July 2018

Special Focus: Refinery of the future

Digitalization for the refinery and plant of the future

Hydrocarbon Processing Industry (HPI) companies are using digitalization in concert with improved processes and skills to squeeze more productivity from existing assets.

Agnihotri, R., IBM Corp.

Hydrocarbon Processing Industry (HPI) companies are using digitalization in concert with improved processes and skills to squeeze more productivity from existing assets. Low oil prices have been one of the drivers1 for companies to invest in analytics, deploy automation more quickly, move information technology (IT) to the cloud and invest in big data.

The same low oil prices have spawned many downstream greenfield projects. However, with the onset of Industry 4.0, owners of capital projects want to build plants that will not become technologically obsolete in the 4 yr–5 yr it takes to design and build them.

Existing brownfield sites seek to use digitalization to promote HPI-leading performance in safety, profits, compliance and sustainability, and to expand their horizons by developing new business models.

The design, construction and operation of the digital “plant of the future,” as well as how to keep it refreshed, are explored here.

Vision for the HPI

The enterprise of the future in refining and chemicals is envisioned to have the following capabilities:

  1. An intelligent, predictive and near-real-time business, where near-real-time data and models drive optimization across the entire hydrocarbon value chain and provide new, actionable, role-based insights, with personalized delivery, to maximize margins, safety and compliance.
  2. Proactive intelligence that discovers and defines impending disruptive or abnormal events (price changes, asset failure, weather, etc.) and mitigates their impact.
  3. Concurrent real-time supply chain management and optimization in planning, scheduling and execution across supply, manufacturing, distribution and retail to maximize profits and reduce costs, inventory and working capital.
  4. A focus on production excellence to manufacture at or above capacity, reduce feed costs and energy consumption, produce higher-value products, reduce or eliminate losses, maximize profits from blending products and reduce off-grade product in transitions, among others.
  5. Continuous improvement in asset availability, reliability and integrity through increased preventive, predictive and prescriptive maintenance and reduced turnaround (TA) time, providing specific, complete, multi-modal information
    for the task at hand.
  6. A digital twin of physical assets that serves as an information source (e.g., business and process models, data, documents, drawings, 2D/3D models) and can provide asset lifecycle management, a cloud-based, integrated asset data solution through the life of an asset.
  7. Sustainability and compliance with environmental and other regulations to conduct business in a social and environmentally responsible manner for the betterment of the community.
  8. Reduced risk to personnel safety and health through use of technology and automation such as drones, safety applications, intelligent inspection and emergency response in hazardous situations.
  9. Product lifecycle management for innovation to increase market share, meet customer requirements and track, diagnose and correct any potential problems.
  10. New business models to continually reinvent the enterprise by taking advantage of new technologies and artificial intelligence (AI) that are disrupting the HPI.

A schematic of the plant of the future’s capabilities and enabling technologies is shown in FIG. 1.

FIG. 1. Plant of the future capabilities and enablers.

In recent surveys,2,3,4 HPI company executives indicated that they are now incorporating digital technologies as part of their business strategy (FIG. 2).

FIG. 2. Chemical companies view a combination of digital technologies as critical to their strategies.

Building the information management digital plant/enterprise

With Industry 4.0, operating companies face a sea of technology suppliers, not only of traditional business, manufacturing, engineering, instrumentation and control/IT systems, but also Internet of Things (IoT) sensors, cloud providers, edge devices, mobility and collaboration tools, software as a service (SaaS), AI and analytics.

For both greenfield and brownfield sites, these technologies and services must be deployed as part of an overall information management (IM) design that best enables the objectives and operations of the HPI enterprise or plant.

Main information contractor

To address these problems, the envisioned role of a main information contractor (MIC) is to build the IM environment and channel disparate technologies for best use in the design and operation of the refinery/plant of the future.

The MIC must have deep expertise in asset operations, be a leader in the architecture and systems required for this new digital landscape, agnostically manage a diverse ecosystem of suppliers and have the proven experience to help clients transform to a digital enterprise.

