Holistic look at SCC phenomena

This is the second trip to ‘our wonderful world of corrosion’. It deals with the many phenomena of environmentally assisted cracking of metals. There is a plenty of literature on their particular embodiments. But what is rather missing – understanding their common grounds and drivers. Did you ever ask yourself: ‘What is the Stress Corrosion Cracking (SCC) or ‘hydrogen embrittlement’ in fact? How does it actually work and how can we prevent or diagnose it? Continue reading for a holistic view.

What is common in all corrosion processes?

I hope that the reader will excuse all the simplifications introduced here for brevity and clarity. There is a plenty of technical literature offering more details and rigours to check these ideas.

The common mechanism of all corrosion kinds is a process of ion exchange described using chemical reactions. In these reactions, ions re-combine to produce new substances. For example, the reaction of salt dissociation in water proceeds both ways, depending on its thermodynamics:

What is important here – the exchange process requires and produces free ions, like H+ and Cl-.

A chemical reaction which produces free electrons e- (and triggers an electric current in an electrolyte) is called electro-chemical. For example, three reactions of iron corrosion in water are:

Iron oxidizes, and hydrogen reduces. If the electrons flow away, what unbound substance remains here? H+ hydrogen ions, which are in fact protons – the smallest nuclei ever, able to penetrate any matter!  Yes, they can combine into the hydrogen gas H2, but what happens if their electrons gone? The protons’ penetration into the metal!

Now, to finish up with chemistry, let’s look how metals react with acids:

Thus, the common mechanism of any corrosion is an ion exchange reaction, which can have different embodiments in various environments. More importantly, the hydrogen ion production accompanies all these reactions, in various extents according to their kinetics. And these H+ ions might be those responsible for SCC, and also, for many so called ‘embrittlement’ phenomena, as we will see shortly.

Classification of all damage mechanisms

Now we should overview the whole spectrum of operational damage mechanisms, taking an aerial snapshot (of an integrity engineer) rather than a magnification glass (of a corrosion engineer).

Namely, most literature specialized in operational damage mechanisms focuses on particular pairs metal-environment, and produces a long list of the potential diseases, though with a little common roof visible over them. Specifically, the spectrum of operational damage possibilities is split into many examples of corrosion and cracking viewed as independent particular phenomena. I was interested in finding more commonalities and eventually realized this scheme:

This diagram is not exhaustive. For example, I’m still unsure where the microbial corrosion goes. Should that be separate mechanism or and aggravator of electro-chemical corrosion? What are your thoughts?

The ‘embrittlement’ phenomena

It is also remarkable: Why are there so many ‘embrittlements’ in different blocks of this diagram?

Namely, the word ‘embrittlement’ is not specific enough. It can relate to any process resulting in material rupture, and thereby confuse several mechanisms in one term, like these:

• High Temperature Hydrogen Attack (HTHA) of steels is a chemical combination of hydrogen ions H+ with iron carbides FeC. This combination produces methane molecules CH4, which are far bigger than H+, and tear the material apart from inside. As a result, additional residual stresses accumulate and promote crack growth to end up with ruptures under external stresses. One may call this effect as ‘embrittlement’ of the material, though with no clear references to the actual mechanism. The name HTHA is far more specific and explanatory.

Though, as we seen above, a high temperature may not be a pre-requisite. The H+ ions are produced at normal temperatures as well (eq. 1 to 4), and they can exploit the same idea: accumulate in material internal voids and discontinuities (like micro-cracks) and combine back into H2 molecules there, by means of borrowing free electrons from the metal. Similarly, H2 molecules are much (approx. 100 000 times) bigger than protons, and the internal pressure tears the material apart, thus propagates micro-defects in a surprisingly successful way. Voila, this is the explanation of all SCCs and many ‘embrittlements’. For example:  

• ‘Hydrogen embrittlement’  more specifically called Wet H2S cracking, proceeds according to the very eq. 4 mechanism (with some additional equations written for Nickel of stainless steels). Furthermore, NACE and API guidelines subdivide it into a variety of particular embodiments: HIC, GHSC, SZC, SWC, HSCC (hydrogen SCC), SSC (sulphur stress cracking). What I’m trying to convey here – their very mechanism is the same H+ charge, and this understanding is essential for the damage control.  

Quest: Would you reckon the metal in the below picture ‘embrittled’ or not, and why did it happen?

Stress Corrosion Cracking (SCC)

Firstly, we need to think about the SCC definition – that is a metal cracking under stress in a corrosive environment. Thus, two processes are needed for SCC: external mechanical loading and a corrosive environment acting as an aggravator by means of the protons penetration and combination.

One well known instance of residual stresses assisting fatigue is the welding induced residual stress, which is typically alleviated by a follow-up heat treatment (PWHT). This is yet a purely mechanical damage (‘Fatigue’ in the right column of the diagram). What needs to be added to convert it into an SCC mechanism is welding in a humid environment (such as a tropical rain) to electrolyse the water:

Subject to good fabrication control, SCC instances in chemical industries are often due to the various aggressive media listed in the middle column of the diagram, but not limited to. This is because violent electro-chemical corrosion instances in non-neutral pH solutions can yet produce the hydrogen charge sufficient to manifest as micro-cracking. For example, a relatively weak carbonic acid produced on a combination of CO2 and H2O (eq. 3) is still quite harmful for low carbon steels, and SCC potential can’t be excluded in a wide range of acid and alkaline corrosion mechanisms:

By the way, the pH stands for the concentration of hydrogen ion [ref encyclopaedia Britannica] and shows the exponent of that concentration in a litre of liquid, e.g. 1?10-pH gram/litre. In this way, seawater pH=9 means that there is roughly 1?10-9 gram of hydrogen ions in 1 litre of the water. Though, these ions H+ are yet bound chemically, unless a chemical reaction frees them up (eq. 1).

