Analysis and Prevention of Corrosion-Related Failures S.R. Freeman, Millennium Metallurgy, Ltd Introduction CORROSION is the deterioration of a material by a reaction with its environment. In a study by the U.s Federal Highway Administration in cooperation with the National Association of Corrosion Engineers(NACE), it was estimated that the annual cost of corrosion was between $121 and $138 billion in 1998 in the United States. These costs included cost of corrosion-control methods, equipment, and services; cost of labor attributed to corrosion management; cost of use of more expensive materials to lessen corrosion damage; and cost of lost revenue, loss of reliability, and loss of capital due to corrosion deterioration. Only selected industrial sectors were analyzed in the study. When extrapolated to all U.S. industries, the total cost estimate is $276 billion, or more than 3% of the U.S. gross domestic product(Ref 1). This great cost is a measure of the importance of corrosion management and an indication of the significance of potential cost saving that corrosion abatement can yield Reference cited in this section 1."Corrosion Costs and Preventive Strategies in the United States, FHWA-RD-01-156, Federal Highway Administration 2002 Analysis and Prevention of Corrosion-Related Failures S.R. Freeman, Millennium Metallurgy Ltd Electrochemical Nature of Corrosion The articles in this Section are devoted to the identification and analysis of corrosion- related failures, the categorization of corrosion failures by form and mechanism, and the application of preventive measures. The mechanisms of corrosion are described in more detail in Corrosion, Volume 13 of the asm handbook However, as a brief introduction, the electrochemical nature of corrosion can be illustrated by the attack on zinc by hydrochloric acid. When zinc is placed in dilute hydrochloric acid, a vigorous reaction occurs; hydrogen gas is evolved and the zinc dissolves, forming an acidic aqueous solution of zinc chloride. The reaction is Zn+2HCl→ZnCl2+H (Eq I Since the chloride ion is not involved in the reaction, this equation can be written in the simplified form Zn+2H+→Zn2++H2 (Eq2) Zinc reacts with the hydrogen ions of the acid solution to form zinc ions and hydrogen gas. Equation 2 shows that during the reaction, zinc is oxidized to zinc ions and hydrogen ions are reduced to hydrogen. Thus, eq 2 can be conveniently divided into two reactions: the oxidation of zinc and the reduction of hydrogen ions Oxidation(anodic reaction)Zn-Zn+2 Reduction(cathodic reaction) 2H 2e H2
Analysis and Prevention of Corrosion-Related Failures S.R. Freeman, Millennium Metallurgy, Ltd. Introduction CORROSION is the deterioration of a material by a reaction with its environment. In a study by the U.S. Federal Highway Administration in cooperation with the National Association of Corrosion Engineers (NACE), it was estimated that the annual cost of corrosion was between $121 and $138 billion in 1998 in the United States. These costs included cost of corrosion-control methods, equipment, and services; cost of labor attributed to corrosion management; cost of use of more expensive materials to lessen corrosion damage; and cost of lost revenue, loss of reliability, and loss of capital due to corrosion deterioration. Only selected industrial sectors were analyzed in the study. When extrapolated to all U.S. industries, the total cost estimate is $276 billion, or more than 3% of the U.S. gross domestic product (Ref 1). This great cost is a measure of the importance of corrosion management and an indication of the significance of potential cost saving that corrosion abatement can yield. Reference cited in this section 1. “Corrosion Costs and Preventive Strategies in the United States,” FHWA-RD-01-156, Federal Highway Administration, 2002 Analysis and Prevention of Corrosion-Related Failures S.R. Freeman, Millennium Metallurgy, Ltd. Electrochemical Nature of Corrosion The articles in this Section are devoted to the identification and analysis of corrosion-related failures, the categorization of corrosion failures by form and mechanism, and the application of preventive measures. The mechanisms of corrosion are described in more detail in Corrosion, Volume 13 of the ASM Handbook. However, as a brief introduction, the electrochemical nature of corrosion can be illustrated by the attack on zinc by hydrochloric acid. When zinc is placed in dilute hydrochloric acid, a vigorous reaction occurs; hydrogen gas is evolved and the zinc dissolves, forming an acidic aqueous solution of zinc chloride. The reaction is: Zn + 2HCl → ZnCl2 + H2 (Eq 1) Since the chloride ion is not involved in the reaction, this equation can be written in the simplified form: Zn + 2H+ → Zn2+ + H2 (Eq 2) Zinc reacts with the hydrogen ions of the acid solution to form zinc ions and hydrogen gas. Equation 2 shows that during the reaction, zinc is oxidized to zinc ions and hydrogen ions are reduced to hydrogen. Thus, Eq 2 can be conveniently divided into two reactions: the oxidation of zinc and the reduction of hydrogen ions: Oxidation (anodic reaction) Zn → Zn2+ + 2e (Eq 3) Reduction (cathodic reaction) 2H+ + 2e → H2 (Eq 4)
An oxidation or anodic reaction is indicated by an increase in valence or a release of electrons. a decrease in valence charge or the consumption of electrons signifies a reduction or cathodic reaction. Equations 3 and 4 are partial reactions; both must occur simultaneously and at the same rate on the metal surface. If this were not true, the metal would spontaneously become electrically charged, which is clearly impossible The corrosion of zinc in hydrochloric acid is an electrochemical process. That is, any reaction that can be divided into two or more partial reactions of oxidation and reduction due to the transfer of electrical charge is termed electrochemical. Dividing corrosion or other electrochemical reactions into partial reactions makes them simpler to understand. Iron and aluminum, like zinc, are also rapidly corroded by hydrochloric acid Thus, the problem of hydrochloric acid corrosion is simplified because in every case the cathodic reaction is the evolution of hydrogen gas according to Eq 4. This also applies to corrosion in other acids such as sulfuric, phosphoric, hydrofluoric, and water-soluble organic acids such as formic and acetic. In each case, only the hydrogen ion is active, the other ions such as sulfate, phosphate, and acetate do not participate in the electrochemical reaction When viewed from the standpoint of partial processes of oxidation and reduction, all corrosion