JXTG boosts valve reliability with Connected Services

Paul Studebaker

It’s well established that predictive maintenance is the best way to keep equipment online and minimize breakdown with their accompanying unplanned outages, collateral damage and safety issues. But it’s not always practical to do it yourself.

“We know that a valve alert monitoring program is important to avoid sudden failure by earlier identification of potential failures. Tracking the valve condition over time with predictive analytics lets us take action before it fails,” said Kazuya Koizumi, instrumentation and reliability engineer, JXTG Nippon Oil & Energy Corp. Koizumi presented with Clinton Schneider, director, valve reliability solutions, Emerson, and Sumitaka Ichiki, manager, Plantweb marketing, Emerson Japan, at the 2019 Emerson Global Users Exchange in Nashville, Tennessee.

JXTG operates 13 refineries and chemical plants. Its refinery in the petrochemical complex in Chiba, near Tokyo, is Japan’s largest at 139 KBD crude distillation unit (CDU) capacity, and has been operating since 1968.

In 2008, AMS ValveLink was installed on 144 critical valves. “These are not block or bypass valves, but severe service valves where reliability must be high,” Koizumi said. The valves were diagnosed annually, with Emerson reviewing the performance diagnostic tests. Offline diagnostics were performed during turnarounds.

“Our utilization of valve condition monitoring was limited to annually for several reasons,” Koizumi said. The reasons were lack of internal resources, lack of expertise in valve diagnostics interpretation, and having more than 1,000 nuisance alerts per month. As a result, be said, “Unfortunately, it was impractical to sustain a monitoring program using only JXTG resources.”

The solution is valve condition monitoring through Emerson’s Connected Services offering. Recently added at Chiba, an online ValveLink monitoring system sends valve condition and alert information through the cloud to Emerson, where experts and expert systems perform analyses. Recommended actions are sent back to our site via email, where they are performed by JXTG technicians, if necessary in collaboration with the local Emerson representative.

The Connected Services program was executed in three steps. “First, we met with Emerson to share issues and align roadmaps. Second were the actions needed to optimize the alert threshold values,” Koizumi said. Emerson representatives visited the site and collected data weekly. They performed alert analyses on the five valves with the most alerts, and optimized the valve alert settings.

“Before optimization, we had more than 1,674 valve alerts per month and afterward, 39 alerts per month for a reduction of 98%,” Koizumi said.

The third step was a trial of connected services—a remote proof-of-concept to check the feasibility of the hardware and software by executing the monitoring and reporting process remotely. Doing this required a minor modification of the AMS network by adding a VPN router and setting up an AMS client on the Microsoft Azure cloud environment. Emerson service staff accessed the data remotely to review AMS alerts, and recommended actions from the analyses were reported to JXTG weekly by email.

The connection included a firewall, and Emerson accessed with read-only permission.

“Before optimization, we had more than 1,674 valve alerts per month and afterward, 39 alerts per month for a reduction of 98%.” Kazuya Koizumi of JXTG Nippon Oil & Energy Corp. on the company’s successful use of Emerson Connected Services to remotely monitor valve performance.

One early finding was an air leak on a sequential valve of a hydrogenization unit with frequently-operated valves. The alert was for mass air flow, but valve response was normal. The recommendation was to check for system leakage. “We found a leak before it affected unit operation,” Koizumi said.

A second finding was on a heat exchanger bypass three-way valve. The alert was for travel deviation: the valve was travelling 9% for a full-close signal. The recommendation was to test actual travel in the field and if verified, a valve overhaul. “We verified the problem but determined it was not affecting operations, so we deferred the overhaul to the next outage,” Koizumi said.

Today, by increasing AMS utilization from yearly to weekly with minimum additional workload, JXTG Chiba has achieved $500,000 in loss avoidance by finding problems early, and 98% reduction in nuisance alerts.

“We see an opportunity to improve the results by time-series analysis of valve condition to detect gradual changes—degradation—of valves and make better estimates of when it will noticeably impact control, reliability or safety,” Koizumi said. “We plan to further secure the system with a data diode to physically restrict it to one-way communication, and we would like to expand the monitoring system to include operational-critical valves and machinery vibration.”

1 Reply

  • Control valves is a great example of where outside expertise can be helpful, and when the expert doesn't have to come to site it becomes so much easier. 4-20 mA/HART and fieldbus positioners makes it possible. Vibration is another example of analysis requiring expertise. Wireless vibration transmitters makes it feasible to monitor vibration on smaller equipment like pumps, lesser compressors, cooling towers, air cooled heat exchangers, and fans/blowers etc., not just your largest and most critical turbo machinery. Vibration monitoring alone may not be sufficient so you can deploy additional wireless transmitters for complete condition monitoring of equipment to detect cavitation, overheating, corrosion, leaks, and fluid losses etc. Using IIoT-based connected services with experts in a central location to monitor equipment in a fleet of plants. But even simple steam traps can benefit greatly. Learn how other plants use connected services from this essay: