Back

Question

Submit an Answer

To what extent do you see the deployment of digital approaches to maintain and operate facilities while leveraging artificial intelligence (AI) in the restructured operations of the petrochemical industry?

Mar-2023

Answers

Trine Dabros, Topsoe, trar@topsoe.com

Raw WPOs are highly contaminated and, therefore, can be fed directly to FCC or steam crackers only at very high dilution rates. To increase the recycled content in a feed stream, hydrotreating and, optionally also, hydrocracking are required to bring the WPO onto specification for downstream processing. These hydrotreating and hydrocracking steps must be tailored for contaminant removal and property adjustment to enable the processing of this new type of feedstock. These steps (upgrading via hydroprocessing) can be understood as a pretreatment necessary to integrate WPO into the existing production facilities.

However, some caution is needed when using the term pretreatment, as several of the steps in plastic recycling require some kind of pretreatment. This can, for example, be the sorting and cleaning steps required for plastic waste before pyrolysis, or it can be post-pyrolysis single contaminant clean-up steps, as is the case for use of sorbents. The contaminant levels in common mixed plastic WPO can be successfully upgraded via tailored hydroprocessing without any additional pretreatment steps required, but a combination of different steps could have economic benefits.

Mar-2025
Rainer Rakoczy, Clariant, rainer.rakoczy@clariant.com

Utilisation and conversion of waste plastic pyrolysis oil is of increasing industrial interest. There are multiple options from a technical standpoint utilising these materials with the highest desire to get the materials back to steam crackers to follow a circular economy concept. Nevertheless, a low-hanging fruit may be the utilisation of older or smaller process equipment in a refinery or a refinery complex with certain access to petrochemical equipment utilising small quantities of an available WPO source to treat or pretreat it for application through coprocessing in the hydrocracker or catalytic cracker. Clariant has expanded the proprietary Clarity, HDMax, and even Hydex portfolio to handle this demanding feedstock and convert it towards feedstock for the aforementioned processes.

Mar-2025
Chris Ploetz, Burns & McDonnell, cploetz@burnsmcd.com

When used as a petrochemical feedstock, the composition and physical properties of raw WPO, also called waste plastic pyrolysis oil (WPPO) or plastic pyrolysis oil (PPO), can cause various challenges in downstream processes. The nature of these challenges varies depending on the disposition of the oil as steam cracker feed, FCC feed, or hydroprocessing unit feed. Usage of these oils in a steam cracker or FCC supports circularity in the polyolefins market (for example, high-density polyethylene [HDPE], low-density polyethylene [LDPE], and polypropylene [PP]), whereas usage in a hydroprocessing unit (with subsequent processing through a reformer and aromatics complex) supports circularity in the aromatic derivatives market (for example, polyethylene terephthalate [PET], polystyrene, and nylon).

From the perspective of downstream processing as a petrochemical feedstock, notable characteristics of WPO include the following: high vapour pressure, low flash point, wide boiling range with heavy tail, high pour point, high levels of unsaturation (including diolefins), chemically bound oxygen and nitrogen, chemically bound halogens (primarily chlorine due to PVC in waste plastic), metals and other heteroatoms (for example, silicon and phosphorus), and particulates (reactor solids consisting of carbonaceous char and calcium halides).

All of these properties can be problematic in downstream processing, but the issues are magnified if the feed is to be 100% pyrolysis oil. In lieu of this, many refinery and petrochemical operators are considering pyrolysis oil blending at relatively small fractions with traditional feedstocks in order to reduce the adverse impacts of raw pyrolysis oil while still gaining credit for recycled content via certification from International Sustainability and Carbon Certification (ISCC) Plus or other third parties.

Specific pretreatment needs upstream of a hydroprocessing unit should focus on addressing diolefins, metals, silicon, phosphorus, and particulates. Diolefins should be saturated in a selective hydrogenation unit (SHU) to avoid oligomerisation at high temperatures in the reactor feed preheat train. Metals, silicon, and phosphorus should be removed using guard beds to avoid poisoning and pluggage/fouling of the main reactor catalyst bed.

Primary particulate removal should be accomplished at the pyrolysis facility, but users of WPO should also install filtration systems to prevent plugging of exchangers, catalyst beds, and control valves. Within the hydroprocessing unit, metallurgy should be evaluated for the presence of chemically bound chlorine and other halogens, which will react to form hydrochloric acid (HCl) and hydrofluoric acid (HF) within the reactor. These acidic compounds will ultimately be removed with the acidic sour water decant streams at the cold separator vessels downstream from the reactor.

