mol⁻¹, and the subsequent acrolein-to-acrylic acid reaction releases approximately 252 kJ·mol⁻¹. Both stages employ multi-tubular fixed-bed reactors cooled by forced-circulation molten salt, but the second-stage acrolein oxidation is particularly constrained by the maximum allowable catalyst peak temperature. Hot spot in this context refers to the highest local temperature measured in a representative catalyst tube by an axial multi-point thermocouple assembly. The acrolein hot spot temperature constraint is therefore the envelope of limits that prevents excessive carbon oxide formation, irreversible catalyst sintering, tube-wall creep damage, and polymerization fouling. A two-stage acrylic acid plant operates economically only when the hot spot is maintained within a narrow band between light-off and destructive oxidation.

The industrial hardware used for both stages consists of vertical catalyst tubes with inner diameters in the range of 25 mm to 38 mm in first-stage service and 21 mm to 31 mm in second-stage service, tube lengths of 2.5 m to 4.5 m, and tube counts between 10,000 and 40,000 in large-capacity trains. Molten salt circulates on the shell side at a bulk temperature of 250 °C to 330 °C. Normal first-stage peak temperatures are generally between 370 °C and 410 °C; normal second-stage acrolein peaks are between 300 °C and 360 °C. The maximum allowable peak is commonly 420 °C to 430 °C in the first stage and 370 °C to 380 °C in the second stage. The narrow margin between the normal second-stage peak and the upper limit is the acrolein hot spot temperature constraint in practice. This constraint is not fixed by a single property but by simultaneous limits on catalyst selectivity, salt stability, metal temperature, and acrolein polymerization propensity.

Why Does the First-Stage Reaction Zone Generate a Critical Hot Spot?

The first-stage reactor receives a preheated mixture of propylene, air, and steam. The propylene concentration is normally limited to 7 vol% to 9 vol% to remain outside the flammable envelope, while oxygen is supplied in slight stoichiometric excess relative to the desired acrolein reaction. The catalyst is a bismuth molybdate-based formulation that provides multiple oxidation states for selective allylic oxidation. The reaction network includes selective oxidation to acrolein, consecutive oxidation to acrylic acid, and parallel total combustion to carbon oxides. Acrolein selectivity is favored within a narrow temperature window because the apparent activation energy for total oxidation is larger than that for selective oxidation; as the local bed temperature rises, the rate of carbon monoxide and carbon dioxide formation increases more rapidly. The hot spot propagates axially because the feed concentration and oxygen partial pressure are highest at the tube inlet, and the heat removal rate is limited by the gas-side film and the catalyst bed effective thermal conductivity. The peak temperature is usually observed between 0.5 m and 1.5 m from the tube inlet, depending on catalyst activity, gas velocity, and salt bath temperature. When the catalyst ages, deactivation of the upper bed shifts the reaction zone downward, and the recorded hot spot at a fixed thermocouple may decrease while the actual peak moves toward the outlet. This migration must be distinguished from true peak reduction by comparing axial temperature profiles at multiple thermowell elevations.

Published kinetic data for specific commercial catalyst formulations is limited because licensors treat rate constants and heat generation profiles as proprietary. However, the general behavior can be inferred from the known reaction stoichiometries and industrial temperature profiles. Propylene conversion in the first stage is typically maintained between 90% and 98%, with acrolein yields of 78% to 88% and acrylic acid yields of 5% to 10% in the first-stage effluent. The remaining carbon leaves as carbon monoxide, carbon dioxide, heavy by-products, and unconverted propylene. The overall heat release includes both the desired partial oxidation and the undesired combustion reactions; complete combustion of propylene releases approximately 1,926 kJ·mol⁻¹, and combustion of acrolein is similarly strongly exothermic. A local hot spot above 430 °C in the first stage accelerates thermal restructuring of the bismuth molybdate phases and can produce permanent selectivity loss. The first-stage peak is therefore constrained not only to maintain acrolein yield but also to preserve the catalyst’s ability to deliver a stable acrolein stream to the second stage.

