Spotlight: How to... use intelligent metal tubes in industrial processes
Erika Hedblom, Manager of Intelligent Tube Systems for Sandvik, Sweden, discusses how internal sensors and cloud computing can provide deeper process insights.
In the age of smartphone technology, we’ve come to expect a high level of real-time digital insight as standard. From the exact location of our approaching taxi to the latest traffic situation on our route home, data is giving us more control in our everyday lives. The same principle applies in industry, where an ever-increasing drive for efficiency and process optimisation is pushing up demand for accurate information, which is being delivered by new and exciting technologies.
Stainless steel tubing is ubiquitous in a variety of sectors, ranging from chemical processing and industrial heating to power generation and petrochemical plants. It plays host to crucial processes and so requires a closely controlled environment to ensure safety and efficiency are maintained. Until now, technicians wishing to monitor the inside of tubes have had to rely in large part on intuition, listening and guessing. But the advent of intelligent tubes now heralds a major change.
What happens inside the tube?
Conventional external sensors can provide some information but are unable to give insights into the internal environment of tubes, pipes and connectors, not least because the often harsh environments result in failure of the sensors themselves. The stresses experienced by industrial tubing include temperature fluctuations, physical strain and vibrations. These pressures are particularly pronounced in extreme or inaccessible environments such as petrochemical applications and boiler systems. Over time, damage and eventual failure result, leading to costly downtime, as the offending section of tubing has to be identified, located, and repaired or replaced.
But what if technicians could monitor and assess the state of the processes – and the tube material itself – in real-time? That opens up a world of opportunity to proactively identify problem areas and address them before they become an issue, protecting the material from exposure to conditions beyond its design limit, and also maintaining the optimal environment for the process at hand.
With this vision in mind, Sandvik Materials Technology developed the Sentusys intelligent tubing system, which uses internally embedded sensors in combination with cloud computing to provide detailed real-time insights.
How does intelligent tube monitoring work?
Highly sensitive sensors are integrated into the wall of the metal tube itself, where they are protected from the harsh conditions while being near enough to the action to provide an accurate picture of conditions within the cavity.
The Sentusys system consists of four key components that work together to collect, process, store and analyse data continuously:
- Tubes with integrated sensors and cables
- Signal conditioning equipment
- Cloud-based data storage and calculations
- Data monitoring and analysis.
Embedded as they are within the metal wall of the stainless steel tubing, these sensors are shielded from environmental conditions, but well-placed to detect and monitor the processes taking place inside the tube. Even in extreme situations, such as harsh chemical reactions or sites of high thermal transfer, these sensors collect data on factors including temperature fluctuations, physical strain on materials and vibration.
The signal process and connectivity device receives signals from the sensor tubes, with one or more devices required depending on how many tubes are used. It uses 4G technology and local output for real-time production process control. The data is also uploaded from the sensors to a cloud-based data centre, built on Microsoft Azure, for additional analysis and storage.
Technicians are provided with an up-to-the-minute view of conditions, as well as being able to look back and review trends and changes. Based on the information, necessary changes can be implemented, for example adjustments in temperature or, if necessary, tube repair or replacement. A new level of accuracy and reliability in material condition assessments is now available to industry, with material failure analysis available for troubleshooting.
Challenges that had to be overcome
A number of obstacles had to be overcome in order to make the system work. Robustness is essential – the sensors had to be strong enough to cope with all kinds of harsh chemical, thermal and physical conditions in a range of industrial applications. So they are engineered to be fully load carrying, while at the same time not affecting the material certification of the metal in which they are embedded.
With regard to the material of the tube itself, a range of different stainless steel alloys are deployed, depending on the environment. For example, highly corrosive conditions, such as in marine drilling, require the use of duplex or hyper-duplex stainless steel grades. The versatility of the system allows for intelligent tube to be made from most stainless steel alloys and in a range of sizes from 15–110mm in diameter, and up to 15m long. The main material accommodation that must be made is the addition of 2mm to the thickness of the tube wall, in order to contain the sensors.
How technicians can use intelligent tubes
One facility using this technology is Bomhus Energi AB in Sweden, where it has been deployed in a 150MW bubbling fluid bed boiler. With a temperature of less than 250ºC, low pressures, air inside the tube and fumes outside, the conditions are ripe for the formation of dew, threatening corrosive damage to the pipes. Having replaced 1m of conventional piping with intelligent tube, technicians are now able to monitor temperatures and keep them above dewpoint, staving off corrosion and the resultant cost in downtime and replacement.
What next for intelligent tube?
There is little doubt that collection, processing and analysis of real-time data will develop even greater sophistication and accuracy. Sandvik is exploring the potential applications of the stored data for machine learning, data mining and artificial intelligence to help further optimise processes and drive even greater efficiency.