Digital Twinning in Pneumatic Conveying: The Future of Process Control

Digital twinning is an influential concept at the heart of the fourth industrial revolution, which is a virtual representation of a physical system capable of simulating and mirroring real-world conditions.

This sophisticated concept opens up many possibilities for industries, particularly in pneumatic conveying, where energy-saving, process quality enhancement, and predictive analytics are involved.

Revealing the world of digital twinning starts with a fundamental question: what is a digital twin?

At its core, a digital twin is a virtual copy of a physical system capable of representing a component, a whole machine, an entire plant, or even an expansive city.

The scale and type of digital twins vary widely, depending on their respective applications. However, it is crucial to grasp that digital twins extend beyond simple 3D visualisation. It represents the underlying dynamic of the system and maintains real-time communication with its physical counterparts.

As with any novel and rapidly evolving technology, the concept of a digital twin seems unclear due to various definitions rooted in the context. To provide clarity, especially in the process industry context, consider the following definition provided by the author:

In the context of the process industry, this common goal can range from energy-saving, preventing specific conditions to enhancing the quality of a process in various aspects.

Digital Twin, Digital Model, and Digital Shadow: What's the Difference?

Investigating more deeply, the distinction between a mathematical model with a 3D model and a digital twin need explanation. A system can be represented with mathematical and 3D models integrated, but this does not qualify as a digital twin. Such a representation is called a digital model, primarily due to its lack of real-time communication with the physical counterpart.

An evolution of the digital model is the digital shadow. Here, the physical system feeds information to the digital model in real time. However, the absence of a reciprocal feedback communication loop back to the physical system still prevents it from being classified as a digital twin. Once this two-way communication is established and the digital and physical versions have a common goal to achieve, we have a digital twin.

Digital Twins in Pneumatic Conveying

In the domain of pneumatic conveying, digital twins can serve many purposes. For example, consider a typical pneumatic conveying system which includes a blower/compressor, blow tank/hopper, and a receiver (see figure 1).

Among the major challenges with such systems, two critical challenges are reducing power consumption and particle breakage. These issues primarily arise due to operation levels set intentionally higher (typically by 20%) than design conditions, implemented to prevent system blockages.

This over-operation, unfortunately, leads to increased power usage and higher particle velocities, consequently elevating particle breakage.

One of the best solutions to address those issues is to develop a real-time control system where a pneumatic conveying system controls its particle velocity at the minimum conveying velocity in real-time.

Such a system typically works independently, around a predefined set point, optimising operation conditions, saving energy and reducing particle breakage. However, integrating digital twinning into such systems enhances their capabilities drastically, showing the industry a future of improved efficiency and sustainability.

Enhancing Control and Predictive Power with Digital Twins

The scope extends beyond a real-time control system after successfully implementing a digital twin in pneumatic conveying. It evolves into a comprehensive solution, comprising a virtual 3D representation of the system, equipped with set of mathematical models describing the system dynamics, and enabling real-time bidirectional communication between the digital and physical counterparts.

This implementation offers features beyond real-time control around a set point. The mathematical models in the Digital Twin can be leveraged for predictive analysis based on the current state and historical data of the pneumatic conveying system.

For instance, Digital Twin can give predictions such as there is a 60%  chance of getting a blockage during a specific period of time for a given set point and material. Moreover, it can take automatic corrective action, such as adjusting the set point to reduce or eliminate blockage risk based on user requirements.

The digital twin is not just a mere control system but a more comprehensive tool providing visual or 3D representations and many other features. If implemented on a cloud or local server with remote access, users or engineers can also monitor and control the system remotely.

Proactive and Reactive Operations with Digital Twins

The integration of digital twins in pneumatic conveying unlocks the potential for proactive and reactive operational measures. Once automated with a digital twin, the system can identify potential blockages, initiate preventive measures, and take reactive steps if needed, such as cutting the blower connection, triggering alarms, or executing additional actions if an explosion risk is identified.

Implementation of a Digital Twin in Pneumatic Conveying

The concept of a digital twin may not be novel, but its implementation in industry is relatively recent. There are no hard and fast rules for its application; it relies on the user's own conceptual frameworks, methods, and tools.

A fundamental framework for digital twin implementation is demonstrated in Figure 2, illustrating the division between the physical space, the digital space, and an optional web space that facilitates remote monitoring and control via 3D graphical representation.