can be classified into a few generalized reactions. The anodic reaction in every corrosion reaction is the oxidation of a metal to its ion. This can be written in the general form where n is the number of electrons released A few examples are Ag→Ag+e Zn→Zn2++2e (Eq7) Al→Al+3e In each case the number of electrons produced equals the valence of the ion There are several different cathodic reactions that are frequently encountered in metallic corrosion. The most common cathodic reactions are: Hydrogen evolution 2H+2 H (Eq9) Oxygen reduction(acid solutions) O2+4H+4e→2H2O (Eq10) Oxygen reduction(neutral or basic solutions) O2+2H2O+4e→4OH Metal-ion reduction Metal deposition M+e→M (Eq13) Hydrogen evolution is a common cathodic reaction because acid or acidic media are frequently encountered Oxygen reduction is very common, because any aqueous solution in contact with air is capable of producing this reaction. Metal-ion reduction and metal deposition are less-common reactions and are most frequently found in chemical process streams. All of the above reactions are quite similar; they consume electrons The above partial reactions can be used to interpret virtually all electrochemical corrosion problems. Consider what happens when iron is immersed in water or seawater that is exposed to the atmosphere(an automobile fender or a steel pier piling are examples). Corrosion occurs. The anodic reaction is (Eq14) Since the medium is exposed to the atmosphere, it contains dissolved oxygen. Water and seawater are nearly neutral. and thus the cathodic reaction is: O2+2H2O+4e→4OH (Eq15) Remembering that sodium and chloride ions do not participate in the reaction, the overall reaction can be obtained by adding eq 9 and 12 Thefileisdownloadedfromwww.bzfxw.com
An oxidation or anodic reaction is indicated by an increase in valence or a release of electrons. A decrease in valence charge or the consumption of electrons signifies a reduction or cathodic reaction. Equations 3 and 4 are partial reactions; both must occur simultaneously and at the same rate on the metal surface. If this were not true, the metal would spontaneously become electrically charged, which is clearly impossible. The corrosion of zinc in hydrochloric acid is an electrochemical process. That is, any reaction that can be divided into two or more partial reactions of oxidation and reduction due to the transfer of electrical charge is termed electrochemical. Dividing corrosion or other electrochemical reactions into partial reactions makes them simpler to understand. Iron and aluminum, like zinc, are also rapidly corroded by hydrochloric acid. Thus, the problem of hydrochloric acid corrosion is simplified because in every case the cathodic reaction is the evolution of hydrogen gas according to Eq 4. This also applies to corrosion in other acids such as sulfuric, phosphoric, hydrofluoric, and water-soluble organic acids such as formic and acetic. In each case, only the hydrogen ion is active, the other ions such as sulfate, phosphate, and acetate do not participate in the electrochemical reaction. When viewed from the standpoint of partial processes of oxidation and reduction, all corrosion can be classified into a few generalized reactions. The anodic reaction in every corrosion reaction is the oxidation of a metal to its ion. This can be written in the general form: M → M n+ + ne (Eq 5) where n is the number of electrons released. A few examples are: Ag → Ag+ + e (Eq 6) Zn → Zn2+ + 2e (Eq 7) Al → Al3+ + 3e (Eq 8) In each case the number of electrons produced equals the valence of the ion. There are several different cathodic reactions that are frequently encountered in metallic corrosion. The most common cathodic reactions are: Hydrogen evolution: 2H+ + 2e → H2 (Eq 9) Oxygen reduction (acid solutions): O2 + 4H+ + 4e → 2H2O (Eq 10) Oxygen reduction (neutral or basic solutions): O2 + 2H2O + 4e → 4OH- (Eq 11) Metal-ion reduction: M 3+ + e → M 2+ (Eq 12) Metal deposition: M + + e → M (Eq 13) Hydrogen evolution is a common cathodic reaction because acid or acidic media are frequently encountered. Oxygen reduction is very common, because any aqueous solution in contact with air is capable of producing this reaction. Metal-ion reduction and metal deposition are less-common reactions and are most frequently found in chemical process streams. All of the above reactions are quite similar; they consume electrons. The above partial reactions can be used to interpret virtually all electrochemical corrosion problems. Consider what happens when iron is immersed in water or seawater that is exposed to the atmosphere (an automobile fender or a steel pier piling are examples). Corrosion occurs. The anodic reaction is: Fe → Fe2+ + 2e (Eq 14) Since the medium is exposed to the atmosphere, it contains dissolved oxygen. Water and seawater are nearly neutral, and thus the cathodic reaction is: O2 + 2H2O + 4e → 4OH- (Eq 15) Remembering that sodium and chloride ions do not participate in the reaction, the overall reaction can be obtained by adding Eq 9 and 12: The file is downloaded from www.bzfxw.com
2Fe+2H2O+O2→2Fe2++4OH→2FeOH2 Ferrous(Fet) hydroxide precipitates from solution. However, this compound is unstable in oxygenated solutions and is oxidized to the ferric(Fe)salt 2Fe(OH)2 +H2O+-O2- 2Fe(OH) The final product is the familiar rust The classic example of a replacement reaction, the interaction of zinc with copper sulfate solution, illustrates metal deposition Zn + Cu Zn-+ Cu (Eq18) or, viewed as partial reactions Zn→Zn2++2e The zinc initially becomes plated with copper and eventually, given enough time and reactants, the products are copper sponge and zinc sulfate solution During corrosion, more than one oxidation and one reduction reaction may occur. when an alloy is corroded its component metals go into solution as their respective ions. More importantly, more than one reduction reaction can occur during corrosion. Consider the corrosion of zinc in aerated hydrochloric acid. Two cathodic reactions are possible: the evolution of hydrogen and the reduction of oxygen. Since the rates of oxidation and solutions containing dis c Increasing the total reduction rate increases the rate of zinc solution. Therefore, acid reduction must be equal solved oxygen normally will be more corrosive than air-free acids. Oxygen reduction simply provides another means of electron disposal. The same effect is observed if any oxidizer is present in acid solutions. A frequent impurity in commercial hydrochloric acid is the ferric ion(Fe), present as ferric chloride. Metals corrode much more rapidly in