Because most digital-transformation initiatives are brownfield projects, there’s a lot to be said for starting small and then scaling up to a larger implementation, whether it’s regionally or corporate-wide.
Many organizations fail to achieve digital transformation because they invest heavily to make a big bang and then wind up in pilot purgatory, explained Jonas Berge, senior director of applied technology, Emerson. He and Ng Hock Cheong, head of maintenance at Denka’s styrenic resins plant in Seraya, Singapore, explained during a 2019 Emerson Global Users Exchange session how the 23-year-old chemical plant avoided the pitfalls of purgatory by reducing equipment downtime and steam consumption in small pilot projects.
The Denka plant faced many challenges with manual operation and equipment failure. The affected machinery that was identified for the projects included steam traps, cooling towers and centrifugal pumps. And field-personnel safety was also marked for improvement.
Previously, steam traps were monitored in a yearly manual survey using a portable tester. Steam-trap failures went undetected, which affected energy consumption and product heating. This would lead to product-quality issues and ultimately a time-consuming cleanout.
The cooling towers posed a different type of challenge. Plant maintenance technicians were unable to measure vibration on the gearboxes because they resided inside protective shrouds. Surprise failures would occur, and the damage required difficult and expensive repairs, which resulted in lost cooling capacity.
A monthly manual survey of centrifugal pumps was being done using the portable vibration tester, too. Sometimes serious damage would occur, and expensive repairs were needed due to extreme operating conditions.
Finally, a man-down situation at another site had gone undetected, so it was decided to use the Seraya site to pilot a real-time locating system (RTLS) to assess its feasibility, as well.
Phase 1 of the project included a seven-month IIoT steam-trap proof of concept, which began in August 2015. It’s now been in full operation since March 2016. In Phase 2, the plant increased the number of steam traps monitored and then added cooling-tower monitoring and centrifugal-pump monitoring, as well as the commencement of RTLS testing.
“The plant was moving from manual processes to automation-software-based and data-driven ways of working,” explained Berge. “They have an existing DCS, which we had to protect. That was one of the criteria for creating the open architecture. We created and added a digital operational infrastructure (DOI), which is independent of the DCS.” The additional DOI includes pervasive sensing at the base, with pervasive networking. Analytics and reporting are built on top of that.
“It started with the IoT sensors, and the data was aggregated into software,” explained Berge. “Then it takes in data from the DCS. The last step is integration with the ERP. Digitalization was executed in phases, not a big-bang approach.”
“Steam consumption was reduced by 7%, and the assured product heating improved product quality and yield.” Denka’s Ng Hock Cheong explained how starting small can yield big results at the 2019 Emerson Global Users Exchange.
Reducing steam loss is critical in any facility. The Denka plant began by adding two-in-one sensors to 148 steam traps for acoustic and temperature monitoring. Samples were taken every minute. “The analytics and reporting are done across the cloud using industrial protocols,” explained Berge. “This model was very revolutionary in its time.” The AMS 9420 wireless vibration transmitters belong to Emerson, and they are included as part of the subscription-based service. The sensors are WirelessHART. Data goes to a cellular router and then into a Microsoft Azure cloud instance that is managed by Emerson. A periodic report, identifying which steam traps need to be replaced, is generated from the cloud analytics. The project was then expanded to 214 steam traps, covering the majority of steam traps in the plant.
The cooling-tower solution was implemented on the cooling-tower gearboxes. “They were instrumented with vibration sensors to predict bearing issues,” explained Berge. The data could be analyzed in the cloud via IIoT-based connected service or analyzed “on-prem” at the edge using the Plantweb dashboard. “The data drives maintenance activity, such as an oil change,” said Berge. Like the cooling-tower gearboxes, the centrifugal pumps were instrumented with vibration sensors for monitoring, using an hourly sample.
The new man-down solution was targeted at warehouse personnel, who were issued location tags. On-prem software offers man-down alarming, rescue locating, evacuation-mustering headcount and geofencing. With the proof of concept now successful, it is being considered to scale to the main plant eventually.
“Steam consumption was reduced by 7%, and the assured product heating improved product quality and yield,” explained Ng. “The cooling-tower monitoring reduced maintenance cost and downtime. Time between cooling-tower fan overhauls went from three years to five years.”
The same result occurred for the centrifugal pumps, extending time between overhauls from three years to five. And the condition-based monitoring helped to reduce downtime.
The man-down system promises faster detection of safety situations. “We have faster response time for the rescue team,” said Ng.
“Overall, we wanted to reduce manual operation,” explained Ng. “With the digital transformation of these pilots, we were able to improve the plant operation, maintenance and reliability. We will include many more in the future.”