Best-of-class companies successfully render their experience globally to:

  • Work with clients to review their business objectives, and target operating model and business processes [e.g., planning, procurement, manufacturing, asset management, sales and distribution, and health, safety and environment (HSE)].
  • Design and implement IM systems for the site/enterprise [e.g., enterprise resource planning (ERP) business systems, manufacturing operations management (MOM), or manufacturing execution systems (MESs), engineering systems, mobility, industrial IT (access control, CCTV, etc.) and IT infrastructure]. These systems now include the cloud, SaaS, analytics and AI.
  • Design and implement the architecture that connects all relevant structured and unstructured data and disparate systems.
  • Aggregate and integrate enterprise information for collaboration.
  • Provide cybersecurity for IM systems and the IT/operational technology (OT) interface.
  • Implement asset lifecycle management from inception to decommissioning, utilizing a digital twin that is a virtual representation of the physical assets.
  • Design and implement turnkey fit-for-purpose command/monitoring centers that can enhance collaboration and cognitive analytics while leveraging remote subject matter experts (SMEs).

Brownfield sites. All MIC activities are applicable for brownfield sites, except for those relating to plant design and construction, and engineering contractors.

Brownfield sites can use digitalization to convert legacy data to digital form, standardize and make it available for wider use throughout the enterprise for better decision-making across different functional groups. Digitized plant data is now available for analysis beyond the control or records rooms, and helps in management of change (MoC). Older, time-consuming processes can be modified to harness mobility, the cloud and analytical forecasts.

Greenfield sites. In a capital project, IM systems must be installed within a regimen of rules, procedures and schedule that are aligned with the engineering, procurement, construction (EPC) and commissioning of the site.

The MIC’s objective in greenfield plant design and construction is to integrate engineering, operational, information and digital technologies to not only build a plant, but also to operate it efficiently. IM is geared to shorten commissioning and startup times and assist the EPC contractor in the build phase.

The MIC assists and liaises with engineering contractors, suppliers and the main automation contractor (MAC) for design and build activities during the EPC and commissioning phases.

The MIC can utilize digital and cognitive techniques to bring efficiencies to the EPC phases. Examples include:

  • Save project time by managing engineering and vendor data, from inception in project drawings and documents and 2D/3D models, to direct digital consumption in operational systems.
  • Improve safety and efficiency through a “safe site” app for personnel safety, predicting project health and performance based on history, and a dashboard for project management.
  • Reduce time and costs via cognitive analytics apps as “advisors” for cognitive procurement, predictive asset optimization, remote inspection of assemblies and field inspection.
  • Improve engineering contractor efficiency through track-and-trace for materials management.
  • Reduce costs by setting up the information environment as a cloud with all systems, including engineering data, as a service.

Benefits of digital enterprise

Digitalization in the downstream segment provides an array of benefits consistent with real-time, proactive response and better decision-making. In a recent survey, HPI executives stated that they see diversified business value from new technologies (TABLE 1).

As an example, the IoT, the cloud and analytics have been applied to provide track-and-trace in the supply chain for field assets that are needed and transported across the globe, saving significant cost and time.

Companies are focused on increasing production output by 1%–3% by understanding variability in performance by plant and by shift, and by driving uniformity in response to deviations in actual vs. model-based expectations. They are also improving asset availability to 98% for greater commercial gain in the competitive marketplace.

The continuous leveraging of technology is used to drive improved, safer inspection techniques, and data collected from those digital processes are used to eliminate forced outages and supply disruptions.

Companies are using real-time insights to optimize working capital and establish more efficient cost structures. Plants are constantly improving the largest cash operating expense outside of feedstock costs, with 5%–7% energy and fuel savings from improved modeling and insights from cognitive-enabled energy programs.

TABLE 2 shows several changes caused by a shift to a new digital paradigm.

Reinventing the enterprise in the digital economy

Digitalization does not just lead to accurate data or faster decisions. Disruption has fundamentally changed the chemicals and petroleum industries, driving new business models and causing enterprises to reinvent themselves.