such impure acid because there are two cathodic reactions hydrogen evolution and ferric ion reduction q21) The anodic and cathodic reactions occurring during corrosion are mutually dependent, and it is possible to reduce corrosion by reducing the rates of either reaction. In the above case of impure hydrochloric acid, it can be made less corrosive by removing the ferric ions and consequently reducing the total rate of cathodic reduction. Oxygen reduction is eliminated by preventing air from contacting the aqueous solution or by removing air that has been dissolved. Iron is nearly inert in air-free water or seawater because there is limited athodic reaction possible If the surface of the metal is coated with paint or other nonconducting film, the rates of both anodic and cathodic reactions will be greatly reduced and corrosion will be retarded. A corrosion inhibitor is a substance that, when added in small amounts to a corrosive, reduces its corrosivity. Corrosion inhibitors function by interfering with either the anodic or cathodic reactions, or both. Many of these inhibitors are organic compounds; they function by forming an impervious film on the metal surface or by interfering with either the anodic or cathodic reactions. High-molecular-weight amines retard the hydrogen-evolution reaction and subsequently reduce corrosion rate. Adequate conductivity in both the metal and the electrolyte is required for continuation of the corrosion reaction. Of course, it is not practical to increase the electrical resistance of the metal because the sites of the anodic and cathodic reactions are not known, nor are they predictable. However it is possible to increase the electrical resistance of the electrolyte and thereby reduce corrosion. Very pure water is much less corrosive than impure or natural waters. The low corrosivity of high-purity water is due to its high electrical resistance and few reducible cations Passivity. Essentially, passivity refers to the loss of chemical reactivity experienced by certain metals and alloys under particular environmental conditions. That is, certain metals and alloys become essentially inert and act as if they were noble metals such as platinum and gold. Fortunately, from an engineering standpoint, the metals most susceptible to this kind of behavior are the common engineering and structural materials, including iron metals such as zinc, cadmium, tin, uranium, and thorium have also been observed to exhibit passivity effect r nickel, silicon, chromium, titanium, and alloys containing these metals. Also, under limited conditions, othe
2Fe + 2H2O + O2 → 2Fe2+ + 4OH- → 2Fe(OH)2 (Eq 16) Ferrous (Fe2+) hydroxide precipitates from solution. However, this compound is unstable in oxygenated solutions and is oxidized to the ferric (Fe3+) salt: 2Fe(OH)2 + H2O + 1 2 O2 → 2Fe(OH)3 (Eq 17) The final product is the familiar rust. The classic example of a replacement reaction, the interaction of zinc with copper sulfate solution, illustrates metal deposition: Zn + Cu2+ → Zn2+ + Cu (Eq 18) or, viewed as partial reactions: Zn → Zn2+ + 2e (Eq 19) Cu2+ + 2e → Cu (Eq 20) The zinc initially becomes plated with copper and eventually, given enough time and reactants, the products are copper sponge and zinc sulfate solution. During corrosion, more than one oxidation and one reduction reaction may occur. When an alloy is corroded, its component metals go into solution as their respective ions. More importantly, more than one reduction reaction can occur during corrosion. Consider the corrosion of zinc in aerated hydrochloric acid. Two cathodic reactions are possible: the evolution of hydrogen and the reduction of oxygen. Since the rates of oxidation and reduction must be equal, increasing the total reduction rate increases the rate of zinc solution. Therefore, acid solutions containing dissolved oxygen normally will be more corrosive than air-free acids. Oxygen reduction simply provides another means of electron disposal. The same effect is observed if any oxidizer is present in acid solutions. A frequent impurity in commercial hydrochloric acid is the ferric ion (Fe3+), present as ferric chloride. Metals corrode much more rapidly in such impure acid because there are two cathodic reactions, hydrogen evolution and ferric ion reduction: Fe3+ + e → Fe2+ (Eq 21) The anodic and cathodic reactions occurring during corrosion are mutually dependent, and it is possible to reduce corrosion by reducing the rates of either reaction. In the above case of impure hydrochloric acid, it can be made less corrosive by removing the ferric ions and consequently reducing the total rate of cathodic reduction. Oxygen reduction is eliminated by preventing air from contacting the aqueous solution or by removing air that has been dissolved. Iron is nearly inert in air-free water or seawater because there is limited cathodic reaction possible. If the surface of the metal is coated with paint or other nonconducting film, the rates of both anodic and cathodic reactions will be greatly reduced and corrosion will be retarded. A corrosion inhibitor is a substance that, when added in small amounts to a corrosive, reduces its corrosivity. Corrosion inhibitors function by interfering with either the anodic or cathodic reactions, or both. Many of these inhibitors are organic compounds; they function by forming an impervious film on the metal surface or by interfering with either the anodic or cathodic reactions. High-molecular-weight amines retard the hydrogen-evolution reaction and subsequently reduce corrosion rate. Adequate conductivity in both the metal and the electrolyte is required for continuation of the corrosion reaction. Of course, it is not practical to increase the electrical resistance of the metal because the sites of the anodic and cathodic reactions are not known, nor are they predictable. However, it is possible to increase the electrical resistance of the electrolyte and thereby reduce corrosion. Very pure water is much less corrosive than impure or natural waters. The low corrosivity of high-purity water is due to its high electrical resistance and few reducible cations. Passivity. Essentially, passivity refers to the loss of chemical reactivity experienced by certain metals and alloys under particular environmental conditions. That is, certain metals and alloys become essentially inert and act as if they were noble metals such as platinum and gold. Fortunately, from an engineering standpoint, the metals most susceptible to this kind of behavior are the common engineering and structural materials, including iron, nickel, silicon, chromium, titanium, and alloys containing these metals. Also, under limited conditions, other metals such as zinc, cadmium, tin, uranium, and thorium have also been observed to exhibit passivity effects