Molten Salt Bath Equipment and Multi-Tubular Reactor Geometry

Multi-tubular reactors used for acrylic acid service are designed with thousands of parallel vertical tubes supported by tubesheets and surrounded by a molten salt bath. The shell-side salt is typically a ternary mixture of potassium nitrate, sodium nitrite, and sodium nitrate with a melting point near 142 °C and a maximum bulk operating temperature near 450 °C. Salt pumps with capacities from 4,000 m³·h⁻¹ to 10,000 m³·h⁻¹ circulate salt through the shell side, and internal baffles or annular distribution devices direct flow across the tubes. Heat removal from the tube wall depends on the salt-side coefficient, which is generally in the range of 800 W·m⁻²·K⁻¹ to 1,500 W·m⁻²·K⁻¹ for forced circulation nitrate salts. The gas-side coefficient inside a catalyst-packed tube is much lower, typically 100 W·m⁻²·K⁻¹ to 300 W·m⁻²·K⁻¹, and therefore controls the overall heat transfer. The effective overall heat transfer coefficient from the catalyst bed to salt is commonly 60 W·m⁻²·K⁻¹ to 200 W·m⁻²·K⁻¹. A low overall coefficient means that a sudden increase in reaction rate raises the catalyst temperature before the tube wall and salt bath respond. This thermal lag is a key reason why hotspot alarms are often set only 20 K to 30 K above normal peak temperature.

Thermowell assemblies are placed in a subset of tubes, typically 5 to 20 tubes per reactor, with multi-point thermocouples at axial intervals of 0.25 m to 0.5 m. Type K or Type N thermocouples with Inconel 600 or 601 sheaths are common. The measured thermowell temperature approximates the gas temperature inside the thermowell, not the maximum pellet center temperature; the actual intraparticle center temperature is higher by 5 K to 15 K depending on catalyst particle size and effective thermal conductivity. This intraparticle gradient is frequently overlooked but is essential for setting a safe tube wall limit. If the thermowell peak reads 370 °C, the catalyst pellet center may already exceed 380 °C. Tube materials are normally austenitic stainless steels conforming to ASTM A213 TP321 or ASTM A213 TP347H, with tube wall thickness of 2.0 mm to 3.5 mm and design metal temperatures commonly 450 °C to 500 °C. The hot spot limit is therefore set below the tube design temperature, and the thermocouple reading is adjusted for the expected intraparticle offset.

Controlling Tube Wall Temperature Through Forced Salt Circulation

Shell-side salt flow distribution is a major field variable. Poor flow distribution produces tube-to-tube salt temperature differences that appear as scattered hot spot readings across the bundle. Process licensors normally specify the maximum acceptable tube-to-tube peak variation, commonly 15 K to 20 K, with corrective action triggered above 25 K. Salt circulation pumps are usually vertical turbine or cantilever designs with wetted components in low-carbon stainless steel or higher-nickel alloys to resist nitrate stress corrosion. The salt pump discharge pressure is typically in the range of 0.15 MPa to 0.35 MPa to overcome baffling and flow resistance. Production-scale experience has shown that salt circulation pumps are a common bottleneck because a short power interruption can allow salt to stagnate in the lower tubesheet zone and cool below its melting point; restarting with solidified salt inventory is a recognized operational hazard. The hot spot constraint therefore includes a minimum salt circulation velocity, not merely a setpoint temperature. Maintaining flow across the lower tubesheet region is particularly important because the reaction zone can become concentrated there if the catalyst bed has settled or if the upper bed is diluted excessively.

The acrolein second-stage reactor generally operates with a lower salt bath setpoint than the first stage, often 250 °C to 300 °C, because the molybdenum-vanadium catalyst is thermally stable but highly active for total oxidation at elevated temperature. The salt must remain above the gas dew point, which depends on steam partial pressure, to avoid condensation of acrylic acid or maleic acid on the tube walls. At the same time, raising salt temperature beyond 330 °C in the second stage narrows the margin between the salt and the maximum allowable catalyst peak and increases the risk of acrolein polymerization in stagnant boundary layers. Field data from forced-circulation salt systems indicate that a tube-to-tube peak spread above 25 K is typically associated with either baffle erosion or non-uniform inlet nozzle flow, and the usual correction is mechanical salt flow redistribution or partial replacement of damaged baffle plates rather than catalyst reloading.