FIG. 3. Disruption has fundamentally changed the chemicals and refining industries.

A series of industry C-suite studies2,3,4 have supported such findings. As shown in FIG. 3, changes due to disruption include:

  • Organizations are thinking beyond the old value chain to imagine new ecosystems
  • Digitalization interconnects products, value chains and business models
  • New entrants with new business models have transformed to next-generation operations, accelerating the transition to as-a-service infrastructure and applications to drive agility and savings
  • The pervasive use of mobile technologies and wearables is forcing a redesign of functions and internal processes.

Examples of process disruption include:

  • Petroleum—The companies’ remote command room/centers, using the IoT, can monitor pump electrical variable-speed drives or other equipment from the cloud; commercial drones with sophisticated electro-optical and infrared sensors can be used for ground surveillance and equipment inspection coupled with visual analysis; sensors and drones monitor hazardous incidents with gas leaks; wearables remotely monitor personnel location and health, as well as skids for remote maintenance and invoicing.
  • Chemicals—Radio frequency identifications (RFIDs) for rail cars to speed up freight; real-time visibility of hazardous materials transport; an optical sensor that can measure distance to an object; sensors and IoT connectivity to enable farmers to optimize water, energy and inputs; and a cloud platform ecosystem that allows companies to work together to manage health and nutrition for livestock.

From digitalization to reinvention. Digital reinvention is not fragmented nor specific; it involves a fundamental reimagining of how a refining or chemicals organization operates and how it engages with its environment. 

Studies reveal that companies are moving from an organization-centric economy to first an individual-centered economy (with insightful customized experiences), and then to an “everyone-to-everyone” economy characterized by:

  • Value creation driven by collaboration and connectedness
  • Multi-directional communication
  • Consumers becoming an intrinsic part of organizations.
FIG. 4. Chemical and refining companies need to embrace digital drivers.

Envisioning new models and developing the means and resources to achieve them are recommended (FIG. 4).

FIG. 5. Foundational principles of modern enterprises.

Pillars of the digital plant/enterprise

A modern enterprise utilizes the foundational principles of visibility, pervasive connectivity and collaboration, and actionable insight (FIG. 5).

Visibility and data. Beyond the data that refineries and process plants already generate, a proliferation of IOT-related information has emerged; an estimated 25 B devices will be installed by 2020. These devices include:

  • Embedded sensors and devices that measure process variables, vibration, asset or personnel location, motion or orientation
  • Videos, drones, smartphones, kiosks and weather data
  • Geographic information systems (GISs)
  • Implanted medical devices and unstructured information from social media, emails, live streams, etc.
FIG. 6. Putting it all together: the IoT from chip to the cloud to applications. An MIC manages the ecosystem and helps companies build and deploy IoT end-to-end.

Having data is not enough; an MIC must build an end-to-end IoT environment that utilizes that data within solutions (FIG. 6).

Pervasive connectivity, integration and collaboration. Unique value is created by the ability to integrate data and systems, and extract insights to create an intelligent environment that benefits the business, and then to offer it as a service.

A modern architecture requires an IoT platform that can perform big data ingress and management, integration and analytics, while managing the accompanying security risks. This architecture utilizes open-standards-based communications (such as MQTT and HTTPS), advanced capabilities for data storage, caching and transformation, integration middleware, and dashboard and console reporting. The IT architecture must handle both virtual cloud and on-premise solutions [SaaS, Big Data (Data Lake)], and should integrate with leading cloud platforms so that customers are not forced to choose proprietary tech stacks.

For risk management, the IoT platform should include core security features for devices, data and connections; enforce appropriate levels of security and privacy to IoT solutions; and use security analytics. The platform must support a secure decentralized system, such as Blockchain.

Integration. Within enterprises, legacy and commercial off-the-shelf (COTS) applications have been integrated horizontally (across units and sites) and vertically (across control/MES/business layers) via middleware.