Passivity, although difficult to define, can be quantitatively described by characterizing the behavior of metals that show this unusual effect. first consider the behavior of what can be called an active metal that is a metal that does not show passivity effects. The lower part of the curve in Fig. I illustrates the behavior of such a metal. Assume that a metal is immersed in an air-free acid solution with an oxidizing power corresponding to 6,int a and a corrosion rate corresponding to this point. If the oxidizing power of this solution is increased, say, by adding oxygen or ferric ions, the corrosion rate of the metal will increase rapidly. Note that for such a metal the corrosion rate increases as the oxidizing power of the solution increases. This increase in rate is exponential and yields a straight line when plotted on a semilogarithmic scale as in Fig. 1. The oxidizing power of the solution is controlled by both the specific oxidizing power of the reagents and the concentration of these reagents. Oxidizing power can be precisely defined by electrode potential, but is beyond the scope of this discussion Transpassive 6asN8oco60 Passive Active 100010.000 Corrosion rate Fig. 1 Corrosion characteristics of an active-passive metal as a function of solution oxidizing power(electrode potential) The behavior of this metal or alloy can be conveniently divided into three regions: active, passive, and transpassive. In the active region, slight increases in the oxidizing power of the solution cause a corresponding rapid increase in the corrosion rate. However, at some point, if more oxidizing agent is added the corrosion rate shows a sudden decrease. This corresponds to the beginning of the passive region. Further increases in oxidizing agents produce little if any change in the corrosion rate of the material in the passive region. Finally at very high concentrations of oxidizers or in the presence of very powerful oxidizers, the corrosion rate again increases with increasing oxidizing power. This region is termed the transpassive region It is important to note that during the transition from the active to the passive region, a 10 to 10 reduction in corrosion rate is usually observed. Passivity is due to the formation of a surface film or protective barrier that is stable over a considerable range of oxidizing power and is eventually destroyed in strong oxidizing solutions Under conditions in which the surface film is stable. the anodic reaction is stifled and the metal surface is protected from corrosion. For example, stainless steel owes its corrosion-resistant properties to a passive surface film. The naturally occurring passive film is usually enhanced with immersion in a hot nitric acid solution or steam. For example, stainless steel surgical implants develop this passive layer when the implants are sterilized in steam The exact nature of this barrier that forms on the metal surface is not well understood. It nay be a very thin, transparent oxide film or a layer of adsorbed oxygen atoms. However, for the purposes of engineering application, it is not necessary to understand completely the mechanism. To summarize, metals that possess an active-passive transition become passive(very corrosion-resistant) in moderately to strongly oxidizing environments. Under extremely strong oxidizing conditions, these materials lose their corrosion- resistant properties Thefileisdownloadedfromwww.bzfxw.com
Passivity, although difficult to define, can be quantitatively described by characterizing the behavior of metals that show this unusual effect. First, consider the behavior of what can be called an active metal, that is, a metal that does not show passivity effects. The lower part of the curve in Fig. 1 illustrates the behavior of such a metal. Assume that a metal is immersed in an air-free acid solution with an oxidizing power corresponding to point A and a corrosion rate corresponding to this point. If the oxidizing power of this solution is increased, say, by adding oxygen or ferric ions, the corrosion rate of the metal will increase rapidly. Note that for such a metal, the corrosion rate increases as the oxidizing power of the solution increases. This increase in rate is exponential and yields a straight line when plotted on a semilogarithmic scale as in Fig. 1. The oxidizing power of the solution is controlled by both the specific oxidizing power of the reagents and the concentration of these reagents. Oxidizing power can be precisely defined by electrode potential, but is beyond the scope of this discussion. Fig. 1 Corrosion characteristics of an active-passive metal as a function of solution oxidizing power (electrode potential) The behavior of this metal or alloy can be conveniently divided into three regions: active, passive, and transpassive. In the active region, slight increases in the oxidizing power of the solution cause a corresponding rapid increase in the corrosion rate. However, at some point, if more oxidizing agent is added, the corrosion rate shows a sudden decrease. This corresponds to the beginning of the passive region. Further increases in oxidizing agents produce little if any change in the corrosion rate of the material in the passive region. Finally, at very high concentrations of oxidizers or in the presence of very powerful oxidizers, the corrosion rate again increases with increasing oxidizing power. This region is termed the transpassive region. It is important to note that during the transition from the active to the passive region, a 103 to 106 reduction in corrosion rate is usually observed. Passivity is due to the formation of a surface film or protective barrier that is stable over a considerable range of oxidizing power and is eventually destroyed in strong oxidizing solutions. Under conditions in which the surface film is stable, the anodic reaction is stifled and the metal surface is protected from corrosion. For example, stainless steel owes its corrosion-resistant properties to a passive surface film. The naturally occurring passive film is usually enhanced with immersion in a hot nitric acid solution or steam. For example, stainless steel surgical implants develop this passive layer when the implants are sterilized in steam. The exact nature of this barrier that forms on the metal surface is not well understood. It may be a very thin, transparent oxide film or a layer of adsorbed oxygen atoms. However, for the purposes of engineering application, it is not necessary to understand completely the mechanism. To summarize, metals that possess an active-passive transition become passive (very corrosion-resistant) in moderately to strongly oxidizing environments. Under extremely strong oxidizing conditions, these materials lose their corrosionresistant properties. The file is downloaded from www.bzfxw.com
