1
Localized Corrosion in Complex Engineering Environments

Although substantial progresses made over the past century in corrosion science and engineering have significantly improved the understanding and control of materials corrosion and degradation of engineering structures, localized corrosion remains a tenacious structural health concern for a wide range of industrial and civil infrastructures such as oil and gas pipelines, piers and ports, water, wastewater and sewer systems, renewal energy production, and transportation facilities. In particular, localized corrosion in complex and variable industrial environments remains a persistent issue that has not been sufficiently managed and controlled. This chapter discusses the characteristics and sources of localized corrosion complexity in practical engineering conditions through the analysis of corrosion cases, from simple to more complex forms. It is shown that localized corrosion complexity is often associated with dynamic changes in corrosion mechanisms and kinetics over an extended period of exposure to variable environments. Localized corrosion complexity makes the prediction, prevention, and management of industrial structural failure a significant challenge.

1.1 Localized Corrosion Complexity

1.1.1 Localized Corrosion and Its Complexity in Engineering Systems

Localized corrosion is usually referred to as corrosion in which there is an intense attack at local sites on a metal surface. A typical example of localized corrosion is pitting corrosion that rapidly attacks a tiny area of an engineering structure such as an oil and gas pipeline, leading to rapid penetration and failure of the whole structure. Localized corrosion is often linked to the local breakdown of a passive or protective film on the surface of a metallic structure, causing accelerated dissolution at localized sites on the structure [1]. In theory, localized corrosion shares the same electrochemical principles that are developed to explain uniform or general corrosion by pioneers of corrosion science, among them Evans [2], Fontana [3], Pourbaix [4], and Tomashov [5]. However, localized corrosion has some characteristics that make it significantly more complex to understand and much more difficult to control than uniform or general corrosion. Over the past decades extensive research has been carried out to understand and model localized corrosion, in particular pitting, by many corrosion scientists and engineers. Among them, Frankel [6] provided an overview of pitting processes including the breakdown of passive films, metastable pitting and pit growth, as well as critical factors influencing pitting corrosion such as alloy composition, environment, potential, and temperature. Macdonald [7] presented the point defect model to explain the growth and breakdown of passive films on metal and alloy surfaces in contact with aqueous solutions, and for the development of a deterministic method for predicting localized corrosion damage. Marcus et al. [8] considered diverse mechanisms of passive film breakdown at the oxide grain boundaries. Soltis [9] highlighted that there is a clear separation of the passivity breakdown/pit initiation process from the pit propagation, which can be considered in terms of the well-known pitting localized acidification model [3, 10]. Newman [11] reviewed the use of statistical methods and semi-empirical models, and the fundamental deterministic processes that occur during localized corrosion.

In numerous literatures, pitting of stainless steels is frequently taken as a typical example to describe the characteristics of localized corrosion, probably because of their wide and countless engineering applications. Susceptibility to pitting is well known to be a major weakness of passive alloys including stainless steels and aluminum alloys when they are exposed to some environmental conditions. In order to explain and predict pitting corrosion of stainless steels, many contested pitting models have been proposed over the past decades [6-11], although currently there are still diverse views on pitting initiation and propagation processes. Nevertheless there are some undisputed general observations regarding pitting corrosion characteristics that are often also applicable to other forms of localized corrosion:

• The initiation of pitting corrosion on stainless steels involves a very small pit nucleus that grows over periods of the order of seconds. The cause of the initiation of pitting corrosion is still not entirely clear; however, it is clear that manganese sulfide inclusions play a critical role for stainless steel type 316 [12]. The initiation of pitting is often described as a random or stochastic process with respect to time and location. However, this author has a view that pitting of a specific metal in a particular environment is not an accident, it is a deterministic event determined by the thermodynamic instability of the metal in the environment, regardless of the size and the shape of a metal specimen or an engineering structure [13]. This explains the fact that a small corrosion probe can simulate and detect pitting on a much larger structure surface exposed to the same environment.
• Pits become more stable as they become larger and a local acidic environment is developed in the local cavity area. This is because for very small pits the local acidity in small cavity can be easily neutralized by diffusion into the bulk solution. As the pits grow larger the diffusion distances increase and the cavity become isolated, and it gets harder for the acidity in pits to change through diffusing away. Therefore when pits are established, they are difficult to be stopped; however, solution movements and the introduction of inhibitors could have an effect on this. This fact also suggests that pitting growth would be a three-dimensional event, three-dimensional corrosion models and probes could be necessarily for modelling, simulating and probing such three-dimensional corrosion behavior.
• It is generally believed that pitting requires a passive external surface that can provide a high potential to cause the current to flow into the pit. If the external surface is active, this driving force is not available and therefore pitting would not grow. For this reason, passive metals such as stainless steels are susceptible to pitting because of the passive film of the external surface. Metals such as carbon steel only pit if the solution (e.g. alkaline solutions) passivates it, it would not pit if the solution (e.g. neutral salt solutions) only corrode it generally. Therefore variability in environment (e.g. pH changes) could change carbon steel behavior completely, creating a major uncertainty in corrosion behavior. This characteristic suggests that allowing the formation of local passive and active environments is important for localized corrosion probes. However, it should be noted that if there is an external current source, such as a piece of coupled noble metal, that can behave as cathode to cause the current to flow into the pit, a passive external surface might not be necessary for pitting to occur.
• There is a critical pitting temperature that is considered to provide a robust predictor of the pitting potential [11]. Below the pitting temperature, metal would be unable to maintain active pitting corrosion dissolution at anodes because passivation intervenes, even in the most aggressive possible microenvironment. This characteristic suggests the importance of environmental condition simulation, especially temperature, in localized corrosion testing and monitoring.

Awareness of these characteristics is important not only for predicting and managing pitting but also for understanding many other forms of localized corrosion that have somewhat similar processes and mechanisms as pitting. For instance, crevice corrosion is known to have a similar mechanism as pitting, although it is easier to initiate than pitting because of the pre-existing local environment at the crevice [3, 10]. More extensive discussion on pitting and crevice corrosion characteristics and mechanisms can be found in prime corrosion science and engineering textbooks such as those in references [2, 3, 10]. It should be noted however that although extensive knowledge about localized corrosion has been acquired over the past decades, there is a still major knowledge gap in the field: there is insufficient knowledge of complex localized corrosion in variable engineering environments. Theories and methods discussed in the historical literature are generally limited to specific forms of localized corrosion (e.g. pitting of a stainless steel) in a defined corrosion environment (e.g. a sodium chloride solution). This is because localized corrosion knowledge reported in the historical literature is often developed based on observations and data from simplified and accelerated laboratory experiments where corrosion occurs under controlled environmental and electrochemical polarization conditions. Observations of localized corrosion are usually in relatively small dimensional and short time scales using a range of conventional visual, physical, and electrochemical techniques. Electrochemical methods used for localized corrosion studies are generally designed under steady-state condition hypothesis, which are often ineffective, if not incapable, for probing dynamic and localized interplay between corrosion mechanisms and changing environmental conditions. For instance, most of...