April 2021

Special Focus: Clean Fuels

Leveraging digital technologies to create the smart renewable diesel facility

Many companies are modifying existing crude refineries or building grassroots renewable diesel facilities to produce drop-in, green renewable diesel from a variety of agriculturally derived triglyceride feedstocks.

Many companies are modifying existing crude refineries or building grassroots renewable diesel facilities to produce drop-in, green renewable diesel from a variety of agriculturally derived triglyceride feedstocks. The key drivers for this global trend are regulatory, tax and social pressures to lower the carbon intensity of transportation fuels.

The renewable diesel production process offers many challenges that can impact safety, reliability and profitability. These challenges range from feedstock availability, variability, gumming, metals, corrosion and wax formation in the pretreatment unit (PTU) or feed pretreatment train that can affect safety, operability, catalytic-based yield and profitability of the renewable diesel unit (RDU). The RDU also has a high-reaction exotherm and is vulnerable to several types of corrosion, such as high-temperature hydrogen attack (HTHA), which can be addressed using approaches such as analytics, integrity operating windows, real-time decision support and automated safety shutdown systems.

Companies that are moving into the production of renewable diesels can mitigate the safety and processing challenges in the PTU and RDU by adopting digital technologies used by leading hydrocarbon processing industry (HPI) companies. Such operators have achieved a 2%–4% increase in effective capacity by improving asset reliability; providing real-time, proactive decision support; reducing operating and maintenance (O&M) costs by 3%–5%; increasing EBITDA performance by 3%–5%; and improving safety.

These leading HPI companies are using a critical real-time data infrastructurea that enables subject matter experts (SMEs) to configure no-code operational digital twinsb of asset classes (e.g., pumps and reactors) and develop smart applications like HTHA analytics. The asset class templates are then used as building blocks to create an operational digital replica of their PTU, RDU and associated supporting production infrastructure such as hydrogen sources, utilities and logistics. This digital infrastructure enables a layered approach to descriptive, diagnostic, predictive and prescriptive analytics, and provides proactive real-time, exception-based decision support.

This article will present how this leading digital technology can enable operational intelligence and continuous improvement, and also increase flexibility, capacity and profitability for renewable diesel production in both grassroots and retrofit scenarios.

The challenges of renewable diesel manufacturing: Safety, operations and profitability

Two primary approaches exist to the development of renewable diesel production. Each will be presented, with associated reasoning and a similar method for leveraging the operational data infrastructurea.

Approach 1: Existing petroleum refinery retrofit

Companies are moving forward with modifications of existing petroleum refineries for cost, time to market, margin and best alternatives to shutting down the petroleum refinery, with the associated environmental and regulatory challenges. However, modifications of existing hydrotreating facilities and associated infrastructure present the following six challenges1 that a real-time data infrastructurea can help mitigate:

