Across FCC, steam cracker, raffinate-2, and MTBE residual C4 streams, the distribution of n-butenes, iso-butylene, paraffins, and diolefins is not constant enough to allow a fixed metathesis operating point without feed linearization. The reactive 2-butene concentration in mixed butenes can vary from less than 10 vol% in raw steam cracker C4 to 55 vol% in raffinate-2 after butadiene extraction and isobutylene removal. This variability directly changes the heat release, the equilibrium driving force, and the inert paraffin concentration in the recycle loop. FCC-derived C4 typically contains 12–16 vol% 1-butene, 15–20 vol% cis- and trans-2-butene, 15–20 vol% isobutylene, and 35–45 vol% mixed butanes, with 0.2–0.5 vol% 1,3-butadiene. Steam cracker C4 differs sharply: 40–50 vol% 1,3-butadiene must be selectively hydrogenated before any metathesis catalyst, while 2-butene may be only 5–10 vol% until extraction and isomerization. Raffinate-2, obtained after selective butadiene extraction and isobutylene removal, contains 25–30 vol% 1-butene, 45–55 vol% 2-butene, and 10–15 vol% butanes as the primary unreactive ballast. On-line gas chromatography following ASTM D2163 quantifies C1–C5 hydrocarbon distribution; trace diene breakout in the C4 feed to metathesis is more appropriately measured by ASTM D2712. Without stabilization of the 2-butene feed concentration, the metathesis reactor experiences variance in apparent WHSV and ethylene:2-butene ratio that cannot be compensated by simple flow ratio control. The commercial fix is not a surge drum alone but a combination of selective hydrogenation, double-bond isomerization, and distillation or adsorption that delivers a metathesis feed with a C4 olefin composition close to the values shown in Table 1.

Feed linearization is therefore not a single unit operation but a sequence of hydrogenation, isomerization, and purification steps that must be tuned to the raw C4 source. The target metathesis feed is a 2-butene-rich stream with diene content below 10 ppmw, isobutylene below 0.3 wt% where possible, and paraffins controlled by a purge strategy that prevents inert accumulation without excessive butene loss.

What Impurity Profile Separates Polymer Grade from Chemical Grade Propylene?

Polymer-grade propylene is defined less by a single distillation cut than by a set of poison limits that are acceptable to Ziegler-Natta and metallocene catalyst families. Chemical-grade propylene may tolerate 2–5 wt% propane and 500–5000 ppmv ethylene without immediate polymerization failure, but polymer-grade material is typically specified at 99.5 wt% propylene minimum, with total olefinic and paraffinic impurities below 0.5 wt%. The specification matrix in Table 2 consolidates the limits most often invoked in licensing contracts and polymer supply agreements. Propylene purity is determined by ASTM D2712; non-condensable gases are separated by gas-solid chromatography under ASTM D2504; total sulfur is oxidized and measured by ASTM D3246; methanol is quantified by ASTM D4864; water is measured by ASTM D6304. These analytical methods are not interchangeable because the detection threshold for methanol in ASTM D4864 is below 1 ppmv, whereas a conventional hydrocarbon GC without an oxygenate column may pass a methanol-contaminated stream as clean. The most severe polymerization poisons are carbon monoxide, carbonyl sulfide, and arsine, but in metathesis-derived propylene the dominant risks are ethylene carry-over, methylacetylene/propadiene, and oxygenates introduced during feed pre-treatment or drying. A polymer-grade limit of 0.1 ppmv CO is required because Ziegler-Natta catalysts exhibit irreversible active-site poisoning at CO concentrations as low as 0.2 ppmv in the polymerization reactor, a boundary that leaves no analytical margin for error. Metallocene catalysts are often more sensitive to oxygenate and sulfur compounds, forcing a second guard bed of activated alumina and molecular sieve. The same polymer-grade specification applies to propylene used in high-activity polypropylene plants with loop reactors, gas-phase fluidized beds, and bulk polymerization units; the only variation is that gas-phase processes may specify a tighter water limit of 2 ppmv because water acts as a co-catalyst poison and a chain-transfer agent. The metathesis product must therefore be designed not only for the C3 splitter overhead purity but for a post-fractionation trace purification train.

The table reflects typical licensor specifications; individual supply agreements may tighten one or more limits based on the polymerization process and the catalyst system. Metathesis-derived polymer-grade propylene is particularly sensitive to methylacetylene and propadiene because these compounds can be formed by side reactions in the pre-treatment section if the hydrogenation reactor is not operated with sufficient hydrogen-to-diene ratio.

