Feb-2024

Revolutionising refining with digital twins

Exploring applications and outputs across the refinery landscape. Refiners are at the crossroads of innovation and challenge.

Michelle Wicmandy, Jagadesh Donepudi and Rodolfo Tellez-Schmill
KBC (A Yokogawa Company)

Article Summary

They are facing disruptions ranging from oil price volatility to the complexities of the global energy transition. Adding to this complexity, Offshore Technology claims India is expanding its pipeline network to more than 29,600 km by 2025. This significant expansion is roughly three-quarters of the Earth’s circumference. As the industry confronts these uncertainties, securing the integrity of this expanding pipeline infrastructure becomes crucial for meeting the nation’s growing energy demands while reducing risks and accidents that can harm people, profits, and property.¹

Refiners are at the crossroads of innovation and challenge. They are facing disruptions ranging from oil price volatility to the complexities of the global energy transition. Adding to this complexity, Offshore Technology claims India is expanding its pipeline network to more than 29,600 km by 2025. This significant expansion is roughly three-quarters of the Earth’s circumference. As the industry confronts these uncertainties, securing the integrity of this expanding pipeline infrastructure becomes crucial for meeting the nation’s growing energy demands while reducing risks and accidents that can harm people, profits, and property.¹

Navigating new refining challenges
To navigate these new challenges, the refining industry is revamping how it produces, uses, and manages energy.² Although existing assets have data acquisition capabilities, the hurdle lies in reviewing, cleaning, and assessing this data. This process is necessary to determine both current and future operating conditions, as well as meet safety and environmental regulations.

In this journey, Indian refiners are in the process of implementing digital twins across various refinery process units for long-term sustainability.³ This initiative centres around creating digital twins for diverse applications, which provides real-time visualisation of key performance indicators (KPIs) and benchmark parameters. By using digital twins, refiners can improve the plant’s efficiency and productivity while reducing miscommunication, data waste, and labour costs. Essentially, digital twin technology is revolutionising the way refiners operate and paving the way to long-term profitability.⁴

Furthermore, it is evident that all aspects of an entire refining supply chain are highly interrelated and complex. Thus, integrating digital twins into the supply chain delivers added value, too, by optimising processes, energy consumption, and control applications such as real-time optimisation (RTO) and advanced process control (APC) systems. In the supply chain, these applications help bridge the gap between forecasting and actual operations.4 Validating these gaps, or delta vectors, uncovers the disparities between the planned and actual operations in terms of demand, inventory, and production. By validating these delta vectors, supply chain managers can quickly assess and address gaps in their models and processes to accommodate changes in inputs and outputs.¹

In regard to process optimisation, which is integrated with supply-side optimisation for power, steam, and utility balances, energy demand takes centre stage. The comparison between linear programming (LP), actual data, and simulation enables automated vector updates and model recalibrations via artificial intelligence (AI) and machine learning (ML) methods.

The following discussions explore various applications, including KPI visualisation, production accounting, LP model updates, process optimisation, real-time optimisation, and corrosion monitoring. The digital twin architecture includes connecting process models through open platform communications unified architecture (OPC UA) with historians to ensure proper calibration.

Digital twin technology
Digital twins offer a solution to transform the oil and gas industry by improving efficiency and reducing risk. According to researchers,⁵ these virtual models of physical assets seamlessly connect with real-time data across assets, columns, reactors, pipinmg, and equipment. Despite changes in crude quality, catalyst composition, and process conditions, digital twins continuously analyse industrial data to predict and optimise processes.⁴ Their perpetual operation brings multiple benefits, such as asset monitoring in planning and scheduling studies, refinery-wide flow sheeting, real-time optimisation, and more.

Furthermore, digital twins set benchmarks for both the quantity and quality of units. These benchmarks are then transmitted to the RTO/APC layer for optimisation on a global scale.⁶ This iterative process involves ongoing validation and adjustments to maximise benefits derived through the APC in a closed loop. The APC, armed with its dynamic process model, aims to stabilise operations and reduce fluctuations. It effectively implements the desired setpoint from the RTO to achieve closed-loop optimisation.⁵ The optimiser identifies the optimal operational state and communicates it to the APC.

Implementing a digital twin starts with identifying possibilities and choosing a pilot configuration with the highest ROI. After implementation, the digital twin becomes an integral part of the enterprise’s digital backbone.⁴ The final step involves monitoring the value created and modifying the digital twin to maximise economic benefits.

Refiners apply digital twins in various applications to improve plant performance. These applications include visualising KPIs for performance tracking, reconciling data in production accounting, updating LP models in the supply chain, optimising processes to improve yield and energy, conducting real-time optimisation through quick gain calculations, and managing corrosion to monitor equipment and system degradation. These applications underscore the value of digitalisation in the refining process and are addressed in the remainder of this study.

Digital twin architecture
The digital twins are process models connected through OPCs with historians such as IP.21, Exa Quantum, OSI PI, or any other real-time data gateways, as shown in Figure 1. The models are calibrated using test data to ensure energy and mass balance accuracy. After calibrating the model, it is scheduled to run, and the results appear on dashboards. Other applications use these to generate advanced analytics.⁴

The success achieved from this system depends on whether the model is accurate and current. An outdated model limits the operation’s potential, resulting in value leakage, lost opportunities, and substantial financial costs.