How to Conduct a Corrosion-Related Failure Analysis
Conducting a root cause analysis to identify the exact cause of a corrosion failure - and determine appropriate remedies - requires a detailed and thorough analysis.
Conducting a root cause analysis to identify the exact cause of a corrosion failure and to determine the appropriate remedies to prevent future occurrences requires a thorough and detailed failure analysis. Failure analysis specialists can be involved at any point of a product life cycle, including during the design, manufacturing, service, and ultimately, when a part fails. For the purpose of this article, we will only focus on corrosion or environmentally related failures, which can include stress corrosion cracking, corrosion fatigue, erosion corrosion, pitting corrosion, crevice corrosion, galvanic corrosion, under-deposit corrosion, microbial influenced corrosion, hydrogen embrittlement and filiform corrosion.
The Key Steps in Corrosion Failure Analysis
The steps in a failure analysis often include:
- Information gathering
- Preliminary visual examination and documentation
- Non-destructive testing
- Characterization of material properties through mechanical, chemical and thermal testing
- Selection of samples for subsequent analyses
- Macroscopic examination of fracture surfaces, secondary cracking and surface condition
- Microscopic examination
- Selection, preparation and examination of cross sections
- Identification of failure mechanisms
- Testing to simulate failure
- Data review, formulation of conclusions and reporting
The Information Gathering Stage
For a corrosion-related failure, information gathering is very critical in identifying the service parameters and may include examination of process data and logs as well as operational procedures. Such information is extremely helpful in developing an analysis approach as well as a central part of the root cause determination. Requisite information includes:
- The application, use and function of the failed part
- The specified material and, if it's welded, welded consumables specified along with the welding procedures applied
- Service life history and life of the failed component
- Whether there have been previous failures (including the type and frequency of failures).
- Maintenance and inspection records
- Operational temperature, including temperature excursions and service pressure
- The number of shutdowns, including emergency shutdowns and their durations
- Lay-up procedure (either wet or dry) if specified for outages
- Startup and shutdown procedures
- Service environment, both internal and external
- Design and repair changes
- Neighboring metallurgy
- Design temperature and pressure
For example, temperature excursions can provide insight into high-temperature corrosion, and unusual process conditions may indicate an unexpected presence of a corrodent or a more corrosive operational environment, while the maintenance/inspection log may show rapid metal loss or corrosion rate. Note that design parameters may not reflect operational conditions; any deviations can be a source of failure. Any cleaning procedures to remove hazardous chemicals must be reported as these might have altered corrosion/chemical deposits. Upon shutdown, corrosion scales such as sulfides can be oxidized, which might impact subsequent data interpretation. In addition to start p and shutdown procedures, lay-up procedures are specified with the aim of excluding either moisture (dry lay-up) or oxygen (wet lay-up) for some critical equipment during process outages. Corrosion processes during outages are quite common, especially when chemical conditions are not controlled. This information is the starting point for a root cause analysis.
The visual examination and documentation of the failed component is the cornerstone of the failure analysis since the failure can be related to the overall component design and operation. For field failures, the component can be quite large and photographic documentation can be quite useful to illustrate and document observations; this is particularly the case when the field investigator is not the analyst in the lab. The field investigation may include a selection of samples to be forwarded to the lab and these locations must also be documented.
If the failure is not visually obvious, non-destructive evaluation may be performed to identify location(s) of leaking. The particular non-destructive method selected mainly depends on the type of flaw to be found (a crack, a pit, a blister, etc.). Often dye penetrants are used, but this procedure may leave residue on corrosion products, which might complicate interpretation during the subsequent analysis.
Sample selection for analysis is often constrained. Samples may be sent to multiple labs or only a small sample can be removed because of repair requirements. Scale samples must be removed in such a manner as to prevent contamination, their locations documented, and the sample from each location stored in a new and different polyethylene bottle. Removal of the metallurgical samples should be done to minimize elevated temperature exposure or damage to the sample. Where possible, several inches to a foot should separate the failure from the cut metal. High temperatures from cutting can modify the metallurgical microstructure (i.e., sensitization of stainless steel) and impact corrosion products. A sample far from the failure location may be useful for comparative purposes and should be included in the documentation process. Analysis of the process solution may also be useful for determining the potential of unexpected corrodents. Again, glass bottles should be avoided as they often break during shipment. For gases, silica coated metal cylinders are useful.
Examining the Data
Once all of the samples are received in the laboratory and the documentation has been reviewed and discussed with the field investigator, detailed examination under optical microscopy will provide an initial understanding of the failure. One should start with the unaided eye and advance to a stereoscopic optical microscope. Photographic documentation of pits, cracks and the corrosion product location(s) and fracture surface will assist in selecting samples for subsequent analysis as well as for further understanding the cause of the problem. For example, pitting along the liquid/vapor interface of a condenser tube can be indicative of dew point corrosion and is best documented with a digital camera. When corrosion scales vary in color, documenting the locations of samples for analysis can be very useful in determining the corrosion progression.
Analysis of selected corrosion products is very informative of the corrosion species. The most common analytical tools are the energy dispersive x-ray spectroscopy (EDS) associated with a scanning electron microscope (SEM) or by x-ray diffraction (XRD). With EDS, the elements associated with the corrosion products are identified; with XRD, the crystalline species are determined. Thus, for example, by XRD, Fe2O3 can be distinguished from Fe3O4 and the process conditions to produce that species defined.
Metallographic analysis can be very useful for identifying stress corrosion cracking, corrosion fatigue or microbial induced corrosion. Polishing a pit in cross-section and analyzing the corrosion products at the base of the pit can provide insight into the cause of the pitting. Such sample preparation should be done dry to avoid water solubilizing the scale. Care in selecting the mounting media is critical so as to minimize edge rounding during polishing and elemental interference. For example, most epoxies contain chlorine and thus should be avoided if chloride corrosion is suspected.
Fractography, arguably the most valuable step in the failure analysis of a fractured component, has been used for centuries in the field of metallurgy. After examination of the fracture surface by optical microscopy, the next step is to examine selected locations in the SEM. Corrosion products on the fracture surface can be identified with the EDS. At the crack tip, the fracture morphology can be identified as either transgranular or intergranular, or the presence of fatigue striations can be determined. Care must be taken in cleaning a fracture surface to prevent any damage or when opening secondary cracks.
Finding the Root Cause of Corrosion
Once the failure mode has been established and the report written, the next step is to use this information, along with the documentation of the operational conditions, to conduct the root cause analysis and to determine how to prevent such future occurrences. The root cause analysis team should cross functional boundaries in the organization. The analysis is designed to identify the inter-relating causes of the failure and to recommend appropriate corrective action(s) to prevent future occurrences. This includes fully understanding how to fix the problem and potential underlying issues. Prevention might include modifying designs or the core process operation and not just changing metallurgy. Follow-up evaluations are a critical step to determine whether the corrective action(s) actually solved the problem or created additional issues.