Analysis and Prevention of Corrosion-Related Failures S.R. Freeman, Millennium Metallurgy Ltd Analysis of Corrosion-Related Failures Corrosion can be categorized as being uniform or localized Uniform corrosion is the most common form of corrosion. Its mechanism is electrochemical and is identified as a consistent loss of material over the entire exposed surface of the material in question. This is the type of corrosion that affects the greatest number of structures, such as buildings, bridges, pipelines, and outdoor equipment. It attacks the most surface area and the most costly. Uniform corrosion is also the easiest to analyze, to predict corrosion rates for, and to establish preventive measures and maintenance schedules for. Preventive techniques include painting(coatings) inhibition, and cathodic or anodic protection. The service life of the affected component can normally be estimated with a reasonable degree of accuracy, and catastrophic failures can be avoided Localized corrosion, such as crevice and pitting corrosion, intergranular corrosion, selective leaching, erosion- corrosion, and stress-corrosion cracking (SCC), act on a small portion of a component. The rate of localized corrosion is often orders of magnitude greater than that of uniform corrosion. Evidence of localized corrosion by virtue of its physical size and location, is often difficult to detect. These factors combine to make localized corrosion mechanisms insidious. If undetected they can lead to catastrophic system failures Certain environments induce specific corrosion failure modes. Hydrogen damage, liquid- and solid-metal- induced embrittlement, biologically and microbiologically influenced corrosion(MIC), and high-temperature corrosion are also addressed in articles in this section While categorizing local corrosion by form, mechanism, or environment aids understanding, failures are often attributed to a combination of modes. The combination of failure modes makes prediction of localized corrosion failures more difficult Corrosion failure involves the same general steps as other failures, as described in the article" Practices Failure Analysis"in this Volume. However, one potential difference between a general failure and a failure related to corrosion is the need with the latter for immediate preservation and protection of all evidence Corrosion failure also may require sampling and testing of corrosion products immediately, such as in the case of MIC, where viable cultures can provide the most meaningful result. When possible and cost effective, a site visit is most desirable. A site visit may provide the investigator with information that otherwise may have been omitted or overlooked Corrosion failures often relate to the material selection and the environment. details of the material pecifications, quality-assurance records, installation and maintenance records, and a history of the environment are all useful resources in resolving corrosion failures. Information regarding system upsets or diversions from the normal operating environment should be provided. A comparison of the actual material utilized with the material design specifications should also be performed History of Failed Parts. Obtaining a history of the operation of the failed component is crucial to solving the cause of failure. The operating environment, any changes in the environment, and temperature excursions should all be obtained. Any notation of previous failures or operating anomalies is useful. When possible, engineering drawings and sketches should be reviewed Information should be provided regarding any testing that was performed by the plant personnel. For instance, liquid-penetrant examination can result in chemical contamination of a surface. The use of paints or dyes to mark components may also alter the corrosion resistance of a part and the chemical composition of a corrosion product. It is also important that the investigator retains pertinent corrosion evidence before movement or disassembly of a component. The chain of events leading to, and occurring after, the failure should be documented On-Site Examination and Sampling. On-site examination should include a walking tour of the failure area Photographic documentation should be made to depict the conditions after the failure. If applicable, viewing a similar undamaged assembly or operation on the site where the failure occurred may be instructive Photographic documentation must be performed with special attention given to capturing the true colors of the corrosion products. Macrographs of the corrosion deposit should include any layering effects in the deposits
Analysis and Prevention of Corrosion-Related Failures S.R. Freeman, Millennium Metallurgy, Ltd. Analysis of Corrosion-Related Failures Corrosion can be categorized as being uniform or localized. Uniform corrosion is the most common form of corrosion. Its mechanism is electrochemical and is identified as a consistent loss of material over the entire exposed surface of the material in question. This is the type of corrosion that affects the greatest number of structures, such as buildings, bridges, pipelines, and outdoor equipment. It attacks the most surface area and is the most costly. Uniform corrosion is also the easiest to analyze, to predict corrosion rates for, and to establish preventive measures and maintenance schedules for. Preventive techniques include painting (coatings), inhibition, and cathodic or anodic protection. The service life of the affected component can normally be estimated with a reasonable degree of accuracy, and catastrophic failures can be avoided. Localized corrosion, such as crevice and pitting corrosion, intergranular corrosion, selective leaching, erosioncorrosion, and stress-corrosion cracking (SCC), act on a small portion of a component. The rate of localized corrosion is often orders of magnitude greater than that of uniform corrosion. Evidence of localized corrosion, by virtue of its physical size and location, is often difficult to detect. These factors combine to make localized corrosion mechanisms insidious. If undetected they can lead to catastrophic system failures. Certain