.png)
Plants often leave recoverable margin behind on pyrolysis gasoline because feed shifts, sulfur poisoning, and single-loop control create coordination gaps across the upgrading train. AI optimization closes those gaps by learning from plant data to coordinate two-stage hydrogenation and guide routing between gasoline blending and BTX extraction. Plants can extend catalyst run lengths, lift BTX recovery, and progress from advisory mode toward automated control at their own pace.
Pyrolysis gasoline, or pygas, is the naphtha-range, aromatics-rich liquid byproduct of steam cracking. It can be blended into the gasoline pool for its high octane value or routed to BTX extraction for benzene, toluene, and xylenes. For olefins complexes, that routing choice carries real margin consequences.
Deloitte's chemical industry outlook describes a sector facing overcapacity, soft demand, and pressure on profitability, with producers focused on operational efficiency and extracting more value from existing assets. Pygas sits high on that list, and getting the operational decisions right on feedstock, hydrogenation severity, and product route is where most of its recoverable margin lives.
Raw pygas is unstable, feed composition shifts with cracker economics, and selective hydrogenation performance shapes every downstream recovery option. Sites need real-time decision support to keep hydrogenation severity and product routing aligned as feed conditions move. Pygas processing sits among the more consequential levers in olefins plant operations.
Pygas value depends on feedstock, hydrogenation selectivity, and the chosen recovery path.
The sections below show how these shape route selection and day-to-day value capture.
Pyrolysis gasoline is a naphtha-range, aromatics-rich stream produced when a steam cracker breaks down heavier hydrocarbons to make ethylene and propylene. Its composition spans C5 to C12 and includes aromatics (primarily benzene, toluene, and xylenes), olefins, diolefins, styrenics, and paraffins. It carries a high octane number, which supports direct blending into the finished gasoline pool. Its high aromatics content also makes it a valuable feedstock for BTX extraction, primarily to recover benzene and, to a lesser extent, toluene.
Pygas composition is a direct function of cracker feed, and that composition sets the ceiling on what downstream processing can recover. Diolefins, styrenics, and reactive olefins in fresh cracker liquid will polymerize and form gums if not stabilized, so selective hydrogenation is typically the first step after quench. Beyond stabilization, the aromatics share decides whether BTX extraction is economic at all.
Naphtha-fed crackers produce pygas that can contain substantial BTX, with reported steam cracker gasoline BTX content around 56–80 wt% in one chemical engineering source. Ethane crackers generate only a few percent pygas by yield, and that pygas carries a different aromatics profile.
Heavier feedstocks yield more liquid co-products per unit of cracker throughput; lighter feeds maximize ethylene selectivity at the expense of pygas volume. Benzene usually carries the primary recoverable value because mixed xylenes recovery from pygas is often uneconomical at typical pygas xylene content and high ethylbenzene levels.
When a cracker shifts from naphtha to LPG for ethylene margin reasons, pygas throughput drops. The remaining feed arrives with different diolefin content, sulfur speciation, and carbon number distribution, and the pygas hydrogenation unit has to reoptimize around a composition disturbance it did not create and cannot predict from its own instrumentation.
Managing those handoffs is one of the harder parts of feedstock variability management in an integrated olefins complex.
The pygas upgrading train has three sections. First-stage hydrogenation stabilizes the reactive stream, second-stage hydrogenation removes the last olefins and sulfur, and fractionation splits the stabilized product into recoverable cuts. How those three sections are run determines the margin available at every route choice that follows.
GHU-1 selectively converts diolefins, styrene, and some olefins in liquid-phase fixed-bed reactors. It must stabilize the stream without destroying the aromatics that carry the product value, which sets the operating window for everything downstream.
Incomplete conversion in GHU-1 degrades every section that follows. Residual diolefins move into GHU-2 and extraction equipment, where they contribute to gum formation and fouling on reactor beds, exchangers, and column internals.
GHU-2 completes olefin removal and hydrodesulfurization of the C6–C8 heart cut. Its core constraint is selectivity. Sulfur and olefin conversion need to be pushed while aromatic ring saturation stays near zero.
Both reactions generate similar exothermic responses in the reactor bed, so temperature measurement alone can't cleanly separate them, and reactor selectivity becomes harder to hold with single-loop control.
After hydrogenation, sites choose among several processing routes. Standard BTX extraction recovers benzene, toluene, and xylenes from the C6–C8 cut. Some facilities recover styrene before full hydrogenation where market economics support the additional processing. Others integrate thermal hydrodealkylation to convert toluene into more benzene, or saturate aromatics for cracker recycle when olefin value is stronger than aromatics value.