Prior to the metathesis reactor, the mixed butene stream is normally hydrogenated selectively over a palladium-on-alumina catalyst to reduce 1,3-butadiene and C4 acetylenics to butenes. The hydrogenation reactor is operated with a slight excess of hydrogen relative to diene stoichiometry, and outlet diene concentration is held below 10 ppmw, while acetylene and methylacetylene are held below 5 ppmw. A residual diolefin concentration above 50 ppmw at the metathesis catalyst inlet produces rapid oligomer formation and shortens the metathesis run length by fouling the first guard bed. After selective hydrogenation, the 1-butene must be isomerized to 2-butene because the metathesis reaction with ethylene proceeds selectively with 2-butene. Double-bond isomerization is conducted over MgO, hydrotalcite, or promoted alumina at 150–300°C and a liquid hourly space velocity of 1–4 h⁻¹ depending on the catalyst. The isomerization reactor is placed immediately before the metathesis guard bed, and its outlet is controlled to an equilibrium 1-butene/2-butene ratio at reaction temperature. The metathesis catalyst itself is typically tungsten oxide on silica, although molybdenum oxide on alumina and rhenium oxide on alumina are also active. Tungsten oxide on silica is preferred in commercial operations because it offers sufficient activity at 260–400°C, low acid-catalyzed oligomerization, and tolerance to traces of sulfur that would poison rhenium. The supported tungsten oxide catalyst is often used in a fixed-bed, multi-bed adiabatic layout with interstage heating, because the reaction is only mildly exothermic and requires the inlet temperature to be kept above the catalyst light-off. The guard bed before the first metathesis bed contains a low-activity hydroisomerization/metathesis composite or a sacrificial WO₃/SiO₂ bed and is replaced or regenerated more frequently than the main catalyst. In this pre-treatment section, the operational failure mode most commonly observed is not catalyst selectivity loss but pressure-drop build-up across the selective hydrogenation reactor caused by green oil deposition when the feed contains C5 dienes or when hydrogen distribution is poor. The guard-bed pressure drop on commercial units has been reported to increase by 0.2–0.5 bar over a 12-month campaign, after which the bed is bypassed and replaced.

When Feedstock Contains More Than 0.3 wt% Isobutylene

In commercial operation where isobutylene is not extracted upstream, the behaviour of the recycle loop changes fundamentally. Isobutylene does not undergo productive metathesis with ethylene to propylene under standard WO₃/SiO₂ conditions, but it is not completely inert. On acidic sites of the metathesis catalyst and on the isomerization catalyst, isobutylene can oligomerize to C8–C12 branched olefins and can crack to C3 and C5 by-products. The result is an accumulation of isobutylene in the C4 recycle stream and a continuous formation of heavy oligomers that deposit on the catalyst surface. The threshold of concern is a fresh feed isobutylene concentration above 0.3 wt% when the C4 recycle ratio is high; below this level, the purge of normal butane and isobutane typically carries isobutylene out of the loop at a rate that prevents autocatalytic oligomer growth. Above 0.3 wt%, the steady-state isobutylene concentration in the recycle can reach 5–10 wt% even though the fresh feed value is below 1 wt%, because the normal butane purge also rejects some isobutylene but not enough. The practical correction is upstream removal by conversion to MTBE or ETBE, by selective adsorption, or by acid-catalyzed dimerization to diisobutylene if the C8 olefins have a market outlet. If such removal is not installed, the metathesis unit must operate with a deliberately larger C4 purge, which reduces propylene yield by rejecting unreacted 2-butene along with the paraffins and isobutylene. The purge is monitored by ASTM D2163 and is typically set so that the isobutylene concentration in the recycle does not exceed 3 wt%. In this operating mode, the gas plant loses not only the isobutylene purge but also the co-purged 2-butene, and the overall propylene yield drops by 2–5 percentage points compared with an isobutylene-free feed. Published data for the exact relationship between purge rate and catalyst life at the 0.3 wt% threshold are limited, because licensors evaluate isobutylene management on a feedstock-specific basis rather than a universal limit. The safest design for polymer-grade propylene production is therefore to specify an upstream isobutylene removal unit when the fresh mixed butene stream contains more than 0.3 wt% isobutylene, unless a high-purity isobutylene co-product is already integrated into the site butadiene extraction complex.