Additionally, any user of WPO needs to consider the high vapour pressure (if unstabilised), low flash point, and high pour point. The vapour pressure of unstabilised pyrolysis oil can preclude storage in atmospheric tanks. The low flash point (typically <<100˚F) requires pyrolysis oils to be treated as a flammable liquid despite being relatively heavy. The high pour point (substantially above summer ambient temperatures) requires heat tracing or other methods of maintaining adequate storage and process temperatures to avoid pluggage due to wax build-up.

Although these pretreatment steps add cost to a project, new technology is not required; all these strategies are within the general experience of the refining and petrochemical industry. Catalyst and adsorbent providers are actively working to optimise their products and services to meet the needs of the pyrolysis oil market. Each WPO user should craft these considerations into a tailored pretreatment scheme that meets the needs of their specific application, considering the actual properties of the candidate pyrolysis oil and the needs of the downstream process.

Mar-2025
Woody Shiflett, Blue Ridge Consulting, blueridgeconsulting2020@outlook.com

WPOs contain a myriad of contaminants that are highly variable depending on what waste plastics constitute the pyrolysis process feedstock and what type of pyrolysis process is employed (thermal and catalytic). Some of these contaminants are in the form of particulates. Many of these contaminants can be removed simply by depth filtration in pretreatment reactors or beds, as has been reported in a joint Ghent University/Pall Corporation study. Mixed polyolefin pyrolysis oils tested have shown some 80% of metals removed in this manner and exhibit 40-60% less coke formation downstream.

Fossil fuel feed contaminants tend to be limited to Ni and V in the heaviest stocks (vacuum gasoil [VGO], deasphalted oil [DAO], and residue), Fe in many feeds from upstream corrosion products, or Si in lighter coker-derived feeds (naphtha and kero). WPO can introduce high levels of Na (as 10s-100s ppm), higher levels of Si and Fe (10s of ppm), some Pb (~ <10 ppm), and significantly high levels of chlorine (Cl) (100s of ppm). Clearly, in any case, some significant pretreatment is and will be required.

Most technology providers and catalyst suppliers actively engage in guard catalyst and ‘hydrodemetallisation’ catalyst development to meet the needs of emerging feedstocks, with the renewables co-processing and hydroprocessing area being a somewhat recent example over the prior decade or two.

WPO processing guard catalyst development is and will be following. Speciality guard material innovators and suppliers, such as Crystaphase (Houston, TX), are and will be tailoring specialised trapping guard systems to address these needs. As WPO processes enter full commercial-scale applications, more detailed physical and chemical characterisation of contaminants will be needed in order to design and optimise appropriate pretreatment and guard material processes and products.

Mar-2025
Scott Sayles, Becht, ssayles@becht.com

Waste plastic oil (WPO) has potential impurities that cause catalyst deactivation. The types of impurities depend on the plastic type being fed to the liquefaction device. Typical feed contaminants are nitrogen, oxygen, olefins, phosphorus, silicon, and chlorides. For example, polyvinyl chloride (PVC) has the most difficult composition, mainly due to the chloride concentration and some metal stabilisers, while polypropylene has the least. The waste plastic received is a mixture of all plastic types. Some sorting is used to remove the hardest-to-process plastics, but the resulting feed is typically a mix of plastic types.

The level of contamination that reaches the hydroprocessing reactors determines the rate of catalyst deactivation. The higher the contamination, the shorter the run length that is observed; this is similar to fossil fuel deactivation. The individual contaminants have individual deactivation rates, and they are also cumulative. The contaminants are at higher concentrations and lower boiling ranges than the equivalent fossil fuels, resulting in higher overall catalyst deactivation. The plastic liquefaction step does not seem to impact the contaminant concentration. However, a pretreatment unit such as that used for renewable feeds is not required.

The method of producing the plastic oil also determines the level of contamination, with hydroliquefaction (HTL) removing more contaminants and pyrolysis retaining more contaminants in the liquid phase.

