Local and general corrosion of well pipe can be monitored in real time over the Internet.

Internal corrosion continues to be a major problem in oil and gas surface facilities and pipelines. Nearly half of all production pipeline leaks come from internal corrosion, compared to 15% from external corrosion, 10% from third-party damage and 25% from all other combined causes. Even more importantly, data indicate localized corrosion causes 75% to 90% of all corrosion failures in industrial field and plant applications. This underscores the need for corrosion assessment.
Increasingly, operators are using direct assessment methodologies to reduce leaks and catastrophic failures. These methods provide a direct indication of system health and integrity. If operators can demonstrate that they have an organized monitoring program, they might not have to perform more frequent, costly and invasive monitoring techniques. An ongoing corrosion monitoring and management program is one way of reducing overall risk while lowering surveillance costs.
However, for a corrosion-monitoring program to be effective, it needs to look for corrosion in the right places. And it must be able to distinguish different types of corrosion phenomena. For example, corrosion monitoring techniques should differentiate between general and localized (pitting) corrosion. Pitting results from loss of protection on the metal's surface and the development of local anodes and cathodes that drive the corrosion process.
One problem with conventional techniques is they only measure the current associated with the overall corrosion process. Previously, direct examination of corrosion specimens was necessary to obtain information about pitting corrosion tendencies. However, newer techniques look at the local fluctuations in corrosion signals (electrochemical noise) in addition to the general corrosion current (Figure 1). These methods provide a quick analysis of localized tendencies before they result in general thinning or highly destructive pits.
This means monitoring points need to be carefully selected, not merely in convenient locations. Most leaks from internal corrosion are caused when a corrosive fluid is trapped in a particular location and left undetected for a prolonged period. Commonly, these are locations where water can pool, such as regions of low flow (less than 3 m/s), low spots in a pipeline or piping system and poorly draining areas. These also may include areas immediately adjacent to a large rise in elevation, where water and debris can pool and be forced to continually slug through the system. Here, combined water and turbulence can result in erosion corrosion.
High levels of carbon dioxide, hydrogen sulfide and chloride in the production environment also can contribute to corrosion problems. Local problems can develop from a leakage of air into the system from pumps, when chemicals are injected or where the presence and growth of bacteria results in corrosive conditions.
Locations
It is important that corrosion monitoring points be carefully considered in advance. Locations should be based on specific system configurations, such as low spots, local turbulence, injection points and stagnant areas, as well as environmental severity and potential for upsets. Establishing multiple monitoring points is critical so that system upsets can be located and identified.
Monitoring can be performed mistakenly in a vertical pipe section rather than in a horizontal section where water and debris would tend to pool. It may be made on a platform export line rather than in a seafloor pipeline. And it may be where a technician can access the pipeline to remove corrosion specimens instead of in a turbulent confluence of two flow streams.
One of the major limitations of corrosion monitoring is that it only has been viewed historically. That is, the data may have been taken with corrosion specimens exposed to the environment for a relatively long period (30 to 90 days). In that case, the weight loss is tabulated and sent to a corrosion engineer who tells the operator there was a corrosion problem sometime in the past. Even when corrosion monitoring is conducted online and in real time, the data typically are sent to a corrosion engineer who may not have access to the production data. Such correlations could be made, but then the operator receives historical data after the incident is past.
Historically, corrosion equipment has been associated with large boxes of computer equipment that cannot be easily interfaced with process control systems. Additionally, corrosion monitoring systems have not used the same communication protocols or accessed the same network as mainstream process monitoring equipment that goes back to the control room. However, modern instrumentation is much smaller and uses distributed hardware at the monitoring point. This instrumentation also can send signals to the control room directly through the plant data loop (via wireless communications or other SCADA systems) and does not require a separate host computer.
Using data to confirm integrity
Corrosion monitoring data need to be online in real time to provide the greatest value in assessing operating systems and prescribing preventative actions. Additionally, the corrosion data should be directed to those people who have control of the system on a real-time basis and be communicated in a simple form. This way, operators can monitor corrosion as they do temperature, pressure, flow rate and pH.
