In fluidised bed methanol-to-olefins operation, a silicoaluminophosphate molecular sieve with CHA topology, designated SAPO-34, is compounded with clay, alumina sol, or silica sol and spray-dried into microspheroidal granules. The particle size distribution is controlled with a fines fraction below 20 wt% for particles smaller than 40 µm and a Sauter mean diameter of 70 µm to 90 µm as measured by ISO 13320:2020. Methanol cofeed in this context refers to the simultaneous introduction of methanol with steam, recycled oxygenates, or recycled C4–C6 hydrocarbons into the dense-phase bed, transport zone, or both. The fluidised bed reactor is integrated with a continuous catalyst regenerator because SAPO-34 undergoes rapid reversible deactivation by coke deposition in the chabazite cage apertures; methanol conversion in commercial operation is maintained above 99.5% with dimethyl ether slip typically below 0.5 wt% carbon basis. The reaction sequence proceeds through methanol dehydration to dimethyl ether, formation of a methylbenzene hydrocarbon pool within the CHA cages, and subsequent olefin elimination via paring and side-chain methylation routes. Ethylene and propylene are the primary products, with carbon selectivity to C2= plus C3= commonly falling between 80% and 85% on a carbon basis at complete oxygenate conversion. The exothermic heat release for the overall methanol-to-hydrocarbon conversion is approximately 1.4 MJ kg⁻¹ of methanol, and the fluidised bed provides heat removal through catalyst circulation and steam generation in immersed or external heat exchangers. Cofeed water reduces the methanol partial pressure, suppresses the equilibrium formation of dimethyl ether, and changes the concentration of adsorbed methoxy species, thereby influencing the distribution of polymethylbenzene intermediates. These effects are material because the CHA framework contains 8-ring apertures with crystallographic dimensions of 0.38 nm × 0.38 nm and cages of approximately 0.94 nm diameter, and product diffusion is controlled by both cage occupancy and aperture blocking.

How Does SAPO-34 Cage Architecture Constrain Olefin Diffusion During Methanol Cofeed Transients?

Diffusional constraints in SAPO-34 arise from the narrow 8-ring windows connecting CHA cages. Under methanol cofeed conditions, the concentration of light olefins in the gas phase is not directly proportional to their rate of formation because ethylene and propylene must diffuse through the same apertures used by methanol and water. At high methanol partial pressure, methoxy and methylbenzene intermediates occupy the cage space and reduce the effective aperture cross-section, which increases the residence time of propylene and butenes inside the cage. The measured apparent activation energy for methanol conversion over SAPO-34 reported in open literature ranges from 80 kJ mol⁻¹ to 130 kJ mol⁻¹ depending on silicon distribution, acid site density, and coke content. The diffusivity of propylene through SAPO-34 at reaction temperature is more than an order of magnitude lower than Knudsen diffusion in a straight pore of equivalent size, and this gives rise to secondary olefin methylation and oligomerisation reactions that raise C4+ selectivity when diffusion is impaired. Water cofeed acts as a competitive adsorbate on acid sites and as a gas-phase diluent; it decreases methanol partial pressure and retards the formation of bulky methylated naphthalenes that block cage apertures. However, excessive water does not overcome intrinsic shape selectivity, and at high water-to-methanol ratios above 1.0 the beneficial effect on diffusion is offset by hydrothermal dealumination of the framework and by the energy penalty of steam vaporisation. A narrow processing window therefore exists in which the methanol-water cofeed ratio and dense-bed temperature jointly control both the rate of coke formation and the product selectivity. Published pilot-plant data indicate that at a water/methanol mass ratio of 0.20 to 0.60 and a dense-bed temperature of 450°C to 475°C, combined ethylene and propylene selectivity is stable within 1.5 percentage points, while operation outside this range produces either excess dimethyl ether at low temperature or excess methane and ethane at high temperature.

On commercial-scale fluidised bed MTO units, the gas distributor design determines the initial bubble size distribution, jet penetration depth, and attrition of the SAPO-34 microspheres. Methanol is injected as vapour through a distributor that maintains pressure drop above the bed pressure fluctuation amplitude; the design criterion commonly applied is a distributor-to-bed pressure drop ratio of 0.10 to 0.30. Superficial gas velocity in the dense phase is usually held between 0.2 m s⁻¹ and 1.2 m s⁻¹, which places the operation in the turbulent fluidisation regime for a Geldart A powder. Under these conditions the bed density is typically 500 kg m⁻³ to 800 kg m⁻³, and cyclone inlet solids loading rises as fines are produced by attrition. The cofeed stream changes the volumetric gas flow and therefore the superficial velocity; an increase in water-to-methanol ratio from 0.2 to 0.6 can increase superficial velocity by 10% to 15% at constant methanol throughput. If the velocity approaches the terminal velocity of the 40 µm particle fraction, fine catalyst loss to the product gas increases unless cyclone separation efficiency is maintained above 99.9%. Attrition resistance is measured according to ASTM D5757-11, and fluidised-bed SAPO-34 grades are typically formulated to an air-jet attrition index below 3 wt% h⁻¹; however, the test is performed at ambient temperature and does not fully reproduce steam-induced thermal attrition in the reactor. Co-injection of liquid methanol droplets without complete vaporisation creates localised high-velocity jets and thermal shock, which can fracture catalyst particles and reduce the particle size distribution below design values. Therefore, cofeed vaporisation is performed in a preheater or in a distributor zone with sufficient heat transfer area to ensure that no liquid droplets enter the catalyst bed.