In the Industry 4.0 era, these integration capabilities are extended laterally across an end-to-end process, organization, industry or value chain, assimilating unstructured data not associated with any system—such as Web 2.0-type interconnectivity across people/communities, web searches, etc. This is used to support new business models and customer experience with new, added benefits for the HPI.

Collaboration. The components of collaboration include data, information management and virtual/physical rooms using applications. This collaboration is for groups, specific activities (morning meetings, weather/hurricane), or for a hierarchy of users (business leaders, operations leaders, etc.). In a command/remote monitoring center, users can collaborate to remotely monitor/manage operations, provide emergency response, avoid hazards or leverage SMEs across multiple sites.

FIG. 7. Cognitive technologies are delivering business value and insights to enable decisions.

Analytics and optimization

Businesses are gaining valuable and actionable insight using predictive and cognitive analytics to forecast usage and operating conditions, as illustrated in FIG. 7.

Predictive analytics. A predictive analytics platform (both onsite and SaaS) encompasses a variety of techniques, including statistics, data mining, machine learning and game theory, to analyze large amounts of data to discover relationships that can be used to make predictions of future behavior or events. Such a platform can be applied to the vast amount of unutilized data in a process enterprise to discover new relationships and improve business decisions. User benefits can include increased profits or reduced risk by: identifying product price points that are most consistent with high sales and/or margins; predicting equipment pieces that are likely to fail, or detecting abnormal behavior in process plants.

FIG. 8. Cognitive systems can perceive, reason, relate and learn.

Cognitive analytics. Cognitive systems are analogous to the human brain. Unlike programmable systems that are based on rules that tell a computer how to react, cognitive systems can perceive, reason, relate and learn (FIG. 8).

Cognitive analytics mimic human cognitive capability as “cognitive advisors” through:

  • Natural language processing (NLP)
  • Advanced machine learning to predict, act and learn from previous experiences
  • Visual analytics to boost learning pace and experience development.

These abilities allow a cognitive advisor to initially respond to questions and then make predictions pertaining to the specific domain. The proprietary cognitive technology’s speed of ingesting and interpreting information, coupled with partner company training, has been used to create advisors for numerous functions—e.g., production, crude procurement, asset management, pricing and project health.

Digital supply chain optimization. A true, near-real-time management and optimization of the entire supply chain will require ubiquitous visibility and a concurrent computation of planning, scheduling and execution of real-time procurement, supply, manufacturing and distribution at a global, regional and local level. The appropriate transport, inventory and manufacturing related models and constraints, and a globally refreshed data store, are essential prerequisites. Optimization technologies have been used for a host of custom large-scale supply chain optimization applications in downstream and chemicals. The goal is to use high-performance computing (HPC) to deliver the results to mobile devices on a near-real-time basis.

Cybersecurity. Pervasive digital security environments have been set up and maintained that can prevent and defend against cyberattacks, threats and the denial of service from both external and internal perpetrators. The threat risks increase with greater IT/OT connectivity and mobile devices. The approach discussed here includes: stopping advanced threats by engaging analytics and insights for smarter and more integrated defense; protecting critical assets by using context-aware controls to prevent unauthorized access and data loss; safeguarding cloud and mobile technologies, and utilizing IT transformation to build a new, stronger security posture; optimizing security programs and employing experts to modernize security and reduce complexity and cost.

Blockchain. Blockchain enables immutable, transparent and auditable business transactions among participants and suppliers, distributors and partners. Its most common applications are in finance, trading, raw materials and spare parts procurement, products distribution and materials management in greenfield projects. HP

NOTES

a IBM’s Watson cognitive technology

LITERATURE CITED

  1. Oil & Gas CIO Survey, IDC, February 2015.
  2. “Global ecosystem survey,” IBM Institute for Business Value, 2016.
  3. “IBM chemicals digital transformation study,” IBM Institute for Business Value, 2017.
  4. “IBM petroleum digital transformation study,” IBM Institute for Business Value, 2018.

The Author

Related Articles

From the Archive

Comments