  1. High-reaction exotherm and associated emergency depressurization systems and liquid recycle and quench systems—The operational data infrastructurea is used by leading HPI companies to help mitigate this challenge through proactive, exception-based decision support. The operating windows (OWs) and integrity operating windows (IOWs) can be configured to combine various process parameters and to provide improved awareness of the operating regimes that need attention. The difference between OWs and IOWs is the degree of criticality and the required time to respond to mitigate escalating situations, including the triggering of the emergency depressurization system. FIG. 1 presents an example of an IOW developed and implemented by MOL, a Hungarian energy company, in its six HPI facilities. MOL developed a portfolio of OWs and IOWs based on industry standards and the experience of its SMEs, and by configuring associated OW and IOW digital asset templates. These templates are then put into production across its facilities, enabling changes to be propagated quickly and efficiently.
    FIG. 1. Example of configuring OWs and IOWs in the real-time operational data infrastructure.2
  2. Feed train fouling—Managing polymerization and associated gumming and fouling to optimize asset performance is another common, powerful application of the operational data infrastructurea, which allows the configuration of asset health indexes by SMEs to determine the health of various assets, such as pumps, exchangers, heat tracing and valves. The health indexes can then be used to incorporate indicators of possible issues, such as fouling from polymerization in the exchanger train or a drop in pump efficiency. The health index can be aggregated across the PTU and/or the RDU to roll up to a key performance indicator (KPI) on the summary dashboard, enabling drill-down and diagnostic investigation. Leading HPI companies are also integrating the health index with their maintenance management systems (MMSs) to trigger work orders and to link in metadata found in the MMS, such as the last maintenance date, spare parts inventory, or manufacturer’s make and model information, which are commonly used in the health index calculation. FIG. 2 provides an example of an advanced condition-based maintenance (ACBM) system created using a data infrastructurea. MOL has configured health indexes for all its critical assets (such as all rotating equipment, exchangers, crucial valves and heaters),
    and has integrated the indexes with its MMS.
    FIG. 2. Example of use of the real-time operational data infrastructure for advanced CBM.2
  3. HTHA corrosion—As with other HPI processes that utilize high temperatures and hydrogen in carbon alloy metal environments, HTHA can occur, leading to embrittlement and loss of containment. FIG. 3 illustrates MOL’s use of an HTHA smart asset template, and demonstrates how a data infrastructurea can address various corrosion regimes, including HTHA, carbonic acid and others. The template uses a table lookup for the regressed coefficients from the Nelson curve, which provides correlation of hydrogen and temperatures with various carbon steel alloys to determine when an alloy node is approaching or has entered an HTHA regime. The start and end of this event can be configured, which is, in effect, codification of the SMEs’ knowledge. Notifications can be triggered in both cases, and analytics can determine the length of exposure and the exposure details. A similar approach can be used for any corrosion or process condition.
    FIG. 3. Example of a configured HTHA corrosion smart asset template.2
  4. Carbonic acid corrosion—As a result of the conversion of triglycerides to hydrocarbons, water and large amounts of carbon dioxide are formed. Apart from these substances needing to be handled safely on an individual basis, they can combine to form carbonic acid that can be very corrosive in the liquid effluent air coolers and sour water disposal systems. To address this corrosion regime, the operational data infrastructurea can be used to create OWs and IOWs, in addition to specific carbonic acid analytics and associated health indexes to be used to notify and proactively inform operators so that corrective actions can be taken. The start and end of events can be configured, like the HTHA application, to capture length of exposure, severity and causality to enable more effective decision making, including modifications in operations, inspection and metallurgy.
  5. Catalyst performance and associated yield optimization—To optimize the catalyst performance and yield in PTUs and RDUs, many leading HPI companies are digitally integrating with their catalyst providers to facilitate advanced unit and catalyst performance. To address the issues of data/cybersecurity, ownership and governance, these leading HPI companies are leveraging an extension of the operational
    data infrastructurea as an enabler of this powerful capability. Instead of using other data transfer methods that are labor intensive and have variable lag times, this digital bridge addresses these issues (FIG. 4).
    FIG. 4. Example of optimizing catalysts’ performance with near-real-time remote monitoring.3
  6. Near-real-time modeling and optimization—Another optimization approach being used by leading HPI companies is the deep integration between the operational data infrastructure and rigorous, first-principle simulation models for PTUs and RDUs, including the inclusion of financial information. The operational data infrastructurea and associated smart asset templates are leveraged to provide context and validated data sets to the models. Optimum targets and forecasts are outputs of the models that are put back into the operational data infrastructure as “future data.” This information is used to perform plan vs. actual analytics, and to enable proactive, data-based decision support, including key information such as lost margin opportunities, the asset heath of PTUs and RDUs, and other intelligence (FIG. 5).
    FIG. 5. Example of deep integration between the operational data infrastructure and modeling software.4

Approach 2: A new grassroots renewable diesel facility

Many non-petroleum refining companies are choosing to design, build and operate new renewable diesel facilities that provide greater flexibility to optimize the entire facility, including feedstock sources and necessary purification in a PTU; the design and catalyst selection in the RDU; the design of the required infrastructure, including utilities, hydrogen sources, blending and logistics; and waste disposal.