Operational Boundaries for Methanol Partial Pressure and Water Cofeed Ratios

The methanol partial pressure at the reactor inlet is the product of total pressure and methanol mole fraction. In fluidised bed operation at total pressure 1.0 bar to 3.0 bar, methanol partial pressure is commonly maintained between 30 kPa and 80 kPa. Above this range, coke deposition accelerates because the hydrocarbon pool becomes denser and the rate of undesirable hydrogen transfer from methylbenzenes to olefins increases. Below the range, the reaction rate per unit catalyst inventory falls and the space velocity must be reduced, increasing capital cost per tonne of olefin. The water/methanol mass ratio is typically controlled between 0.20 and 0.60 in steam-diluted operation. At ratios below 0.20, the adiabatic temperature rise from the methanol conversion exotherm can be difficult to control in a dense bed with limited heat removal surface; local hot spots above 500°C promote thermal cracking of butenes and pentenes, increasing methane and ethane. At ratios above 0.60, hydrothermal dealumination of SAPO-34 becomes kinetically significant at temperatures above 450°C, and the equilibrium dehydration of methanol to dimethyl ether is suppressed, causing DME slip unless the contact time is increased. The processing window for the combined temperature and water cofeed variables is narrow; pilot-plant data show that selectivity is relatively insensitive within 450°C to 475°C and 0.20 to 0.60, but shifts by more than 2 percentage points when the temperature deviates by ±5°C at a water/methanol ratio near 0.10. The methanol space velocity expressed as weight hourly space velocity is typically set between 2 h⁻¹ and 10 h⁻¹; higher values reduce conversion and increase DME slip, while lower values increase residence time and favour secondary reactions of propylene to butenes and aromatics. In cofeed operation with recycled oxygenates such as dimethyl ether, the total oxygenate-to-catalyst ratio must include DME carbon content because DME occupies acid sites and contributes to the hydrocarbon pool at a rate comparable to methanol on a carbon basis.

Continuous catalyst regeneration is mandatory in SAPO-34-catalysed methanol cofeed operation because the coke content on spent catalyst can reach 5 wt% to 10 wt% within minutes to hours depending on cofeed composition and temperature. The coke is not a single phase; it consists of polyaromatic deposits formed by the condensation of methylbenzenes and naphthalenes inside and near the CHA cages. Thermogravimetric analysis of spent catalyst under air, conducted according to ASTM D5373-21 for carbon determination, typically reveals a major combustion peak between 500°C and 650°C. The regenerator is operated at 550°C to 650°C to remove coke while minimising hydrothermal degradation of the SAPO-34 framework. Steam from cofed water is present in the reactor, but in the regenerator the gas is air or air plus limited steam; the water partial pressure must be controlled because SAPO-34 is more susceptible to dealumination than ZSM-5 under hydrothermal conditions. Regenerator flue gas oxygen concentration is typically maintained at 1 vol% to 3 vol% excess oxygen to avoid local temperature excursions beyond 700°C, which would collapse the CHA structure and irreversibly destroy acid sites. The catalyst circulation rate is set to maintain a spent catalyst carbon level that does not reduce the pore aperture to less than the size of ethylene and propylene; this is normally achieved at a catalyst-to-methanol mass flow ratio of 10 to 25 in a circulating fluidised bed, although published data for specific commercial configurations is limited. In units with a secondary riser or stripper, the cofed water removes some adsorbed oxygenates by steam stripping before the spent catalyst enters the regenerator, which reduces the carbon load and limits temperature rise in the regenerator. The regenerated catalyst is returned to the reactor through a standpipe and slide valve, where dry air or nitrogen is used to prevent backflow of hydrogen-rich gas. Moisture adsorbed during catalyst transfer or storage at relative humidity above 60% can impair flowability and should be avoided by pre-drying the catalyst hopper and transfer lines.