environments induce specific corrosion failure modes. Hydrogen damage, liquid- and solid-metalinduced embrittlement, biologically and microbiologically influenced corrosion (MIC), and high-temperature corrosion are also addressed in articles in this Section. While categorizing local corrosion by form, mechanism, or environment aids understanding, failures are often attributed to a combination of modes. The combination of failure modes makes prediction of localized corrosion failures more difficult. Corrosion failure involves the same general steps as other failures, as described in the article “Practices in Failure Analysis” in this Volume. However, one potential difference between a general failure and a failure related to corrosion is the need with the latter for immediate preservation and protection of all evidence. Corrosion failure also may require sampling and testing of corrosion products immediately, such as in the case of MIC, where viable cultures can provide the most meaningful result. When possible and cost effective, a site visit is most desirable. A site visit may provide the investigator with information that otherwise may have been omitted or overlooked. Corrosion failures often relate to the material selection and the environment. Details of the material specifications, quality-assurance records, installation and maintenance records, and a history of the environment are all useful resources in resolving corrosion failures. Information regarding system upsets or diversions from the normal operating environment should be provided. A comparison of the actual material utilized with the material design specifications should also be performed. History of Failed Parts. Obtaining a history of the operation of the failed component is crucial to solving the cause of failure. The operating environment, any changes in the environment, and temperature excursions should all be obtained. Any notation of previous failures or operating anomalies is useful. When possible, engineering drawings and sketches should be reviewed. Information should be provided regarding any testing that was performed by the plant personnel. For instance, liquid-penetrant examination can result in chemical contamination of a surface. The use of paints or dyes to mark components may also alter the corrosion resistance of a part and the chemical composition of a corrosion product. It is also important that the investigator retains pertinent corrosion evidence before movement or disassembly of a component. The chain of events leading to, and occurring after, the failure should be documented. On-Site Examination and Sampling. On-site examination should include a walking tour of the failure area. Photographic documentation should be made to depict the conditions after the failure. If applicable, viewing a similar undamaged assembly or operation on the site where the failure occurred may be instructive. Photographic documentation must be performed with special attention given to capturing the true colors of the corrosion products. Macrographs of the corrosion deposit should include any layering effects in the deposits
Laboratory photography in a controlled setting should also be performed to guarantee accurate reproduction of color tones and surface textures The type and extent of on-site sampling depends on the environment and its availability. Piping corrosion in a domestic water service would require water samples from the source(incoming water supply )and from the end use(for example, a faucet). Microbiologically influenced corrosion may apply in certain cases. On-site testing or sample removal is necessary to retain the most accurate information regarding the type and number of bacteria involved. In most cases, removal of corrosion deposits from the surface may be performed in the field If the sample is undergoing extensive laboratory examination, it may be prudent to carefully retain the corrosion in place for laboratory documentation and removal If samples are removed on site, care must be taken to avoid any contamination. Sealed bags, latex gloves, tools for sample removal, buffered containers for water samples, MIC kits, and adhesive tape are useful for sample removal. (See the article"Biological Corrosion Failures" in this Volume for listings of commercially available kits. The analyst should avoid touching any corrosion product directly with bare hands to prevent contamination Physical removal of samples must be done in a manner to avoid further damage to the failed component and to avoid disturbance of corrosion products. In general, cutting must be done with care to avoid alterations of the metallurgical condition of the material and corrosion deposits. Saw cutting is generally preferred over torch cutting, since heating of the sample can affect the material and the corrosion product. If torch cutting is performed, a distance of 75 to 150 mm(3 to 6 in. )should be maintained from the area of interest. Saw cutting should be performed at a slow rate to avoid overheating. Use of lubricants and coolants should be avoided, if possible, to minimize contamination The proper shipment and storage of samples lessens the possibility of destruction of pertinent evidence Wrapping and sealing of the failed component will generally retain evidence and prevent further oxidation or contamination. A neutral environment may be required to reduce activity. Any fracture surfaces should be otected from the potential of rubbing and contact Laboratory Examination. While each type of failure may have unique tests, some general steps should be taken in investigating all corrosion failures All samples must be properly identified and their origin, handling, and processing within the laboratory Photographs should document samples in the "as-received"condition Stereomicroscopic examination of the involved areas should follow. Photographic documentation during stereomicroscopy should also be performed Nondestructive test methods should now be considered. The key to performing nondestructive testing is to avoid any disturbance of scale product until corrosion samples have been removed. Radiography to document casting quality or to evaluate cracking is usually acceptable. However, the use of liquids or dyes is not acceptable until the corrosion samples have been removed The removal of corrosion deposits for evaluation is the next step. The samples should be removed with a noncontaminating tool such as a stainless steel pick. Corrodent samples should be placed in clean, clearly marked containers The corrosion deposits should be analyzed. One of the