The operational challenges, which are fewer because of the flexibility afforded by a grassroots design and build, are still present, and the mitigation opportunities from the use of an operational data infrastructure remain applicable. The operational data infrastructurea can be configured to address the challenges and opportunities of a grassroots design and build. One key advantage of leveraging an operational data infrastructure experienced by many HPI companies is the ability to accelerate startup, improve warranty validation, and reduce the number of applications and solutions by more than 50%.5

Creating the smart renewable facility with an operational data infrastructure. A real-time operational infrastructurea is an agnostic, open, scalable and reliable technology specifically designed for critical operations to deliver operational data in a reliable way to stakeholders and applications. The infrastructure must enable self-service analytics, deliver all required context for operational intelligence, and have the following capabilities (FIG. 6):

FIG. 6. A critical operations integration, applications and analytics infrastructure.
  • Secure integration of time-series operational data from the distributed control system (DCS), supervisory control and data acquisition (SCADA), and Industrial Internet of Things (IIoT) systems
  • Abstraction of diverse tag and asset names into a standard company lexicon and asset hierarchy
  • Integration of metadata, including engineering data and information from the MMS (FIG. 2)
  • Normalization of units of measure, time zones and asset descriptions
  • Configuration of traditional operational applications, such as energy management, environmental compliance and KPI-driven dashboards
  • Use of a “layers of analytics” framework and strategy to provide the analytics foundation via configurable descriptive, diagnostic and simple predictive analytics.

No-code operational digital twinsb configured and supported by SMEs. A digital twin is a replica of a physical asset (such as a heat exchanger, pump or compressor) comprising attributes, calculations, KPIs, empirical correlations and models of varying complexity. Contrary to the hype, digital twins have been around since the 1960s. However, today’s operationally focused digital twins are dramatically more robust and sophisticated in their ease of use, approach and capability to develop, evolve and leverage in a renewable diesel plant.

Most digital twins require IT, data scientists, machine learning, model integration and coding. They also have a limited ability to deal with data volume, velocity, variability and anomalies. They are difficult to scale, and struggle to address the anomalies of physical assets that have variability in vintage, makes, models and levels of instrumentation. Furthermore, real-time operational data and asset metadata (i.e., static information like equipment model and location) typically reside as tags in control systems, as well as in other databases and platforms, with accessibility issues and lack of naming standards limiting access to critical data that could potentially be leveraged to gain insights.

A key capability of the operational digital twin is operational data management that creates a system of record for operational data. The operational digital twin is created by the SMEs over time in an evolutionary way by first creating smart asset digital templates that consist of asset categories, attributes, calculations, event frames and notifications. Attributes are grouped in categories for providing ease of navigation and enabling the drag-and-drop configuration of smart display templates. These attributes consist of data references to real-time data sources like DCS, SCADA, programmable logic controllers (PLCs) and other systems, as well as linked tables into engineering data, the MMS and tabular correlations like the HTHA example. The digital twin provides configurable, no-code calculations and complex expressions by using one or more of the more than 110 time-based functions, such as the function library in Excel.

The smart asset template attributes are placeholders to enable the actual references and link tables when the templates are applied to an actual asset. The smart assets are combined to form a base asset hierarchy that can be used to create relative asset hierarchies for context and ease of navigation. Another key capability of smart asset templates is the ability to have base and relative templates with inheritance to allow for the anomalies commonly found in asset classes.

The digital operational infrastructurea can enable asset anomaly detection by allowing the SME to create or modify anomaly expressions and then to test the expressions by backcasting (i.e., running the expression back into the operational history). Once satisfied with the expression, the SME can then forward-cast this modified expression or event detection algorithm to other assets that utilize the same digital twin template. This powerful capability enables continuous improvement of calculations, expressions and event analytics over time, as well as comparison of similar expression results, KPIs or events as part of the diagnostic process. FIG. 7 is an illustration of an asset hierarchy associated with a no-code digital twin.

FIG. 7. Operational data management in a no-code operational digital twin.

The no-code digital twin can include operational metadata, engineering data and MMS metadata to create a pump curve and overlay with real-time pump performance (FIG. 8). Actual pump performance can be viewed over time to see historical performance and, for future references, be based on other forecasted information. This is also an example of how the output of calculations and expressions can be historized to enable their use in other calculations and analytics.

FIG. 8. Example of an operational digital twin and integrated data references used in analytics and visualization.