What also varies amongst different corrosion agents is the reaction kinetics, and hence – the total amount of produced protons. In this view, given a prolonged exposure, the list of the SCC particular embodiments (middle column in 1st diagram) can be virtually endless (Lemon Juice SCC, LJSCC etc).

‘Mysterious’ cases of SCC

A question ‘Can we have SCC without prior corrosion instances’ has two answers:

1. No, in a controlled environment where other sources of H+ are excluded (CUI, Chloride SCC)
2. Yes, in a complex chemical environments where H+ sources can be overlooked

The option 2 is apparently responsible for some ‘mysterious’ cases of cracking without corrosion:

• Compressor impeller cracked in a ‘sweet’ gas environment
• Stainless steel tubes cracked in a steam service

(please give me more examples!)

These are likely instances of soluble salts present in the stream. The cruelty of salts is that they can deposit on metal surface, and re-hydrate upon moisture availability. This availability can be an upset, fluctuating or irregular situation, thus disabling a prediction of future damage rates. Following re-hydration, the eq. 1 will apply and may deliver that necessary H+ cause for a salt specific SCC embodiment to proceed.

Another possible explanation is a mechanical transmission of ions produced somewhere upstream, which is more likely in conditioning systems using chemicals (water treatment, corrosion inhibition).

And the final option is that the cracking occurred due to metallurgical or mechanical mechanisms.

Embrittlement mechanisms other than SCC

There are few ‘embrittlement’ embodiments which aren’t SCC, but use different mechanisms:

• Low temperature embrittlement (cold embrittlement) manifests itself as carbon steels ductility drop, found below a certain negative temperature, and is measured via impact (Charpy) tests. Degraded ductility minimizes plastic zones at a crack tips, and cracks propagate much faster under same stress level, ending up with early fatigue failures or a catastrophic fracture. Basically, this happens due to weakening inter-atomic bonds, and isn’t relevant to a corrosion or hydrogen charge. This is addressed by refining the steel grain structure, or by material change to stainless steels (cold service) or aluminium (cryo service).
• Sigma-phase, Temper and 475°C ‘embrittlements’ are metallurgical phenomena of alloy phase changes in a certain temperature range, which trigger fracture toughness degradation and reduce the material strength. A metallurgical damage is a sign of a wrong material selection for a high temperature service or an inadequate welding/heat treatment procedure.
• Liquid Metal Embrittlement (LME/LMC) occurs in certain pairs liquid metal – solid metal, for example, liquid mercury condensation on aluminium cryogenic equipment. One explanation of the mechanism is penetration of the liquid metal into the grain boundaries and forming brittle inter-metallic compounds there. Susceptible metal pairs are known and should be avoided. In particular welding spatter or contact with galvanized (Zn) steel caused many instances of stainless steel cracking.

We can see that the common term ‘embrittlement’ is traditionally used for essentially different damage mechanisms and drives some confusion. While not attempting to re-name the conventional damage classification, I suggest the reader to focus on the particular damage nature, to understand its drivers and implications more confidently.

This is why I tend using simply ‘SCC’ to designate a consequence of any ‘pH influenced corrosion’.

So, what do we do with SCC?

Misfortunately, the SCC processes start occurring at a microscopic scale, and addressing these changes by replicas or specimen testing is rarely practical. Notably, the longer a crack is – the faster it propagates, and an MPI/DPI or visual inspection regime has a typical crack detectability threshold not less than roughly 0.1 mm. At this length, highly stressed cracks can be already near-critical, depending on the applied stress levels. As a result, inspection for SCC at sensible regular intervals (months or years) and using traditional methods is rather unable to prevent sudden failures by SCC.

There are some developments of advanced Non-Destructive Techniques (NDT) to identify material micro-cracking non-intrusively, which are very valuable indeed, but limited in applications by high costs and the NDT coverage dilemma (which we discussed earlier here).

One final argument for using something other than regular inspections would be: ‘From any SCC inspection results, what will be likely recommended to minimize the failure risks?’ It can be:

1. Corrosion Resistant Alloys (CRA) for corrosive environments, specifically for high consequence equipment. This option will minimize the H+ production and also corrosion failures risks altogether.
2. Post Weld Heat Treatment (PWHT) of heat affected zones in cyclic service. This will relieve the weld induced residual stresses at least, and can also refine the steel grain structure. Note that fatigue effects are non-linear on the stress scale: a small alleviation of peak stresses can significantly extend the ‘safe’ life of the stressed equipment.
3. Environment change – better stream conditioning and/or corrosion inhibition, if possible.
4. Monitoring the stream composition and setting alarms on surpassing the damage thresholds. This route aligns with the Integrity Operating Windows (IOW) concept, with a subtle difference that only the damage thresholds are looked at, with no attempts to predict the future damage progress analytically. Instead, sounding alarm suggests performing a follow-up inspection for actual integrity condition, which may provide a basis for prognostics. Advanced and perspective NDT techniques can be applied where consequence levels warrant.

All these measures would be most effective being applied preventively, e.g. at the design stage.

This brief article is of a popular character and is intended to create a holistic picture of the problem, rather than particular engineering applications which we disclaim here. Particular integrity problems require more in-depth analysis which we are happy to help with on the basis of our R&D. 

If you found this article interesting, please comment with your thoughts, critics and questions.