Mar-2025
Richard Evans, Delaware UK, richard.evans@delaware.co.uk

Digitalisation works on so many levels within the oil and gas sector. For example, it makes it much easier to create a unified experience among employees, which helps to secure the right people with the appropriate qualifications and experience to implement work. This means it is a lot easier to remain compliant with safety regulations, for example. On another front, technology like digital twins has also matured and is now very accurate and reflective of the operating environment. This technology means planners can remain onshore, therefore reducing the risk of accidents while also saving carbon on unnecessary helicopter flights to rigs.

More recent developments have been around AI and IoT and the collection and analysis of time-series data. These data first have to be normalised and tagged – does the source represent pressure, temperature or even acoustic soundwaves? But once collected, the data must be used, and this is where intelligence is applied. The data need to be referenced against example datasets to be able to spot anomalous signals. An AI model can be trained to spot irregularities and build a picture of what is really happening. Whether there is excess wear in a pipe or gas is not being sufficiently heated, the AI will be able to accurately discover this and suggest preventative maintenance. It is a very granular approach that can diagnose and isolate root causes. Utilising AI means identifying the faulty piece of equipment or process the first time, thereby reducing long hours spent investigating potential causes. This in itself helps to reduce overall maintenance costs and increase reliability.

When real-time data are available that present how the operating environment is performing in the moment, and the design limits and regulatory requirements are embedded into the system, it is possible to make the operating environment much safer, ensure that hydrocarbons stay in the pipes, and also prevent accidents.

Jun-2023
Philippe Mège, Axens, philippe.mege@axens.net

Digitalisation of process operations has developed thanks to data transfer technologies, sensors and soft sensors generating data, and IA, especially with machine learning tools, building systems that learn from data. This leverages process expertise, resulting in digital twins designed to optimise asset operation, thus securing the business decision. That ensures reliable operation and allows us to anticipate maintenance, which reduces its cost, and compare actual performances vs forecasted ones.

This applies to catalysts and also equipment like heaters, compressors, and heat exchangers. Such tools are available in the Axens Connect’In digital application. Several machine learning tools have already been developed and used by operators for octane prediction, recycle gas purity estimation, heat exchanger network optimisation, or generating soft sensors to get streams analysis not directly measured.

Jun-2023
Mark Fronek, Becht, mfronek@becht.com

The hard lessons learned from large-scale, corporate, top down, digital transformation initiatives have put pressure on organizations and vendors to prove value before larger scale implementation is considered. As a result, we see widespread pilot project activity in maintenance and operations. Maintenance history data which has been digitized is driving initiatives in predictive, condition-based maintenance programs. In the field, many organizations are testing wearables, IIoT 4.0, and drones which are using Machine Vision AI. In the operator booth improved access to democratized, visualized data is enabling shorter time to action through leveraging AI to create soft sensors, eliminate false alarms, and more. Access to soft knowledge through natural language processing AI systems is lagging, but I expect the media attention to ChatGPT will motivate extended exploration of the potential in this area.

Mar-2023
Charles Brandl, Honeywell UOP, charles.brandl@honeywell.com

As the refining and petrochemical industry evolves to address challenges related to the energy transition, sustainability and digital transformation (digitalisation) are critical to be viable and competitive. Adoption of digital applications, including cloud-enabled solutions, continues to improve and deliver benefits to the industry. Digitalisation is transforming how we will run sustainable, reliable, safe plants in the future. Advanced analytics (machine learning/AI) is becoming an integral part of digitalisation efforts.

Intelligent, AI-driven applications are at the core of the journey from automation to autonomous operations, enabling self-optimising applications, autonomously adjusting to different operating conditions and environments and orchestrating disparate applications to optimise the operations. Adoption of AI technology is key to delivering on the promise. The commercial success will depend not just on the technology readiness (maturity) but also on the organisational readiness to deploy the applications and sustain the value delivered. This will entail focusing on workforce development and organisational structure that can drive change management, successfully scale up, and deliver value to the business.

Simply bolting machine learning or AI solutions onto specific layers may help with engineering productivity but will not provide an optimisation step change or add millions in new margins to your bottom line. For example, our Performance Services offer digital applications and consulting services that help our customers to run sustainable, safe, reliable, and optimum operations.

Mar-2023
Andrew Ledlie, Solenis, aledlie@solenis.com

Refineries within the petrochemicals industry are increasingly employing digital technologies, including AI, that support their water treatment management efforts to achieve stricter sustainability targets for reducing water use and improving energy efficiency. This trend, because of the usefulness of these technologies, is likely to continue.