Consolidating the concepts, capabilities and applications, FIG. 9 illustrates how smart asset templates are leveraged to create a digital replica of a physical renewable diesel facility and to form the foundation for a portfolio of dashboards, KPIs and advanced decision support capabilities.

FIG. 9. A smart renewable diesel facility configuration from digital twin asset building blocks.

A ‘layers of analytics’ strategy for renewable diesel facilities. Terms such as advanced analytics, machine learning, big data and artificial intelligence (AI) appear pervasively in marketing literature, but can lead to confusion, failed projects and significant lost opportunity costs.

The most successful operators achieve value from analytics by first defining an analytics framework, along with the types of analytics required, and then selecting fit-for-purpose technologies. They use a “layers of analytics” strategy, which considers incremental cost vs. incremental value as they move to more complex analytical methods. The costs include not only the technology, but also the costs associated with lost time to value, scalability, configuration, sustainment and risk of attainment.

The foundation of this “layers of analytics” approach (FIG. 10) relies on the use of an operational data infrastructurea to enable SMEs, not IT, to configure real-time descriptive, diagnostic and simple predictive analytics by using formulas, empirical correlations and rule-based expressions. These lower-level analytics form the foundation for more advanced predictive, prescriptive and adaptive analytics that use machine learning and other methods, and require collaborative support from data science teams.

FIG. 10. Using incremental cost/value evolution through layers of analytics and hybrid data lakes.

These foundational analytical layers generally provide over 80% of the value for about 20% of the cost vs. more advanced analytical layers that only use technologies such as machine learning.

Once higher layers of analytics are utilized, it is imperative to feed back the results of these advanced layers to the lower-level layers as forecasts or targets, where appropriate, to operationalize the advanced analytical output. This is key to the development of the smart renewable diesel, as results from the integration with process simulation optimization models and financial data for real-time gas plant financial optimization are fed back to the operational data infrastructure.a


The growth of renewable diesel, by either building new facilities or modifying existing petroleum refineries, is global, and opportunities are expanding rapidly to address social, regulatory and financial needs and opportunities. Renewable diesel PTUs and RDUs present both operating challenges and opportunities that can successfully be addressed to leverage a proven, powerful and configurable operational data infrastructurea used by many HPI companies worldwide. The key is the ability to enable SMEs to configure smart asset digital twins that can be combined to create an operational digital twin of the renewable diesel facility, including associated utility and infrastructure systems.

The operational data infrastructure forms the foundation for real-time operational intelligence and proactive, exception-based decision support. It also enables the use of modern digital technologies, such as analytics, financial base modeling and optimization, and safe and secure ecosystem digital integration.

The result is the ability to increase safety and to mitigate processing challenges in the PTU and RDU, as well as to increase effective capacity by 2%–4% by boosting asset reliability, reduce O&M costs by 3%–5%, and increase EBITDA performance by 3%–5%. HP


         a Refers to OSIsoft’s PI System
         b Refers to PI System’s Asset Framework (AF)


  1. Chan, E., “Converting a Petroleum Diesel Refinery for Renewable Diesel,” Burns & McDonnell, December 2020, online: https://www.burnsmcd.com/insightsnews/tech/converting-petroleum-refinery-for-renewable-diesel
  2. Komróczki, T., “Supporting Strategic Initiatives at MOL with the PI System Infrastructure,” OSIsoft PI World 2015, April 2015.
  3. Haragovics, M., M. Bubálik, and A. Frouillac, “Digital Bridge (PI Cloud Connect) Between Axens and MOL to Maximize Profit from Conversion Catalysts,” OSIsoft PI World 2018, April 2018.
  4. Tate, S. and J. Rose, “Best Practices in the Integration of Modeling Software with the PI System,” OSIsoft PI World 2019, April 2019.
  5. Mandani, M. and A. Brodskiy, “Overview of the PI System Enabled Integrated Refinery Information System (IRIS) at YASREF,” OSIsoft PI World 2015, April 2015.

The Author

Related Articles

From the Archive



{{ error }}
{{ comment.comment.Name }} • {{ comment.timeAgo }}
{{ comment.comment.Text }}