Environmental cracking has been with us for centuries, and it was during the industrial revolution in the 1850s, with the widespread use of steam boilers, that the heavy consequences of failures demanded attention. In a single year in the U.S. over 50,000 deaths and 2,000,000 injuries from steam boiler failures occurred, and this was the motivation behind the slow process of developing pressure vessel and boiler codes – first issued ~60 years later in 1914 and developed by mechanical engineers to address overload and fatigue design.
While helpful, the dominant cause of the failures was stress corrosion cracking (SCC) and corrosion fatigue (CF), terms that were unfamiliar terms to engineers of that era. The greater complexity of environmental effects made it difficult to address, with the Code simply mandating that SCC be avoided. This relegated the issue to the designers, who interpreted the Code to mean that regions of SCC immunity exist, and can be identified with simple tests representing <0.2% of a 40+ year design life. While the Code has been successful in reducing mechanical failures, now, ~165 years later, it has never seriously addressed the original problem of environmental cracking.
Failures thus continue and our grasp of environmental cracking is stunted. Simple, short-term tests are generally only able to identify severe SCC susceptibility, and tend to focus on initiation with its inherent ambiguity, reliance on near-surface conditions that are not well understood, and uncertain relevant to plant components whose surfaces and stresses are poorly characterized. If they were successful, the design codes would have prevented failures, however extensive cracking has occurred, and life evaluation codes were developed to address cracks detected by leakage, inspection, etc.
Resolution of environmental cracking will only come with the recognition that the design codes are a starting point, and other codes, standards and/or guidelines are needed to manage design and fabrication to address environmental effects. Major sources of vulnerabilities include weak material specifications and associated inhomogeneities; welding, fabrication and grinding; and boiling, creviced, aggressive, and upset operating environments – if these were optimized, perhaps 99% of SCC incidents would cease.
Our belated efforts to quantify and predict environmental cracking are complicated by experimental challenges, very long plant life (>80 years), complexities of the metallurgical – environmental – mechanical factors in components, and the rich diversity of the kinetics and possible mechanisms of cracking. Models tend to assume what controls environmental cracking, generally avoiding rigorous evaluation of their assumptions and ramifications. Our basic understanding of environmental cracking – especially immunity – hinges on incisive experiments, including an ability to detect very low crack growth rates.
This talk highlights our dependency on the highest quality experiments to discern the nature and kinetics of environmental cracking. Examining only very aggressive situations that would induce failures in months is an inadequate approach to defining safe operation for many decades, or understanding the nature of environmental cracking. The possibility that immunity or thresholds exists in stress intensity factor, corrosion potential, water purity, temperature, metallurgical state, etc. need to be evaluated against sophisticated experiments able to discern subtleties of behavior. Some examples of such experiments will be given.
The reality is that SCC often occurs over a very broad spectrum of conditions, including very resistant materials, high purity water, low corrosion potential, low stress, etc. Both plant experience and careful laboratory observation have altered our understanding of environmental cracking, and eroded much of the lore surrounding immunity and thresholds. This is especially true in high temperature water environments, the conditions associated with the original industrial problem.