The latest digital approaches to water treatment in refineries leverage three key areas: instrumentation, remote monitoring, and predictive analytics using AI. AI is becoming a key tool for predicting at an early stage the scaling, corrosion, and fouling tendency within cooling water systems.

In terms of instrumentation, many innovative devices, including sensors, analysers, and controllers, have been developed in recent years. For example, Solenis developed a patented analyser that employs ultrasound to measure accurately fouling and deposition in situ. Consequently, when the analyser is used to measure fouling in heat exchangers, the heat exchangers do not need to be opened as frequently for inspection because the ultrasound device gives a real-time in situ picture of any fouling.

Remote monitoring is a powerful way to provide all key stakeholders instant access to critical information in real- time. This enables faster troubleshooting of emerging problems. Waiting for plant personnel to assemble and provide data or for experts to visit the site is costly when downtime or extended production slowdowns occur. Use of a trusted cloud platform, such as Solenis Cloud, addresses this need and allows all stakeholders to see the flow of problem resolution remotely in real-time, thereby reducing stress and providing peace of mind. This platform uses statistical process control tools and techniques to process and display data, thus enabling refinery operators to easily monitor and optimise the performance of their water treatment programmes.

Lastly, predictive analytical tools that crunch large amounts of data are being adopted more widely, provided the operator feels comfortable sharing their data. Solenis’ HexEval performance monitoring program for heat exchangers is an example of AI enabling decision-makers to identify, with confidence, which heat exchangers pose the greatest threat to reliable operation due to scale, corrosion, and/or fouling. Consequently, plant personnel can develop appropriate plans to optimise heat exchanger efficiency. Digital twins are another form of emerging AI that allows refiners to model the impact of process changes before implementation.

With refinery operators striving to improve sustainability, for example, by reducing water use, these digital solutions are critical to ensuring that production, efficiency, and asset protection are not sacrificed in exchange for sustainability improvements. Reducing water use, for example, often increases scaling, corrosion, and fouling, which all negatively affect energy use, maintenance costs, and downtime. Seeing in real-time or via digital twins how each step change in water use reduction affects key performance indicators is powerful and readily available through industry leaders.

Mar-2023
Geannie Gardner, KBC (A Yokogawa Company), geannie.gardner@kbc.global

Digital technologies are significant enablers of smart manufacturing throughout petrochemical enterprises, from plant floor to boardroom. Consider some examples:
• Looking from the bottom up IIoT sensors, coupled with robotics, drone technology, and AI/ML, enable plant managers to change how rotating equipment is monitored and maintained. Now, operators are freed from daily inspection rounds because the sensors plus AI flag early warnings of trouble, allowing robotics to guide visual inspections.

Reliability improves with continual monitoring. Combining predictive AI algorithms with better data directs maintenance efforts to where it is needed, which leads to fewer unplanned shutdowns.

• Operator roles change Additional sensor information and AI guidance are accessible via dashboards and 3D visualisations, readily available simulations (first principles, ML-based or hybrid), plant knowledge bases, and so on. This data provides rich insight into plant performance. Therefore, operators work with true digital twins of their plant, and their role evolves to monitoring, reviewing, and approving the outputs of the various AI/ML applications.

• More flexibility to operations scheduling AI-driven algorithms married to new analytic techniques can be applied in plant control systems, allowing for faster switches between product grades.

• Looking top-down Digitalisation gives organisations consolidated access to information, cutting through traditional silos. This, in turn, allows more automated workflows and hence increased business agility. For example, common scenarios around supply and demand opportunities can be examined by AI-enabled workflows, with planners and schedulers reviewing recommendations rather than running the analyses themselves. Accepted recommendations can be implemented automatically from the ERP system to plant floor, increasing responsiveness.

• Travelling the industrial autonomy journey The opportunity to be fully achieved involves digitising the human experience and knowledge, then coding this data to analyse key decisions, assess the effectiveness of human machine interfaces, and capture key learnings. Manufacturers that embrace these rapidly developing digital technologies and techniques, including AI/ML, are likely to survive and thrive in this volatility, uncertainty, complexity, and ambiguity (VUCA) world. As a result, they should be able to operate more nimbly, with more empowered workforces and report greater returns on capital.

Mar-2023