Corrosion: Understand It to Fight It
EP Editorial Staff | January 23, 2018
Knowing what your operations are up against is crucial in preventing costly, potentially dangerous, damage.
By Neville Sachs, P.E.
The plant’s roof was typical in that it was almost flat, with well-placed drains. One section, though, was not draining properly and vegetation had begun growing in the shallow pond that resulted from the situation. After discussions with the site’s engineering and maintenance staff and a walk-through of the facility, failure-analysis consultants who had been working on other projects at the site uncovered the problem: The troublesome section of the roof was over an area that was frequently wet, and the supporting columns had buckled several inches because of corrosion.
The issue with that plant roof was a close cousin to that shown in Fig. 1. The railroad-bridge-support column in this image is an excellent example of the costly, potentially dangerous, damage that corrosion can cause in infrastructure and other plant assets.
CORROSION COSTS
There have been many studies about the cost of corrosion.
NACE International, formerly known as the National Association of Corrosion Engineers (nace.org, Houston), conducted research from 1999 to 2001 that found direct corrosion costs in the United States amounted to $276 million (about 3.2% of the country’s GDP). In March 2016, NACE released a study that estimated worldwide corrosion amounted to $2.5 trillion (about 3.4% of the GDP) and indirect costs doubled that.
In the 2016 study, NACE estimated that between 15% and 35% of those corrosion-related costs could be eliminated using current technology. Researchers also offered that, comparing corrosion costs in 1975 with those in 1999, an intelligent approach to automobile design and elimination of corrosion appeared to have reduced the cost to American consumers by about 52%.
HOW CORROSION OCCURS
Energy is needed to convert mined ores into useful metals. Corrosion is the natural result of those metals trying to revert back to their original states. Consider, for example, that there’s very little difference between the rust from corroded steel and the iron ores that were originally refined to make that steel.
The actual corrosion process is an electrochemical reaction. Depicting a steel bar in a liquid, Fig. 2, shows how this reaction takes place. In the diagram, corrosion is attacking the anode, with iron ions being released into the solution, while hydrogen is being generated at the cathode. Water (H2O), is made up of two hydrogen ions and one oxygen ion. The iron ions from the anode (the Fe symbols) will ultimately unite with oxygen in the water, whereupon several different types of rust can form.
At the cathode site of the piece, atomic hydrogen is being released. Most of those hydrogen ions then mate with another hydrogen ion and form molecular hydrogen, the readily flammable gas we’re used to thinking about. But some of the ions remain solitary and they are the cause of the many forms of hydrogen damage including hydrogen embrittlement, cracking, and blisters.
For wet corrosion, a liquid must be present to provide the complete circuit required by the electromechanical reaction. Electrons that flow from the cathode to the anode have to eventually return to the cathode, and they do so by traveling through the liquid.
Referring back to the railroad-bridge-support column in Fig. 1, note the presence of a fair amount of silt. This has two substantial effects on the corrosion rate:
The silt holds moisture that allows corrosion to attack the column for a longer period of time than if the steel were dry.
Chemicals such as road salt are in the silt. As the moisture in it evaporates, the chemical concentration increases. The chemicals, in turn, make the water more electrically conductive and significantly increase the rate of corrosion.
Temperature is a third important factor in corrosion. Below freezing, ice can’t conduct corrosion currents. But, as the temperature increases, the corrosion rate increases. A good example is the rapid attack on hot piping with moist insulation. The exact solution chemistry has a major effect, but up to about 175 F (80 C), the corrosion rate usually rapidly increases, then drops off and ceases when the liquid vaporizes.
TYPES OF CORROSION
While more than 99% of corrosion losses are from the wet variety, dry corrosion also occurs, but only at greatly elevated temperatures. A common example would be the formation of oxide scales on a barbecue grill.
In North America, the recognition and description of corrosion have been driven largely by work initiated by Dr. Mars Fontana of The Ohio State University (osu.edu, Columbus, OH). We usually describe corrosion by how it appears. The common categories (or types) are:
Uniform corrosion causes about 80% of all corrosion. It occurs where anode and cathode sites relatively uniformly swap position. Examples include the railroad-bridge-support column shown in Fig. 1, buried steel water lines, nooks and crannies on vehicles where deposits build up, and machine frames and bases in damp areas.
Pitting corrosion manifests as isolated areas of attack. With carbon steel, it may take years before leakage occurs while stainless-steel pitting might progress at a rate of 0.001 in. (0.025 mm)/day. Steel examples frequently include water and wastewater tanks. Stainless-steel examples include external areas with dirt deposits on them.
Galvanic corrosion occurs when two chemically different metals are joined. One is always the anode and continuously attacked, protecting the other piece. A common example involves a joint between steel and copper pipe, where the steel will always be attacked.
Figure 3 shows a bronze fitting and a steel pipe that had been submerged in water. Perforation of the freshly cut pipe threads happened in only nine months.
Selective leaching is essentially galvanic corrosion within a metal. The common industrial application involves buried cast-iron water or waste lines where the graphite in the iron acts as a cathode, and the iron is eaten away, leaving a weak and brittle graphite pipe. When initially excavated, the pipe may appear almost undamaged, but sandblasting will rapidly remove the graphite leaving proof of the mechanism. (A frequent problem with buried-pipe replacement is that the new piece is always anodic to the older sections. The new one will rapidly corrode and leak, and personnel will blame the material, not knowing that the actual problem is their lack of corrosion knowledge.)
Crevice corrosion occurs in a small gap between two pieces of metal. It allows a corrosion mechanism to act in a way that’s similar to pitting corrosion. Although it’s not a common industrial mechanism, it can happen with poor joint control on welded assemblies.