most common analysis techniques is energy dispersive spectroscopy(EDS). This method, used in conjunction with scanning electron microscopy (SEM), provides information regarding the elemental composition of the corrosion deposit Based on the visual examination, a corrosion sample may need to be subjected to microbiological nal These steps may be followed by cleaning and/or other tests After the corrosion deposits selected for examination have been secured, the failure sample may be cleaned However, in some cases it may be desirable to retain the corrosion product intact for metallographic examination. For instance, a component subject to hydrogen damage or caustic corrosion may benefit from an analysis of the corrosion product layering effect. Cross sections should be taken prior to removal of corrosion p Precautions must be taken during the cleaning process to avoid any destruction to the base metal. Generally, the least aggressive cleaning method is initiated first, such as brushing with a soft brush or light air pressure Thefileisdownloadedfromwww.bzfxw.com
Laboratory photography in a controlled setting should also be performed to guarantee accurate reproduction of color tones and surface textures. The type and extent of on-site sampling depends on the environment and its availability. Piping corrosion in a domestic water service would require water samples from the source (incoming water supply) and from the end use (for example, a faucet). Microbiologically influenced corrosion may apply in certain cases. On-site testing or sample removal is necessary to retain the most accurate information regarding the type and number of bacteria involved. In most cases, removal of corrosion deposits from the surface may be performed in the field. If the sample is undergoing extensive laboratory examination, it may be prudent to carefully retain the corrosion in place for laboratory documentation and removal. If samples are removed on site, care must be taken to avoid any contamination. Sealed bags, latex gloves, tools for sample removal, buffered containers for water samples, MIC kits, and adhesive tape are useful for sample removal. (See the article “Biological Corrosion Failures” in this Volume for listings of commercially available kits.) The analyst should avoid touching any corrosion product directly with bare hands to prevent contamination. Physical removal of samples must be done in a manner to avoid further damage to the failed component and to avoid disturbance of corrosion products. In general, cutting must be done with care to avoid alterations of the metallurgical condition of the material and corrosion deposits. Saw cutting is generally preferred over torch cutting, since heating of the sample can affect the material and the corrosion product. If torch cutting is performed, a distance of 75 to 150 mm (3 to 6 in.) should be maintained from the area of interest. Saw cutting should be performed at a slow rate to avoid overheating. Use of lubricants and coolants should be avoided, if possible, to minimize contamination. The proper shipment and storage of samples lessens the possibility of destruction of pertinent evidence. Wrapping and sealing of the failed component will generally retain evidence and prevent further oxidation or contamination. A neutral environment may be required to reduce activity. Any fracture surfaces should be protected from the potential of rubbing and contact. Laboratory Examination. While each type of failure may have unique tests, some general steps should be taken in investigating all corrosion failures: · All samples must be properly identified and their origin, handling, and processing within the laboratory documented. · Photographs should document samples in the “as-received” condition. · Stereomicroscopic examination of the involved areas should follow. Photographic documentation during stereomicroscopy should also be performed. · Nondestructive test methods should now be considered. The key to performing nondestructive testing is to avoid any disturbance of scale product until corrosion samples have been removed. Radiography to document casting quality or to evaluate cracking is usually acceptable. However, the use of liquids or dyes is not acceptable until the corrosion samples have been removed. · The removal of corrosion deposits for evaluation is the next step. The samples should be removed with a noncontaminating tool such as a stainless steel pick. Corrodent samples should be placed in clean, clearly marked containers. · The corrosion deposits should be analyzed. One of the most common analysis techniques is energydispersive spectroscopy (EDS). This method, used in conjunction with scanning electron microscopy (SEM), provides information regarding the elemental composition of the corrosion deposit. · Based on the visual examination, a corrosion sample may need to be subjected to microbiological analysis. · These steps may be followed by cleaning and/or other tests. After the corrosion deposits selected for examination have been secured, the failure sample may be cleaned. However, in some cases it may be desirable to retain the corrosion product intact for metallographic examination. For instance, a component subject to hydrogen damage or caustic corrosion may benefit from an analysis of the corrosion product layering effect. Cross sections should be taken prior to removal of corrosion products. Precautions must be taken during the cleaning process to avoid any destruction to the base metal. Generally, the least aggressive cleaning method is initiated first, such as brushing with a soft brush or light air pressure. The file is downloaded from www.bzfxw.com