Because the metathesis equilibrium limits single-pass conversion, the reactor is never operated at a stoichiometric ethylene-to-2-butene ratio of 1:1. The equilibrium position for 2-butene plus ethylene to two propylene is only moderately favourable at reaction temperatures, so excess ethylene is required to shift conversion upward. In commercial practice, the fresh ethylene-to-2-butene molar ratio is maintained between 1.5:1 and 3.0:1, with the lower end used when the feed is rich in 2-butene and the upper end used when the recycle loop contains high ethylene dilution. The reactor inlet temperature is normally set at 260–400°C for WO₃/SiO₂ catalysts, and the outlet temperature is allowed to rise by 30–70°C depending on the heat of reaction and the bed layout. Reactor pressure is maintained between 20 bar and 35 bar to increase the partial pressure of the reacting olefins and to facilitate downstream condensation without excessive ethylene compression. Weight hourly space velocity based on 2-butene feed is typically 1–4 h⁻¹ for fixed-bed operations; lower values increase equilibrium conversion at the cost of producing more oligomers, while higher values reduce conversion and can move the reactor away from the optimum operating point. Single-pass 2-butene conversion is therefore confined to 60–75%, with the unconverted butene separated and recycled to the reactor. The overall propylene yield based on butenes after accounting for recycle and purge can exceed 90–95%, but only if the separation section returns a low-ethylene, low-paraffin C4 recycle. The main process conflict is that high temperatures above 400°C promote double-bond migration and oligomerisation, while temperatures below 260°C reduce catalyst activity and allow heavier C5+ olefins to condense on the catalyst pores. The margin between catalyst light-off and accelerated coking is therefore narrow, and the interstage heating system must be designed for rapid response to feed composition changes. Control of the reactor section is normally based on on-line gas chromatography of the reactor effluent with a cycle time shorter than 10 min, which provides feedback on conversion, ethylene consumption, and C5+ by-product formation. In this narrow operating envelope, even a 10°C increase in average bed temperature above the upper limit can shorten the run length by increasing the rate of coke deposition, while a 10°C decrease can drop conversion below 55% and produce a propylene product that fails the polymer-grade yield target.

Thermal Deactivation of Tungsten Oxide on Silica

Tungsten oxide supported on silica remains the dominant commercial metathesis catalyst, but its deactivation is not governed by a single mechanism. At normal operating temperatures, the slow accumulation of high-boiling oligomers inside the catalyst pores is the principal cause of activity loss, not the loss of tungsten surface area. The first sign is a gradual decline in conversion at constant reactor inlet temperature, accompanied by an increase in the pressure drop across the first metathesis bed. On a commercial fixed-bed reactor with a run length of 30–60 days, the pressure drop across the first bed can increase from a clean value near 0.15–0.25 bar to 0.45–0.60 bar before the bed is taken off-line for regeneration. The regeneration procedure uses a nitrogen-diluted air mixture with an oxygen concentration that is initially held at 0.5–1.0 vol% and is increased in steps to 2.0 vol% as the coke front moves through the bed. Regeneration inlet temperature is maintained between 450°C and 550°C, with the outlet temperature monitored continuously to prevent runaway coke combustion. Above 550°C, the risk of silica support sintering increases, and above 600°C volatile tungsten species can redistribute from the catalyst surface, reducing the number of active metathesis sites and permanently lowering activity. The most subtle operational issue is water desorption during regeneration: the WO₃/SiO₂ catalyst must be cooled and conditioned with dry nitrogen before returning to olefin service, because residual water vapour at concentrations above 10 ppmv in the regeneration gas can hydrolyse the active tungsten species and delay light-off. In some commercial units, the first guard bed is not regenerated in situ but is replaced as a sacrificial bed because its coke loading, oligomer deposition, and metal pick-up from trace heavy metals in the feed make regeneration unattractive. The main bed is regenerated separately, and the replacement cycle is coordinated with the selective hydrogenation catalyst to minimise the total outage time. Published data for the exact deactivation constant of polymer-grade propylene production units is limited, but industrial experience indicates that the deactivation rate is more sensitive to diene breakthrough, oxygenate ingress, and isobutylene concentration than to the base operating temperature within the normal 260–400°C window.

