Remote and condition monitoring in chemical plants | Automation.com

Remote and condition monitoring in chemical plants

Author: Gianni Minetti, President and CEO,
Paradox Engineering SA

Chemical facilities today face very complex challenges, since they are asked to maximize plant production while ensuring efficiency and cost control. Effective operations management is fundamental and generally deals with monitoring, troubleshooting and problem mitigation – all these processes rely on data collection, information reporting, virtual or factual inspections in the plant to check key parameters and indicators.

Maintenance processes include not only break fix or troubleshooting, calibration, or testing: therefore, not only do they need quick and effective access to context information, but also they have to elaborate a large amount of data generated in remote if not hazardous environments.

In addition, chemical plants struggle with growing complexity as for systems and sub-systems, and are often instrumented for process controlling.

In short, there is a huge volume of data and information to be effectively collected and treated to ensure proactive management, best-in-class maintenance, reliable and efficient operations.

Furthermore, market asks for ever growing performance, and this pushes towards more frequent collections of data which are difficult to obtain from traditional sources. New data points are often required by new regulations (eg. about emissions) or the changing production patterns resulting in growing risks calling for monitoring and control (eg. rotating machines designed to work continuously but forced to new standby/on duty cycles by floating market demad). On the other hand, the high cost of installing field instrumentation often prevents the placement of additional data points which would make plant assets monitoring and management get to the next level. A typical chemical manufacturing facility requires hundreds of thousands of metres of wires and cables for reliable monitoring operations, with costs which are calculated to vary from 120 to 6,000 USD per metre. The actual cable cost depends on the location of the machine relative to existing network infrastructure, the type of cable needed (e.g. fiber optic), conduit engineering (if needed), labor cost rates and trenching if required. Recently, a waste to energy plant calculated thateach new data point delivered through traditional cables/wires would cost roughly 5,000 USD. Such costs clearly represent an investment not only preventing from installing additional data points or sensors, but also pushing companies not to move from wired to wireless sensors before being sure the current wired infrastructure has been extensively amortized and absorbed.

It is evident in fact that even if installation, calibration and commission via wireless technology is relatively inexpensive, and wireless sensors are quite low cost devices, because the investment for wired devices has been so huge most of the companies struggle to deliver more value with that same investment being therefore renitent to substituting wide parts of the infrastructure.

Still, benefits of wireless transmissions are clear and include improved performance (in terms of speed of deployment, for instance), reduced costs for installation and ongoing maintenance, and better overall control thanks to deeper, more granular and timely collection, rapid deployment and virtual ubiquity of wireless data acquisition and transmission.

It looks therefore logical that wireless sensors themselves might not always be a viable option, while wireless transmission seems to be the best choice. So how to conjugate these two elements?

Feeding data into the company’s ICT infrastructure
The answer lies in wireless devices connecting existing data points to the company’s ICT architecture: collecting information from data points, they can reliably face the needs of increasingly frequent sampling, ensuring first class maintenance, operations and asset management without requiring significant investments to the companies. Requiring no change to existing sensor infrastructure, such sort of solutions may also be used for temporary and/or incremental data collection campaigns that are often needed and not performed, once again for prohibitive costs.

Common industrial wireless sensor network solutions actually use RF transmission on 2.4 GHz or above, which is unluckily critical in industrial environments, quite congested, and offering a low penetration factor. We have to consider that industrial environments require by nature high penetration and noise immunity, low power consumption even with high sampling rates and challenging transmission times.

In addition, to make operations smoother, installation and use is expected to be easy and provide fitness-for-use: with shrinking workforce, no long implementation or roll out processes, trainings or additional complexities can be afforded – not to tell about budget constraints preventing from subcontracting to skilled third party engineers. Therefore, an industrial wireless sensor network solution should better require no software to install, no heavy server side operation. In brief, it should provide almost unmanned deployment coupled with modularity, scalability and a self-organizing, self-healing, adaptive and fault tolerant network to ensure best wireless and network performance and  >99% data reliability. Here, a full mesh network is the best answer.

Authentication, authorization, strong encryption, open standards and interoperability should complete the set of characteristics, together with hazard/explosion safe hardware and web-based management applications for configuration and management.

Such a set of features in a solution requiring no change or upgrade to any existing infrastructure, sensor, or data point should quite clearly enable any chemical production facility – and yet any industrial plant – to meet their objectives as of increasing granularity of information, better asset management and improved operations, at a fraction of the cost of wired solutions, which are often economically inadvisable due to plant location being remote or critical.

Production plants might in fact easily be located in rural or geographically outlying locations, but also in urban centers implying limitations that make standard wired technologies economically unfeasible, like for example right-of-way access or line-of-sight. As a consequence, choosing a wireless sensor network solution being ready for narrow band and scalable to broad band is an attractive alternative ensuring highly reliable connectivity, also suitable for day-to-day operations or as a backup to terrestrial systems during disaster recovery situations. Because of the bandwidth economies of scale inherent an existing provider’s large network, working through a satellite network provider – and better through a virtual network operator offering tailored hybrid solutions – is an interesting option to consider, which may well also ensure business continuity applications and redundant communications to backup terrestrial ones. Once again, the reliability a modern plant – and business – requires a first class infrastructure for data collection, transmission and delivery.

The chemical industry represents an example of very high complexity in real-time production and condition monitoring, where wireless sensor network solutions can be particularly successful in providing an effective approach to data collection and transport for overall plant efficiency. Thanks to innovative wireless sensor network technologies chemical companies have the opportunity to leverage all the value of information coming from their assets and production workflow — but lessons from this market are useful and applicable for all industrial organizations having dispersed plants and facilities to be monitored.

About the Author
Gianni Minetti is President and CEO of Paradox Engineering SA, and founded the company in 2005.
Paradox Engineering SA is a highly dynamic Swiss company whose portfolio includes Industrial wireless sensor networks, smart metering solutions, global virtual networks for companies where smart grid, energy management, condition and remote monitoring, industrial process monitoring, engineering and telemetry projects require future proof and cost effective solutions to unlock the value of data. Paradox Engineering SA acts as a one stop shop provider of turnkey solutions and technologies on a truly global scale through a consolidated network of strategic partners
 

Did you Enjoy this Article?

Check out our free e-newsletters
to read more great articles.

Subscribe Now

MORE ARTICLES

VIEW ALL

RELATED