Increasingly, operators are becoming involved in the corrosion control process. Because they routinely look for upset conditions, it is a natural extension of their process control function to relate process and corrosion data. This allows instant correlation of increasing corrosion rate or pitting tendencies to changes in production conditions on a real-time basis. With the addition of a computer network, operators can actively share data with corrosion or process specialists to determine relevant trends while situations are current instead of after the fact.
Varying corrosion inhibitor dosage
Figure 2 shows the rapid process control feedback provided by online monitoring. During a corrosion inhibitor pump upgrade, the monitoring data immediately showed the effect of varying the corrosion inhibitor dosage rate. Prior to the upgrade, inhibitor was being injected at up to 40 ppm, and the average corrosivity value was about 0.35 mm/year. After the pump upgrade, a higher dosage of corrosion inhibitor (70 ppm) was being injected, and the fluid corrosivity in the main oil line dropped immediately to an average value of 0.15 mm/year. When the dosage rate was decreased to 45 ppm to 46 ppm, the corrosivity rapidly increased to an average of 0.25 mm/year.
The effectiveness of the pump upgrade can be seen clearly. The corrosion probe later was used to demonstrate the effectiveness of new inhibitor treatments as they were introduced into operation.
Figure 3 shows that corrosion monitoring can identify periods of localized corrosion using a localization (pitting) index. This method statistically analyzes the scatter of the current response compared to the value of the corrosion current on a time basis. Episodes of high pitting (as observed when the pipeline corrosion inhibitor dosing was stopped) quickly resulted in high values of localization index. Another peak in the localization index was observed after the inhibitor was restarted but before a uniform inhibitor film was established. Similar observations were made when the inhibitor concentration was too low or where the inhibitor formulation did not provide high resistance to pitting corrosion.
Condensate fluid corrosivity
Figure 4 depicts the response of online corrosion monitoring in a condensate line to an upset condition. The larger of the two spikes (labeled A) in the data is associated with an unforeseen gas plant upset. The background corrosion rate value falls well within the prescribed limits of the corrosion management program.
In this case, the value of the monitoring provides confidence that the condensate line is under excellent corrosion control - most of the time. However, it also showed that corrosivity spikes needed added attention to minimize future corrosion problems.
The online aspect of the monitoring means operators are alerted to the effect of plant excursions on corrosion activity in real time. Moreover, operators can implement remedial measures immediately to manage the corrosion.
Water injection header fluid corrosivity
Figure 5 shows the sensitivity of corrosion monitoring to bacterial activity in a water injection system. Corrosion monitoring and corrosion specimens were installed in a water injection header.
In the graph, the corrosion rate data show "noisy" characteristics typical of localized corrosion activity caused by scaling of the material surface. This scale buildup is characteristic of microbial activity, which later was confirmed by microbiological analysis of the scale formed on adjacent corrosion specimens. This real-time information alerted the operator to the problem (even before microbiological work could be scheduled and performed) and gave an indication of the level of corrosion damage sustained before pipes and vessels were cleaned.
Following replacement by a clean probe, the corrosion data showed a dramatic decline in corrosion activity to a level representative of the corrosion control afforded by corrosion inhibitor and biocide chemical treatments. The information available from corrosion monitoring of clean and scaled surfaces enabled the platform integrity team to manage scale deposits at acceptable levels. Probes at this location were cleaned and replaced every 6 months on a rolling basis with the data compared before and after probe-cleaning episodes.
The technology
The technology that made these corrosion monitoring jobs possible came together when InterCorr acquired the SmartCet product line, which uses electrochemical noise technology for assessing general and localized corrosion online in real time.
SmartCet technology is the only monitoring method capable of providing rapid indicators of localized corrosion, sending the data to clients for analysis, interpretation and feedback via modem, wireless or Internet connections. SmartCet is used in combination with corrosion prediction software such as Predict and Socrates to solve field problems, reduce failures and identify locations for cost-efficient corrosion monitoring.
For more information, visit www.corrosionsource.com.

Comments