When Catalyst Beds Are Axially Diluted to Suppress Peak Temperatures

When the first or second stage shows an excessively high hot spot at target conversion, the catalyst bed is usually divided into zones with different activities or dilution levels. The upper bed may be diluted with inert ceramic spheres or rings having particle diameters of 3 mm to 6 mm at 20 vol% to 50 vol% inert. Dilution reduces the volumetric heat release in the inlet zone and allows reactant concentrations to decrease before gases reach undiluted active catalyst. The hot spot is thereby shifted downstream and broadened; the peak temperature may decrease by 5 K to 15 K for a moderately diluted bed at a gas hourly space velocity near 1,500 h⁻¹. Excessive dilution above approximately 60 vol% is usually counterproductive because inert zones increase pressure drop without providing corresponding heat removal and can create radial flow maldistribution. The pressure drop of a diluted bed follows the Ergun relationship, where the pressure drop per unit length is proportional to the square of gas velocity at high flow and inversely proportional to the third power of bed voidage. A decrease in the tube inner diameter from 38 mm to 25 mm lowers the radial resistance to heat transfer and is often preferred before applying aggressive dilution.

Published data for specific commercial dilution patterns is limited because each licensor’s loading diagram is proprietary. However, the underlying transport principles are standard. The effective thermal conductivity of a catalyst bed is typically between 0.2 W·m⁻¹·K⁻¹ and 0.5 W·m⁻¹·K⁻¹, and this low value dominates intrabed heat transfer. The gas-to-wall heat transfer coefficient in a packed tube can be estimated from correlations that account for tube diameter, particle diameter, gas Reynolds number, and bed voidage. A larger catalyst particle increases permeability but lowers external surface area and increases the intraparticle diffusion resistance. Therefore hotspot management through particle size alone involves a compromise between pressure drop and selectivity. In second-stage acrolein oxidation, catalysts are frequently shaped as rings or trilobes with hydraulic diameter of 3 mm to 5 mm, which provide lower pressure drop at the same geometric surface area. Catalyst batches are sampled before loading for equivalent particle diameter and particle crushing strength; a batch drift of 0.3 mm in mean diameter can alter bed voidage by approximately 0.02 and shift the hot spot axial location, requiring a salt bath setpoint adjustment to restore the target peak temperature.

At the second-stage inlet, the combined effluent from the first stage, additional air, and steam create a mixture in which acrolein concentration is typically 5 vol% to 9 vol% and oxygen is 12 vol% to 15 vol%. The second-stage oxidation is highly exothermic, and the local adiabatic temperature rise for complete acrolein conversion can be estimated from the heat of reaction and the gas heat capacity. For a feed containing 6 vol% acrolein and a gas mixture heat capacity near 32 J·mol⁻¹·K⁻¹, the theoretical adiabatic rise exceeds 450 K. This calculation explains why the acrolein hot spot is the limiting parameter in stage two: if the salt bath fails to remove heat from only a small fraction of the tube bundle, the local temperature can approach or exceed the maximum allowable within seconds. The margin between normal peak and interlock is often no more than 10 K to 20 K. Process control therefore relies on feed-forward oxygen trimming, hot spot cascade to salt bath setpoint, and fast oxygen cut when the peak rises above the high-high limit. Flammability limits are also a parallel constraint; propylene and acrolein concentrations are held outside the flammable envelope with the aid of steam dilution and defined limiting oxygen concentration data. Published data for this specific configuration is limited, but the applicable gas mixture flammability framework is addressed in ISO 10156:2017.