Gasoline blending captures octane value but leaves chemical value unrecovered. Benzene content regulations in finished gasoline also limit how much raw pygas can be blended. Extracting benzene first can increase the remaining stream's blending flexibility while also lifting product value.
C5 fractions add incremental value, but recovery economics depend on site scale and available infrastructure. The same applies to deeper recovery paths beyond benzene and toluene. In practice, route selection is an ongoing chemical manufacturing optimization problem shaped by live spreads between benzene, styrene, isoprene, and ethylene, along with feed quality and unit constraints.
Pygas units operate under coupled constraints that conventional control handles poorly. The most consequential failure modes involve catalyst poisoning, selectivity loss, and cross-unit coordination gaps.
A common failure mode is poisoning from sulfur species carried over from the cracker. Elevated CS₂ in particular is known to deactivate palladium-based GHU-1 catalysts, which shortens cycle lengths and pushes operators toward higher reactor inlet temperatures to compensate.
Higher temperatures then accelerate coke deposition, which shortens cycle length further and raises energy cost. Managing that temperature escalation is a meaningful margin lever, but it requires visibility into catalyst activity state, a variable that must be inferred from conversion and pressure-drop signatures rather than measured directly.
Single-loop PID control handles variables independently. It can't coordinate hydrogen-to-feed ratio, reactor temperature profile, and space velocity across two interdependent reactor stages with competing conversion targets.
Tuning a cascade of setpoint controls to hold both GHU-1 and GHU-2 at their selectivity limits while feed composition drifts is a job single-loop control was never designed for.
The problem is also organizational. Maintenance tracks catalyst condition, operations manages reactor setpoints, and planning decides product routing based on market prices. When those groups work from separate data and separate models, the integrated optimum stays harder to see.
Modern data historian best practices help, but a shared dynamic model is what actually puts catalyst state, conversion, and margin in front of all three groups at the same time.
AI optimization addresses the coordination gap that opens when several interacting constraints shift at once and static models lose accuracy. Operator judgment built over decades at the board still decides which recommendations to act on.
Advanced control systems have been applied to pygas hydrogenation for years and can improve performance. But models identified on one feed slate can misrepresent process behavior as conditions drift, and sustaining model maintenance remains an operational burden on many sites.
A learning model built from actual plant data keeps pace with those drifts by updating its representation as the unit runs. The model captures non-obvious interactions between GHU-1 severity, GHU-2 temperature profile, and downstream extraction feed quality.
Plants can start in advisory mode, where the optimization technology recommends setpoint changes and operators decide whether to accept them. Advisory mode preserves operator authority while confidence builds, and it does useful work on day one without requiring a step change in automation.
In advisory mode, the same model that would otherwise write setpoints is available for several other uses:
As confidence in the recommendations builds through feed transitions and catalyst aging cycles, sites can extend the model's remit toward supervised and then real-time optimization under operator oversight. The journey is progressive and can stop at whatever level of autonomy the site is comfortable with.
For process industry leaders seeking a practical way to improve pygas hydrogenation performance and value recovery, Imubit's Closed Loop AI Optimization solution for chemicals and petrochemicals operations learns from plant-specific operating history, builds dynamic models of catalyst behavior and feed-composition relationships, and writes optimal setpoints to the DCS in real time.
Sites can start in advisory mode, keep operators in full authority, and progress toward more automated operation as confidence builds across feed transitions and catalyst cycles.
Get a Plant Assessment to see how AI optimization can lift margin on pygas hydrogenation and BTX recovery.
Each feedstock transition changes diolefin content, sulfur speciation, and carbon number distribution at the same time, so the pygas unit has to reoptimize around a composition disturbance it did not cause. The response spans GHU-1 severity, hydrogen allocation, and fractionation targets. Models that learn from historical transitions can anticipate the setpoint shifts that follow a feed change, improving feedstock optimization rather than waiting for downstream conversion deviations to appear.
Raw pygas contains diolefins, styrene, and reactive olefins that polymerize and form gums during storage and downstream processing. Hydrogenation stabilizes these species so the stream can blend into gasoline without forming gums or feed an aromatics extraction unit without fouling downstream equipment. Hydrogenation runs in two stages. GHU-1 saturates diolefins selectively without destroying aromatics, and GHU-2 removes remaining olefins and sulfur. Skipping or undersizing hydrogenation causes first-pass yield to collapse across the processing train.
Yes. Operational improvements alone can lift pygas value recovery. Tighter reactor temperature control can extend catalyst run lengths, and better coordination between hydrogen allocation and reactor severity can reduce both over-hydrogenation of aromatics and under-conversion of olefins. Real-time routing decisions tied to changing benzene-gasoline price spreads can also improve profit optimization from existing infrastructure with no added equipment.