Downstream of the reactor, the effluent is cooled against reactor feed and condensed in a multistage separation train that removes unconverted ethylene, polymer-grade propylene, and C4 recycle streams in sequence. The first column is a deethylenizer that strips methane, hydrogen, and light gases overhead and sends ethylene-rich liquid to a C2 splitter or an ethylene recycle compressor. The deethylenizer is normally operated at a pressure of 25–35 bar to allow propylene and heavier components to condense without excessive refrigeration. The overhead ethylene stream is recycled to the metathesis reactor after separation from methane and hydrogen, because methane accumulation in the recycle loop reduces the partial pressure of ethylene and forces a higher purge. The next column is a depropylenizer that recovers a propylene-rich overhead and a C4-plus stream. The crude propylene overhead is then sent to a C3 splitter, where propylene is separated from propane and any residual ethane and ethylene. The C3 splitter is the highest energy user in the metathesis block due to the low relative volatility between propylene and propane; a column overhead pressure in the range 15–22 bar and a reflux ratio that can exceed 10 is typical. The C3 splitter overhead is not automatically polymer-grade. Residual methylacetylene and propadiene can survive the splitter overhead if they are not eliminated by selective hydrogenation or adsorption, and these compounds are poison for both Ziegler-Natta and metallocene catalysts. A trace purification train containing activated alumina, molecular sieve, and a fixed-bed oxygenate/methanol adsorber is therefore included downstream of the C3 splitter when the metathesis product is intended for polymer-grade use. The treated propylene is dried to a water content below 5 ppmw and, for gas-phase polymerization, often below 2 ppmw before it is sent to the polymerization reactor. In this separation sequence, the critical equipment failure mode is not the distillation tower itself but reboiler fouling from residual oligomers and heavy olefins that polymerise at the elevated reboiler temperatures. Operators on commercial units maintain the depropylenizer bottom temperature below 200°C and inject a low-dosage polymerization inhibitor into the C4 recycle to prevent fouling. The overhead and bottoms compositions are monitored by ASTM D2712, with the concentration of trace oxygenates and water measured by ASTM D4864 and ASTM D6304 respectively. The propane content of the polymer-grade propylene is the most common off-spec cause during startup, and it is corrected by increasing the C3 splitter reflux ratio rather than by changing the upstream reactor conditions.

Ethylene Recycle Ratio and the Propylene Yield Boundary

The molar ratio of ethylene to 2-butene at the reactor inlet is the primary economic variable for a metathesis-based polymer-grade propylene unit. At low ratios near 1.2:1, the equilibrium conversion drops and the recycle burden of unconverted 2-butene increases; at high ratios above 3.0:1, the propylene yield improves only incrementally while the ethylene recovery section must handle substantially more ethylene recycle. In a typical closed-loop design, the ethylene recycle ratio is set between 1.5:1 and 2.5:1, and the net ethylene consumption is monitored by flow metering on the fresh ethylene feed, with on-line GC analysis of the deethylenizer overhead to track methane and hydrogen accumulation. Methane accumulation is controlled by a purge from the deethylenizer overhead; without a purge, methane can blanket the reactor and reduce the effective ethylene partial pressure even when the ratio controller indicates a constant value. The purge stream contains ethylene that is lost to fuel gas unless a secondary ethylene recovery membrane or pressure swing adsorption unit is installed. The acceptable recycle ratio is therefore bounded on the low side by a single-pass conversion below 55% and on the high side by ethylene compressor capacity and deethylenizer flooding. For a mixed butene feed containing 45–55 vol% 2-butene, an ethylene recycle ratio of 2.0:1 is generally sufficient to maintain a single-pass conversion in the 65–70% range without excessive equipment size. The final propylene yield is also influenced by the butane and isobutane content in the feed, because these inerts occupy reactor volume and must be purged from the C4 recycle. The C4 purge is typically set to keep the total butane concentration in the recycle below 15–20 wt%; higher concentrations dilute the reactive butenes and increase energy consumption in the C4 recycle tower. The product propylene from this configuration is only ready for polymer-grade service after the trace purification train removes residual ethylene, methylacetylene/propadiene, oxygenates, and water, because even 50 ppmv ethylene in the final propylene can alter the polymerization kinetic profile and reduce stereoregularity in Ziegler-Natta systems. The operating boundary for polymer-grade propylene via olefin metathesis of mixed butenes is not set by the metathesis reaction itself but by the ability of the downstream separation and purification train to hold the poison limits under recycle transients, feedstock changes, and catalyst regeneration cycles.