Maintenance tool eases stress during paper mill start-up

When Klabin SA’s greenfield Puma plant started up in March, 2016, it had more than 2,000 field devices with the wrong configuration, 600 loops in manual, 500 bypassed interlocks, an average control loop error of 35%, and more than 550,000 alarms per hour.

Today, a maintenance technician typically goes out in the field and performs either breakdown maintenance, or time-based inspections and preventive maintenance. “Our research shows 63% of trips to the field are for routine inspections (35%) or find no problem (28%),” said Manoj Chandrasekharan, offer director, asset management, process automation, Schneider Electric. “Those are wasted effort.”

Instead, they can do predictive maintenance based on the condition of the asset. “They can monitor assets in real time, know before and act to prevent a failure,” Chandrasekharan said. “Predictive maintenance can be done on any automation asset.”

Chandrasekharan spoke to attendees of his session at Foxboro User Group 2018, this week in San Antonio. “Maintenance might have to deal with 500 to 800 assets per day. How can you know the status of all of them?” he asked. Those assets are made by multiple vendors, and use multiple communication protocols. “You need an application for each type, typically 10 to 12 different applications,” he added. “The life of a maintenance guy is a tough one.

“Mobility and digitalization is a critical need, so the maintenance guy doesn’t have to walk all the way back to his PC for information. We need to minimize manual rounds, and to be able to access information from anywhere.”

A better way to manage field devices

Traditional field device management solutions are narrowly focused on configuration and advanced diagnostics. But optimizing lifecycle performance of field devices to meet business objectives requires a much more holistic approach that goes well beyond tradition. Schneider Electric EcoStruxure Maintenance Advisor is a comprehensive predictive maintenance platform that monitors the condition of plant-wide automation assets; provides real-time reporting of emerging asset problems with context, criticality and corrective action; and has built-in workflow functionality to communicate maintenance actions from anywhere.

Then, maintenance needs a command center. “The operator has a station, accounting has their system, but maintenance has islands of information in work order management, asset information and diagnostics systems,” Chandrasekharan said. “They need their own system, with one source of the truth, a maintenance HMI with maintenance SCADA, to see information, raise work orders, and track them.”

Maintenance Advisor is a single application for a wide range of assets that can monitor transmitters, valves, drives, motor starters and switchgear in real time, automatically, across vendors. It can allow fast decisions by presenting information in the right way. “It can communicate that decision via work orders and assignments to the people who will do the work, and it can do it from wherever you are, you don’t have to go back to the office,” Chandrasekharan said.

An important aspect of Maintenance Advisor is Condition Advisor, which uses embedded diagnostics or programmed algorithms to monitor asset condition in real time; see bad actors; determine the potential cause, severity and impact; and recommend a corrective action.

“Alerts are defined by NAMUR 107, they’re very descriptive,” Chandrasekharan said. Corrective actions can be based on device vendor information or plant procedures. “Work orders can be generated in the CMMS or it can do it on its own,” he added. “Press ‘send’ and it goes to the maintenance planner and is assigned to a technician. When the technician updates its status, it goes back to Advisor.

“You don’t have to wait for everyone to report in the morning meeting, you already know the status.”

Tested in the field

One of the earliest adopters of Maintenance Advisor is Klabin SA, the largest paper producer, exporter and recycler in Brazil. Headquartered in Sao Paulo, it implemented the system on its greenfield Puma unit in Paraná. The Puma plant includes 28 operator and 24 engineering workstations, 32 servers, 41 FCP 280 controllers, and more than 23,000 I/O points, of which 4,254 are HART and 2,936 are Profibus.

When the plant started up in March, 2016, it had more than 2,000 field devices with the wrong configuration, 600 loops in manual, 500 bypassed interlocks, an average control loop error of 35%, and more than 550,000 alarms per hour, with 13 operators.

“We implemented a three-pronged approach,” said Edemilson Bueno de Camargo, process engineer, Klabin SA, who joined the session by videoconference from Brazil. Using an emergency server, they applied EcoStruxure System Advisor to manage alarms and the safety system configuration, and to analyze operator actions. They used Maintenance Advisor to monitor the condition of process instruments, support decisions and support KPIs. And they implemented Control Advisor to monitor, tune and report loop performance.”

The maintenance crew analyzed the alarms and their causes every day. Custom reports were developed to help everyone involved understand the issues, and after a few months, the plant reached 12 alarms per hour per operator.

Since then, Maintenance Advisor has prevented unplanned shutdowns by quickly identifying potential problems and directing maintenance attention. In one example, an air supply regulator on a critical valve in the evaporation plant had been turned off during maintenance. If the valve was actuated before the problem was fixed, the plant would have shut down.

A second example identified a corroded electrical connection in a valve positioner before it failed completely, also preventing a shutdown.

“Our future plans are to update the Control Advisor and Maintenance Advisor software versions to get more capabilities,” said Camargo. “We also plan to add 17 instrument suppliers, provide reports to the management consoles, and add non-DCS instruments to Maintenance Advisor.”