The Transformative Impact of AI on Petrochemical Production

The petrochemical industry is facing a pivotal moment, marked by the growing demand for light olefins such as ethylene, propylene, butanes, and benzene, toluene, xylene (BTX), while the need for traditional fuel products is on the decline. This changing landscape is influenced by the global availability and affordability of crude oil, a key factor in the production of these petrochemicals.

Light olefins, also known as alkenes, are unsaturated aliphatic hydrocarbons characterized by one or more carbon double bonds. These compounds are foundational in the chemical industry, essential for producing a wide range of products that are integral to modern life, including synthetic fibers, coatings, and solvents. The market dynamics for these olefins are shifting significantly, with a report from Fior Markets projecting the global alpha olefins market to grow from USD 8.6 billion in 2020 to USD 13.7 billion by 2028, at a CAGR of 6% during the forecast period of 2021-2028.

In the center of this industry is ethylene (C2H4), a basic alkene used predominantly in the production of polyethylene for the plastic industry. Other alkenes like propylene (C3H6), butene (C4H8), and pentene (C5H10) find their applications in the synthesis of chemical compounds, resin production, alkylate gasoline production, and rubber processing. The optimization of ethylene production is a complex and integrated process, requiring a comprehensive process control system that utilizes various sensors to measure physical, chemical, and process parameters. Online analyzers are particularly important in this regard, as they measure critical quality properties of feed, semi-final, and final products, enabling refineries to make informed decisions about corrective or preventive actions.

The analysis of crude oil quality online is crucial in producing the most economical blend with the highest refining margin. This aspect significantly improves the refining margin and profitability of the entire crude-to-chemicals processing. Innovations like the advanced MOD-4100 crude analyzer represent breakthroughs in online crude oil analysis, performing critical measurements such as salt content, water recovery, API gravity, viscosity and Total Acid Number (TAN).

A strategic approach in maximizing light olefins from crude oil involves running the process at the highest economic performance. Modern AI-driven optimization tools, like the ones provided in Modcon.AI packages, play a key role in this. They enable connectivity, validation, and prediction of main KPIs, aiding refineries in maintaining and improving the management of industrial processes. These tools, particularly artificial neural network (NN) dynamic models, allow for the calculation and prediction of physical properties and chemical compositions for different process streams, guiding refineries in setting appropriate parameters.

AI models combined with accurate measurement of feedstock properties provide a valuable tool for predicting the composition of ethylene and propylene. This data can be used to determine fractionator parameters, aiding in the shift from fossil fuels to renewable energy sources and light olefins more effectively.

Despite the complexity of the process, the performance of the cracker often emerges as the bottleneck, making it a primary target for synchronization with sequential processes. The recent advancements in computing power and sophisticated mathematical solvers have made dynamic real-time optimization a reality. Today's dynamic modelling is based on nonlinear models, with processes grounded in the correlation between design and online optimizer models, as well as theoretical laws pertaining to chemistry, reaction kinetics, and physics.

AI optimizer models combined with accurate feedstock measurement, sensor feedback, and historical data, provide a robust tool for predicting ethylene/propylene composition. This approach is instrumental in determining the KPIs of the fractionator that most influence production efficiency.

Efficient utilization of process data from each unit and mathematical models developed for dynamic optimization enable running the process in an efficient and cost-effective mode. This not only increases production capacity but also reduces energy costs, thereby enhancing the profitability of the ethylene/propylene production unit and strengthening its position in the global petrochemical market.

Refineries face challenges such as variability in gas and liquid feedstocks, volatility in crude oil prices, and constant margin erosion in certain petrochemical chains. Modcon.AI implementation has shown promising results in addressing these challenges, including an 8-10% increase in ethylene yield, a 10-20% improvement in energy efficiency, and a reduction in risks impacting quality, profit, environmental safety, and operational manpower.

Looking forward, the anticipated changes in fuel consumption due to new automotive technologies, government pressures for renewable energy adoption, and the ongoing global pandemic are prompting refineries in the Western world to make far-reaching changes in their production technologies and redefine their competitive strategies. Online crude oil quality analysis plays a significant role in this transition, producing the most economical blend with the highest refining margin and improving the overall profitability of the crude-to-chemicals process. These successes can be attributed to advancements in process analyzers, leading to savings in production, reduced product giveaway, lower operating manpower requirements, and enhanced energy conservation.