1 May 2007

Keep it clean

Corrosion monitoring technology leads to real-time results for specialty materials

Fast Forward

  • Corrosion can lead to lost material and process down time, leaving in its wake cleanup and repair costs.
  • Real-time data makes for proactive corrosion control strategies before major problems begin.
  • Real-time monitoring allowed user to troubleshoot corrosion differences in two similar chemical units.
By Keith Briegel and Russell Kane

Process and pipeline corrosion can cause major problems for petroleum, chemical, and process industries, to name a few. Just look at a recent cost-of-corrosion study analyzing all major sectors of the economy, which estimated the direct cost of corrosion annually in the U.S. to be $300 billion. In the global process industries, the cost appears to be about $50 billion per year and is projected to climb still higher over the next five years.  Real-time corrosion data brings the potential to implement more proactive corrosion control strategies be-fore major problems occur.

Rohm and Haas' Deer Park, Tex., facility produces raw materials for adhesives, paints, coatings, and plastics. The company needed to find a way to troubleshoot corrosion differences in two similar chemical units at the site, which showed differing signs of corrosion damage. While one of the units had low corrosion rates, the other corroded at high rates, causing fast failure of stainless-steel piping. Traditional monitoring techniques, such as using corrosion coupons, failed to detect the root cause of the problem. 

Metal coupons are prepared, pre-weighed, and inserted into the pipe where they are left in contact with the material flowing through the line for three to four months. They are then removed from the pipe, cleaned, and re-weighed to determine metal loss. Although this is an effective method for determining corrosion with a given system, it is, at best, historical data and provides an average corrosion rate over the time of the test.

An alternate method is using ultrasonic testing with smart pigs. When using smart pigs the process or transfer line needs to be shutdown, the pipeline drained, and the pig inserted into the pipe where wall thickness readings are taken. Again, this is an effective method to measure the effects of corrosion, but generates data only after the fact and provides an average corrosion rate over time.

Plant officials knew they had a problem but weren't sure what the problem was. Another problem was a plan for alloy upgrade in materials would cost the plant well over $500,000.

Development in the area of corrosion measurement systems, such as smarter data storage and processing capabilities, have made it possible to conduct more sophisticated techniques in remote plant instruments that previously only were available in the corrosion laboratory. Results are greater accuracy to the corrosion rate measurements and the ability to differentiate types of corrosion, such as uniform versus localized attack. It has also resulted in a more rapid measurement cycle, enabling integration of corrosion as a process variable within the plant distributed control system (DCS). With these changes, the corrosion signal can travel directly to the plant DCS for viewing along with process information.

The Rohm and Haas team decided to install corrosion transmitters, providing online, real-time monitoring technology to give plant operators access to time-trended general and pitting corrosion data, which can be correlated with plant process information. Installing corrosion transmitters gave plant operators access to current, actionable process variable information including a time-trended general (uniform) corrosion rate. It also provided an indication of the mode of corrosion (localized or pitting corrosion) detection called a pitting factor.

The transmitters communicate via the HART protocol and can easily connect to existing control systems. As an input to the process control system, you can alarm, historize, trend, and assign corrosion data to process groups. You also now can seamlessly correlate corrosion data with other process variables, providing a broader and real-time view of plant operating conditions and methods of mitigation through process optimization.

Empowering plant operators

After installing the two probes and connecting them to the control system, operators and engineers at Rohm and Haas were able to identify two process scenarios contributing to the difference in corrosion rates. The first scenario was one unit's corrosion rate was immediately higher after a shutdown. On further inspection, they found a leaky valve was allowing water into the system on shutdown, which raised the corrosivity of the system. They replaced the leaky valve, eliminating this source of corrosion.

The second scenario was higher corrosion in the process when a certain recirculation condition occurred. By making some process changes, the plant was able to avoid this condition on the second unit.

Since implementing this technology, the company has reaped the following benefits:

  • Avoided costly materials upgrade by identifying the root cause of the difference between the units, which allowed the plant to save over $500,000 in capital expenditure.
  • Identified high corrosion rates during a recirculation condition by online correlation of process and corrosion data, which allowed the plant to devise an operating strategy that avoids corrosion.
  • Identified a path for operators to use real-time corrosion data in combination with process information to improve equipment reliability, stability, integrity, and uptime.

Creating tighter working relationships

Whereas corrosion is an understood concept to corrosion engineers, process operators and engineers do not typically know if and when it exists. Coupling corrosion data with process data creates a tighter working relationship between corrosion and process engineers. Including corrosion as an online process variable makes plant personnel aware of the process conditions that can initiate localized corrosion. Examples of such conditions include unintentional aeration by venting of equipment to atmosphere, additions of oxidizing agents and aggressive catalysts, lack of dew point control in normally dehydrated systems, and excessively high velocities in attempts to increase unit productivity. Online corrosion detection in a process control environment will allow plant operators to correlate highly corrosive conditions with the process conditions that relate to the corrosion so they can start to reduce the excessively high corrosion cost.