When Recycled C4 Olefins Are Co-Injected with Methanol in a Commercial DMTO-Style Reactor

Butene and pentene co-injection with methanol is practised in some MTO process configurations to convert low-value by-product streams to propylene and ethylene. The recycled C4 fraction is recovered from the product gas after fractionation and may contain 20 wt% to 60 wt% isobutene, butene-1, butene-2, and small amounts of butadiene and paraffins. When co-fed with methanol into the SAPO-34 fluidised bed, butenes undergo acid-catalysed oligomerisation, isomerisation, and cracking. The cracking of butenes to propylene and ethylene is endothermic and partially offsets the exotherm of methanol conversion, reducing the heat-removal duty in the dense bed. The presence of methanol is important because methanol methylation of the cracked olefin fragments suppresses the formation of dry gas and heavy aromatics. However, excessive butene cofeed increases coke selectivity because butene oligomers are bulkier and more reactive than methanol-derived C1 intermediates. Industrial design data from DMTO-style units suggest that the recycled C4-to-methanol carbon ratio should be maintained below 0.20 to avoid a rapid increase in coking and a loss of light olefin selectivity. The exact ratio depends on the isobutene content and the concentration of dienes, which are more potent coke precursors. In a separate fluidised bed or downer-secondary reactor, the butene-rich stream may be contacted with regenerated SAPO-34 at a short contact time of 1 s to 3 s to promote cracking while limiting oligomerisation. In a single dense-bed cofeed operation, the superficial velocity is increased by the vaporised C4 injection, and the cyclone load is altered. The product gas analysis for such streams is usually performed by gas chromatography according to ASTM D2163-14 for C4 hydrocarbons and ASTM D1945-14 for fixed gases and light hydrocarbons. Published data for this specific configuration is limited, but the general trends are supported by laboratory fixed-bed and fluidised-bed studies.

Catalyst Attrition and Fines Retention Under Cofeed Conditions

Spray-dried SAPO-34 catalyst for fluidised bed MTO service is designed to resist particle attrition while maintaining sufficient macroporosity for reactant access. The attrition index is measured with an air-jet apparatus according to ASTM D5757-11; typical targets for fresh spray-dried MTO catalyst are below 3 wt% h⁻¹, but equilibrium samples after repeated regeneration can exhibit higher attrition due to thermal cycling and steam exposure. The cofeed water accelerates the dehydration of binder phases and can weaken the alumina or silica binder if the binder was not properly calcined. Fines below 20 µm are preferentially lost through the cyclone overflow unless the cyclone dipleg is sealed by a trickle valve or an equivalent device. Catalyst losses are normally tracked by measuring the particulate matter in the product gas quench water and by periodic laser diffraction analysis of collected fines according to ISO 13320:2020. In a large fluidised bed MTO reactor, the catalyst inventory is typically replaced in part by fresh catalyst addition to maintain both particle size distribution and activity; makeup rates are often expressed in kilogram of fresh catalyst per metric tonne of methanol feed. Published data for specific commercial SAPO-34 units is limited, but design basis values between 0.05 kg t⁻¹ and 0.30 kg t⁻¹ are used in conceptual studies. The interaction between cofeed water and the catalyst is strongest at the distributor, where hot steam jets and methanol vapours can cause local particle velocities above 30 m s⁻¹ if the distributor nozzle sizing is incorrect, leading to severe attrition and erosion of reactor internals. Therefore, nozzle exit velocities are usually limited to 30 m s⁻¹ to 60 m s⁻¹ depending on the nozzle design, and the distributor pressure drop is maintained at the upper end of the design range to achieve uniform gas distribution.

Process monitoring in SAPO-34 methanol cofeed operation requires rapid on-line analysis of reactor effluent because the product spectrum responds to small changes in catalyst circulation, cofeed ratio, and bed temperature within minutes. Methanol conversion is inferred from the concentration of dimethyl ether and methanol in the quenched product gas; a rise in DME slip above 0.5 wt% carbon basis indicates coke build-up, poor fluidisation, or inadequate regenerated catalyst temperature. The methane and ethane yields are monitored as thermal-cracking indicators; a sudden increase above baseline by more than 2 percentage points is often caused by a temperature excursion in the dilute phase or cyclones above 500°C. Acetylene and methylacetylene are tracked to distinguish thermal cracking from catalytic hydrogen transfer. The water-to-methanol ratio is measured by mass balance across the feed vaporiser and by on-line moisture analysers, and the set point is adjusted to maintain the dense-bed temperature in the prescribed range. Process gas compressors downstream of the reactor are protected from catalyst fines by cyclone systems designed to achieve a total separation efficiency above 99.9% and by quench water venturi scrubbers. In units that co-feed C4 olefins, the butadiene and isobutene content of the recycle stream is controlled by selective hydrogenation or solvent scrubbing; butadiene concentrations above 1 wt% in the cofeed increase fouling of the reactor distributor and accelerate coke formation. The operational boundary for SAPO-34 is ultimately defined by the hydrothermal stability of the chabazite framework, and the combination of high water partial pressure and regenerator temperature above 650°C must be avoided. Published data for specific commercial SAPO-34 cofeed configurations is limited in certain areas, particularly the long-term effect of alternating oxidising and reducing atmospheres on attrition rate, and this limitation should be addressed through pilot-plant validation rather than extrapolation from FCC catalyst data.