Intergranular corrosion involves galvanic attack at the grain boundaries within a metal. It’s usually associated with a poor choice in materials of construction for chemical processes.
Erosion corrosion is a combination of actions. Corrosion results in an oxide on a metal’s surface. The oxide, though, slows the attack because it prevents fresh corrodent from reaching the surface. If there’s a fast fluid flow that scrubs the oxide off the surface, corrosion continues at a very rapid rate. A common site for erosion corrosion is the outer radius of piping elbows in steel lines with untreated waters and flow rates exceeding approximately 10 ft./sec. (3 m/sec). It’s also been seen in pumps as a result of poor choices of construction materials.
The previous seven categories/types are basically different-looking versions of galvanic corrosion. Two other corrosion types—stress-corrosion cracking and hydrogen damage—result in metallurgical damage leading to often hard-to-detect catastrophic failures.
Stress corrosion cracking (SCC) can occur with almost any metal and is the result of a combination of stress, a chemistry that attacks the metal’s structure, and a susceptible metal. Industrially, although it is sometimes seen with nitrates and steel, the most common situation involves 300 series (austenitic) stainless steels and chlorides.
One interesting example of the interaction of stress, a sensitive material, and attacking chemicals involved a series of large stainless-steel vessels. The problem stemmed from two critical design flaws: The tanks were downwind of several cooling towers, and their roofs were supported by ASTM A 36 steel beams. Thermal expansion caused stresses where the beams were welded to the roof panels, and drying of the mist off the cooling towers increased chloride content in the air. After about 10 years of operation, hundreds of small stress corrosion cracks in the vessels were noted. Consultants monitored this collection of cracks for approximately two years and saw little change in them. When operators increased the pressure in the tanks, however, the number of cracks suddenly increased.
The following real-world examples are representative of more common SCC problems in plants:
A 500-gal. (1.9 m3) tank where careless filling led to overflow: The tank held a 160 F (70 C ) caustic cleaning solution. The overflow ran down the tank’s side and evaporated, eventually reaching a concentration at which cracking occurred.
A series of stainless-steel lines that carried oil for process machinery: Surface temperature of the oil lines exceeded 140 F (60 C). In a scenario similar to the above example, process water dripping on the hot lines caused cracking. Figure 4 shows a view through a low-power microscope of one 1/2-in.-dia. (12-mm-dia.) line. A jagged and irregular stress corrosion crack is readily visible, as are a number of pitted areas.
Hydrogen damage is a particularly diabolical form of attack because it is often impossible to see prior to catastrophic failure. It starts with the corrosion reaction reflected in the image of the railroad-bridge-support column in Fig. 1, and can go on to cause hydrogen embrittlement in metals such as titanium, blisters in low-carbon steel, and cracking in hardened steels.
Hydrogen blisters and cracking appear to be similar mechanisms in that the hydrogen ions are so incredibly small they can float through the atomic structure of steel until they find an irregularity. Because of the void, the ion tends to rest there until another one comes along and the two combine to form a hydrogen-gas molecule, which has a much larger volume than the single ion. The result is an internal pressure that, when an external stress is applied, causes blisters in mild-steel components and cracking in hardened ones. NACE advises that hydrogen cracking can occur in steel with hardness values as low as HRC 23, if sulfur compounds are present. It’s even been seen in leaking steam-line joints with cracks in flange bolts that were HRC 27. Rockwell hardness [HRC] readings are one of the ways of measuring the hardness and strength of a bolt. SAE Grade 5 and ASTM A 325 bolts are commonly in the range of HRC 25 to HRC 34. SAE Grade 8, U.S. socket-head cap screws and ASTM A 490 bolts typically range from HRC 33 to HRC 39. Because of their lower ductility, these high-strength components are much more susceptible to hydrogen cracking in corrosive environments. A cardinal rule of good reliability is to not use high-strength bolts in such areas.
Although hardened bolts are among the most common examples of hydrogen cracking, leaf and coil springs are other frequent victims. The coil spring shown in Fig. 5 had a ground plain end supported on a cup-like frame that was frequently wet. The corrosion is obvious.
KEEP IN MIND
The battle against corrosion is never ending. In summary, if an area is wet and metal isn’t protected, there will be corrosion. What’s worse, the seriousness of the damage caused by this scourge may not be recognized for years. EP
Forget About ‘The Other Guys.’ What About Your Operations?
There’s a tendency by some plant personnel to “look at the other guys,” rather than effectively confront problems in their own operations. With regard to corrosion, that might mean focusing on the idea that other sites’ costs must be higher, or believing that paper mills, with their hot and humid environments, must surely have more potential for corrosion than, say, air-conditioned manufacturing facilities. But two of the most memorable failure analyses I ever encountered actually involved air-conditioned plants.
• The first was a manufacturing operation where a buried water line erupted twice in eight years. Each of those two events shut the site’s machining lines down for several days.
• The second involved a pharmaceutical plant where many millions of dollars of product had to be scrapped because of corrosion in a cooling system’s sensor.
Remember this: Practically anywhere the relative humidity exceeds 60%, there’s a potential for corrosion. No plant can escape it. —N.S.
Neville Sachs has spent many years working in the field of machinery reliability and lubrication for a wide range of industries. The author of two books on failure analysis and a contributor to others, he has written more than 40 articles on these topics. A Registered Professional Engineer, he holds STLE’s CLS certification, among others. Contact him directly at nevsachseng@gmail.com.
To learn more about preventing/managing corrosion, visit: nace.org/resources or corrosion-doctors.org. (This resource operates out of Canada’s Royal Military College, Kingston, Ontario.)
View Comments