In a modern chemical plant, automation and control systems control the entire facility. These arrangements of process equipment and piping are far too complex for operators to personally control every aspect of their operations. Therefore, they rely on a system of data acquisition and associated computer routines and software applications to analyze the data and apply rule-based methodologies for assessing variations in process conditions and prioritizing the response. These systems also provide management of safety and security. This infrastructure has vastly improved the productivity of the modern industrial plant. With current technology and initiatives such as abnormal situation management, the vision is for a steady increase in the number of operating days per year and increase plant productivity levels to over 95%.

Integrating control strategies is everyone's job

As a 2004 survey shows, corrosion is a major factor accounting for petrochemical plant failures. The survey compared its results with those of a similar survey performed in 1984, and the situation appears unchanged over the past 20 years. To meet this challenge while also reaching productivity goals, plants need to integrate corrosion into real-time automation and control strategies in process plant systems.

Once corrosion becomes a regularly used, online, real-time process variable, you can more fully integrate it into this system. You can more easily acquire measurements and displayed data with other key performance indicators in the plant data historian, which leads to the seamless connectivity between varying job functions. As such, you can more easily share corrosion information across a variety of job functions using a site or enterprise network.

To make gains in productivity, corrosion control must become part of everyone's job function in a similar manner to quality control or safety, and the corrosion specialist can provide key real-time input to significant corrosion situations as they occur. You can automatically generate alerts for specific job functions, such as inspection, maintenance, process control, and engineering. Work orders and response reports related to corrosion can become automated functions as well. Fault models can take the input corrosion data and apply rules to direct the fault indications to the most appropriate plant function, in a time frame where changes can avert major damage.

About the Authors

Keith F. Briegel is manager of the corrosion and materials science group at Rohm and Haas in Deer Park, Tex. Russell D. Kane, Ph.D, is director of corrosion services at Honeywell Process Solutions in Houston.

Pigs not smart enough

News about the oil leak from a pipeline in Prudhoe Bay, Alaska, revealed three miles of the pipeline infested with 5,476 potential bad spots, including 176 places where corrosion might have chewed through 50% or more of the pipe wall. The metal pipe was corroding from the inside out.

A March Reuters report said the operator of the Trans Alaska Pipeline System had successfully completed smart-pig runs conducted ahead of schedule because of growing concerns about pipeline corrosion. But Alyeska Pipeline Service Co. has yet to complete its analysis of the data collected by the runs of the smart pigs.

Real-time monitoring

Neither the metal coupons nor smart pigs can see use as process control. One solution to this problem is real-time corrosion monitoring. Using online, real-time corrosion monitors can help measure general corrosion and localized corrosion, or pitting. General corrosion occurs at nearly the same rate across the surface of the material exposed to the atmosphere. Localized corrosion, pitting, is on the surface of the metal and not uniform across the surface. Pitting corrosion makes up 70-90% of equipment and pipeline failures.

Using linear polarization, harmonic distortion analysis, and electrochemical noise in processing algorithms can help alleviate this problem. Linear polarization resistance involves the measurement of the polarization resistance of the corroding electrode to determine the corrosion current. Harmonic distortion analysis measures the resistance of the corrosive solution by applying a low frequency sine wave to the measurement current. Electrochemical noise evaluates the fluctuation in current and voltage noise generated at the corroding metal/solution interface.

Stand-alone management

Cathodic protection methods have been around for over 150 years.  Whether a galvanic system or an impressed current system is used, someone or something is responsible for monitoring the system's performance.  Most recently, impressed current systems have been the most popular choice along above ground and submerged pipelines.  There are several cathodic protection excitation systems available on the market today, but someone still needs to monitor the health and performance of these 24/7 systems to ensure continuous implementation of proper corrosion protection.

One way to get more from cathodic protection is through a net concentrator system (NCS), a stand-alone management system that can monitor several critical measurements of a cathodic excitation system, as well as external variables like temperature or ancillary 4-20mAs and discrete inputs.  Current and voltage measurements from shunts and step-down transformers are the most common parameters that monitored to ensure the protection system is not only operating, but also operating at the right levels.  The NCS can monitor these parameters and, if desired, compare the parameters to internal set points. Then using an internal real-time control engine, it can provide local control,  alarming, or annunciation.  A standard NCS has an on-board data logging system that can store up to 32,000 time- and date-stamped records users can later download for historical performance details and analysis. 

An NCS has an embedded web server so users can view any real-time signals or historical trends via pre-configured web pages. If users don't have local networks available in remote areas, wireless solutions using frequency hopping spread spectrum (FHSS) in the 900Mhz and 2.4Ghz range can offer up to 20 miles line-of-sight communication capabilities.  Dial-up modems can allow use of existing analog phone lines to periodically monitor the system with a laptop or desktop that supports dial-up connections.

This dispatch was compiled from reports by Keith Briegel, corrosion engineer at the Rohm and Haas facility in Deer Park, Tex., Russell D. Kane, Ph.D, director of corrosion services at Honeywell Process Solutions in Houston, Jack Lehner, business development manager at Pepperl+Fuchs in Twinsburg, Ohio, and Gary Mathur, application engineer at Moore Industries in North Hills, Calif.