In the presented framework, the pneumatic conveying system remains connected to a local controller for real-time system control and a local database in the physical domain. Given the considerable volume of data generated by hundreds of sensors, a local database is crucial.

Transferring such raw data to a database in Digital Twin in real time is feasible in a standalone local area Digital Twin. However, it is challenging when implementing the Digital Twin on a cloud due to bandwidth constraints.

The solution here lies in deploying an EDGE computing platform for data processing. This allows for raw data storage while transmitting only the essential variables to the digital twin required by the models within the Digital Twin. This streamlined system can then seamlessly communicate with a graphical user interface, offering a dynamic 3D visualisation system.

Expanding Capabilities and Benefits with Digital Twinning in Pneumatic Conveying

Incorporating digital twins into pneumatic conveying reveals many capabilities and benefits. An important ability is a capacity to operate the conveying system at minimum conveying velocity, reducing both power consumption and particle breakage.

Moreover, the lifespan of moving components, such as bearings and rotary valves, can be enhanced significantly. When subjected to accelerated wear, these components require frequent replacements due to operation at higher conveying conditions.

A Digital Twin mitigates unnecessary higher-level operation and integrates the limitations of these moving parts into the control algorithm, thereby preventing overuse and premature wear. Further, Digital Twin provides a comprehensive description of the pneumatic conveying system.

It can visualise data in real-time such as air velocity at the inlet, current throughput, blower status, pressure drop across the pipeline, particle velocity, and properties of the material being conveyed.

Additionally, they offer status updates on components and materials in the feeding and receiving hoppers. Such detailed insights enable informed decision-making, contributing to the overall efficiency and safety of operations.

One of the significant advantages of a digital twin in pneumatic conveying is its predictability, which depends on current and historical data analysis. It can identify potential blockage scenarios and take proactive preventive measures, which is invaluable.

This predictive ability enables Digital Twin to make informed decisions about conveying materials under certain conditions, reducing the need for constant human monitoring and intervention. Moreover, digital twins can predict the completion time for a current batch, which is an essential feature as the solid flow rate of the material is not fixed throughout the process.

Traditional real-time control systems focus primarily on controlling the system around a certain point, often overlooking throughput considerations. However, a digital twin can predict completion times and consider throughput, enabling better planning and workflow coordination.

Digital Twin can calculate the maximum achievable throughput, helping users understand the limits and thus optimise the process. By clarifying these boundaries, digital twins allow users to set realistic expectations for process completion times, facilitating efficient scheduling and overall operations management.

Further, this advanced technology can provide timely recommendations. For instance, based on the material to be conveyed and the status of the components, a digital twin can predict component failure before initiating the conveying process. It could suggest the replacement of components likely to fail in the mid-operation, turning away unexpected disruptions and blockages.

Digital Twin can propose optimal conveying conditions for different materials using historical data and component statuses. According to the requirements set by the user, such as power consumption, maximum throughput, or minimum conveying conditions, the digital twin provides the best setup.

Digital Twin offer the advantage of algorithmic analysis and can even integrate machine learning. They provide insights into the behaviour of different materials under varying conditions, creating a digital memory of all conveyed materials over the years. This vast databank is invaluable as it eliminates the need for memorisation and provides insights as needed.

The 3D visualisation systems facilitate easy understanding and integration of the process by any operator, reducing the need for extensive training and minimising human errors. Moreover, unlike real plants, where internal processes can be invisible, 3D visualisation can render sections transparent, enabling operators to see inside the system and prevent mistakes.

The Digital Twin also can connect with the supply chain, enhancing the predictive maintenance insights. It provides more than a fixed servicing schedule; it calculates service dates based on historical conveying conditions and the availability of the components in the stock.

This information can be integrated into an Enterprise Resource Planning (ERP) system or any other system, allowing the digital twin to suggest optimal times for part replacements and maintenance work based on the operations schedule.

Conclusion

In conclusion, integrating digital twins into pneumatic conveying systems brings numerous advantages. This advanced technology enhances operational efficiency, reduces power consumption and component wear, and offers real-time, comprehensive insights into the conveying system. Its predictive capabilities allow for proactive measures against potential blockages, ensuring smoother operations.

Furthermore, digital twins enable better planning and optimisation of the process, contributing to improved workflow coordination and overall operations management. The ability to predict maintenance schedules and connect with supply chains leads to more efficient, proactive planning, reducing unexpected breakdowns. With the integration of digital twins, pneumatic conveying systems are set for more efficient, reliable, and future-proof operations.