Solving Corrosion Problems by Treating Metal Deterioration as a Process Variable

 Imagine that you’re called into your plant to solve a recurring problem. The operator states: “We’ve had to replace this valve three times this year due to failures from over pressure.” Your first question is pretty obvious: “How high is the pressure?” The operator shrugs and responds with a perfectly straight face: “We don’t know. We don’t measure it.”

As absurd as this scenario sounds, it happens every day in facilities where corrosion is a problem. Pipes, vessels, and valves are being eaten out from the inside, but there is no attempt to measure where it’s happening or how quickly. How one refinery solved this problem is the topic of my article at automation.com, Case Study: Refiner Solves Persistent Corrosion Problem Using Online Monitoring. The article recounts problems faced by a Preem refinery in Sweden.

This case study focuses on a specific instance of severe corrosion identified on the top of a main crude distillation column in Preem’s Lysekil refinery. Severe corrosion was identified on the top of the main crude distillation column near a complex arrangement of pressure relief valves. The corrosion was found to be caused by a broken injection quill that was used to inject a neutralizing amine into the pipes. Without the proper injection of the amine, the pipes were susceptible to corrosion, leading to elevated corrosion rates. The team installed a new quill and the corrosion stopped, but after some months, corrosion began to reoccur at a high rate of about 47 mils/year, indicating that the amine inhibitor was not the long-term answer, with further investigation required.

So, how did Preem’s engineers get a handle on the problem and identify the root cause? How did they determine the rate of metal loss? They started treating corrosion as a process variable to be measured and analyzed. Real-time metal thickness measurements are possible and very practical via Emerson’s WirelessHART Corrosion & Erosion Monitoring System, using ultrasonic thickness sensors.

The corrosion monitoring system consisted of 17 sensors, each sending data to a gateway, which was in turn hardwired via Ethernet TCP/IP to a PC hosting the Emerson monitoring software. The Emerson software interprets and converts the ultrasonic waveform raw signals into wall thicknesses measurements using its patented adaptive cross correlation signal processing, which eliminates the effects of roughness caused by some corrosion mechanisms. The software is then connected via open platform communications (OPC) to the AspenTech IP.21 process historian, where the sensor values are interrogated and imported using a unique tag for each sensor.

The article goes into more detail on how the engineering team finally figured out the relevant root cause, made possible by watching the corrosion in action via the monitoring software. As is often the case, it turned out to be a combination of factors that they were able to identify by analyzing the data. Having the data from the sensors made all the difference.

Corrosion mechanisms are often very difficult to detect using manual monitoring methods, increasing the risk of pipe and vessel failure. Online monitoring using wireless clamp-on sensors addresses this issue by continuously providing data to specially designed software, which creates trends and uses algorithms to create alerts in advance of failures. This information is made available to notify plant personnel of impending issues so they can perform mitigation.

For more information, visit Emerson’s Corrosion & Erosion Monitoring pages at Emerson.com. You can also connect and interact with other engineers in the Refining and Petrochemical Processing Groups at the Emerson Exchange 365 community.

The post Solving Corrosion Problems by Treating Metal Deterioration as a Process Variable appeared first on the Emerson Automation Experts blog.