Thermal Runaway Boundaries During Acrolein Oxidation Require Redundant Interlocks

The acrolein oxidation stage is protected by a layered instrumented safeguard system. The primary loop is the hot spot temperature controller, which adjusts salt bath setpoint or feed oxygen within a narrow band. A high alarm is typically set at 10 K above normal peak, and a high-high trip at 20 K to 30 K above normal peak. On high-high, the oxygen or air flow to the second stage is automatically reduced or isolated, and inert steam or nitrogen is injected to sweep residual reactants. The trip is implemented in a safety instrumented system conforming to IEC 61511-1:2016, with sensor voting such as two-out-of-three for peak temperature in the hottest tubes. The first-stage and second-stage trips are interlocked with the propylene and air feed valves to prevent formation of a flammable mixture under abnormal conditions. Pressure relief devices for the reactor shell and tube side are specified according to ASME BPVC Section VIII Division 1 or locally applicable pressure equipment codes.

The catalyst tube wall temperature must also stay below the design metal temperature, which is usually 450 °C to 500 °C for austenitic stainless steel tubes. A tube wall hotspot exceeding the design value can lead to creep rupture under internal pressure. The differential pressure between the process gas and the salt bath is monitored; a sudden drop indicates tube leakage, which can inject salt into the catalyst bed and create a severe local exotherm. In practice, a tube leak in the second stage is a serious event because salt contamination accelerates acrylic acid polymerization and can block the tube. The operating team typically reduces feed and begins partial reactor isolation when a hot spot exceeds 370 °C in the second stage or 430 °C in the first stage, even before the high-high trip activates.

What Limits Gas Hourly Space Velocity in the Acrolein Second-Stage Reactor?

The gas hourly space velocity, defined as the volumetric feed flow at standard conditions divided by the catalyst bed volume, is a primary lever for shifting the acrolein hot spot. In the second stage, gas hourly space velocity is usually between 1,200 h⁻¹ and 1,800 h⁻¹. Increasing gas hourly space velocity improves gas-to-wall heat transfer because the Reynolds number increases, but it also increases pressure drop and reduces contact time, requiring a higher bed temperature to maintain conversion. The acrolein conversion per pass in the second stage is typically 90% to 99%, and acrylic acid selectivity is greatest when the bed outlet remains below 360 °C. At high gas hourly space velocity, the reaction zone shifts toward the tube outlet, and the peak temperature may move beyond the axial thermocouple coverage. The maximum gas hourly space velocity is therefore limited not by catalyst activity alone but by the need to keep the hot spot within the measured and controllable region. Pressure drop across the second-stage tube is commonly 0.03 MPa to 0.08 MPa; excessive pressure drop increases air compressor power and can create flow imbalance across the tube bundle if fouling occurs.

The Ergun equation provides a quantitative relation between pressure drop and gas velocity. For catalyst particles with equivalent spherical diameter 3 mm to 5 mm, bed voidage 0.38 to 0.45, and gas mass velocity in the range of 0.5 kg·m⁻²·s⁻¹ to 1.5 kg·m⁻²·s⁻¹, the pressure drop is dominated by the inertial term. A reduction in particle diameter from 5 mm to 3 mm can increase pressure drop by a factor of approximately 3 at constant gas velocity, while improving intraparticle effectiveness and lowering the catalyst center temperature. Consequently, the final loaded particle size distribution is selected by the licensor after pilot-plant measurements of pressure drop, hot spot, and product yields. Published data for this specific optimization is limited, as it is usually contained in proprietary loading diagrams.

Extended operation near the upper hot spot limit causes measurable changes in the second-stage catalyst. Molybdenum can migrate along the bed under high temperature and steam, and the pore structure collapses gradually. Surface area measurements by nitrogen adsorption according to ISO 9277:2010 show that fresh second-stage catalysts typically have BET areas of 5 m²·g⁻¹ to 20 m²·g⁻¹; aged catalysts extracted from hot spots may show local reductions of more than 30%. The loss of surface area lowers activity and can force operators to raise salt temperature, which compresses the remaining hot spot margin. Mercury intrusion porosimetry according to ISO 15901-1:2016 reveals a decline in mesopore volume that correlates with acrylic acid selectivity loss. These analytical results are used to decide catalyst replacement cycles, which are typically 2 years to 5 years for second-stage service depending on the severity of the hot spot history. The acrolein hot spot temperature constraint is embedded in the selection of catalyst replacement intervals and the configuration of forced salt circulation equipment.