Ultrasonic cleaning in acetone is considered a nonaggressive approach. This cleaning method will remove some light surface deposits. Deposits that have been exposed to elevated temperatures will generally require a more aggressive approach. Inhibited dilute acid solutions and citric acid cleaners can be used to clean adherent corrosion deposits. In cases where the protection of the fracture features is critical, softened acetate tape can be used to remove adherent deposits. This method also retains the deposits removed for further examination if necessar If the failure analysis does not involve preservation of a fracture surface, fine sandblasting of the base metal may be useful to remove scale deposits as in the evaluation of the pitting After the corrosion deposits have been removed, additional nondestructive testing may prove useful. Magnetic particle examination, penetrant examination, eddy current testing, and ultrasonic testing are a few techniques that can be employed to explore the quality and condition of the failure and material Microscopic Examination. As noted previously, stereomicroscopy is performed to document the corrosion product, and it also is an appropriate tool for analysis of the fracture surface(if one exists)or other surface after cleaning. This analysis will provide information regarding the failure initiation, cracking, and corrosion-surface patterns such as pitting and wear/erosion. Photographic documentation of the surface should be performed Based on the results of the stereomicroscopy areas can be selected for sem to characterize the fracture features. A determination of the fracture features, such as intergranular cracking, cleavage, and ductile dimples will provide valuable information pointing to the likely cause of failure and the corrosion mechanism Metallography is an essential tool for the examination of the failure specimens. Selection of the most informative cross-section locations is important. Metallurgical examination of a cross section requires mounting, surface grinding, polishing, and examining the sample in the unetched and etched condition under a microscope. Microstructural features and conditions such as cracking and crack progression, pit morphology, selective leaching, surface features, and other characteristics can all be examined. These features provide key evidence regarding the cause of failure and extent of damage As discussed earlier, in some cases it is useful to examine a cross section of the corrosion product with the base material. There are various techniques to retain the scale without pullout during the polishing process. In some cases, plating over the scale prior to mounting the sample or impregnating the mount with resin after preparation can retain the scale for examination. Additional information is provided in the articles"Practices in Failure Analysis"and"Metallographic Techniques in Failure Analysis"in this Volume Corrosion testing includes several categories of tests. Normally, corrosion testing is considered a long-term pproach to investigations regarding material selection. Simulated testing or in situ testing is often performed when a given environment may be unique or where materials may experience unique flow conditions. For ample, piping containing fluid may experience unique flow conditions that may result in erosion-corrosio failures. In that case, sections of pipe of different alloys may be placed on line and monitored. Other common methods of testing include accelerated tests, simulated or pilot testing, and electrochemical tests Typically, a variety of materials are selected for evaluation. A standard, such as carbon steel, may be used to verify the corrosiveness of the environment. Multiple samples of the same material should be tested to verify reproducibility. Sources for corrosion test methods are ASTM International, NACE, and internationally, ISo, EN, and JIS standards. Other testing methods and inspection procedures have been developed on an industry specific basis by industrial and government organizations, such as in the pulp and paper and the electric power industries, to specifically address corrosion issues relating to their specific operating conditions and environments Accelerated tests are performed when an expedited answer is needed to solve a serious corrosion problem and demonstrate an acceptable corrosion-resisting lifetime, provided that correlations can be established Accelerated testing usually involves the exposure to extreme conditions, relative to the actual servie environment,to accelerate the corrosion process. The process may be accelerated with high-stress conditions, elevated temperatures, or highly aggressive solutions While accelerated testing does provide test data sooner, extrapolation of results may be misleading. The acceleration of certain factors during the testing may cause different corrosion mechanisms to become active, and thus the results may not reflect those that would be obtained under actual service conditions. Care must be taken when evaluating results so that economical and acceptable choices are not eliminated due to their rformance in unrealistic environments Simulated-use tests are primarily set up in the laboratory with an environment as close as possible to the environment of interest. In these cases, sufficient time must be allowed for an appropriate response; a short time
Ultrasonic cleaning in acetone is considered a nonaggressive approach. This cleaning method will remove some light surface deposits. Deposits that have been exposed to elevated temperatures will generally require a more aggressive approach. Inhibited dilute acid solutions and citric acid cleaners can be used to clean adherent corrosion deposits. In cases where the protection of the fracture features is critical, softened acetate tape can be used to remove adherent deposits. This method also retains the deposits removed for further examination if necessary. If the failure analysis does not involve preservation of a fracture surface, fine sandblasting of the base metal may be useful to remove scale deposits as in the evaluation of the pitting. After the corrosion deposits have been removed, additional nondestructive testing may prove useful. Magnetic particle examination, penetrant examination, eddy current testing, and ultrasonic testing are a few techniques that can be employed to explore the quality and condition of the failure and material. Microscopic Examination. As noted previously, stereomicroscopy is performed to document the corrosion product, and it also is an appropriate tool for analysis of the fracture surface (if one exists) or other surface after cleaning. This analysis will provide information regarding the failure initiation, cracking, and corrosion-surface patterns such as pitting and wear/erosion. Photographic documentation of the surface should be performed. Based on the results of the stereomicroscopy, areas can be selected for SEM to characterize the fracture features. A determination of the fracture features, such as intergranular cracking, cleavage, and ductile dimples, will provide valuable information pointing to the likely cause of failure and the corrosion mechanism. Metallography is an essential tool for the examination of the failure specimens. Selection of the most informative cross-section locations is important. Metallurgical examination of a cross section requires mounting, surface grinding, polishing, and examining the sample in the unetched and etched condition under a microscope. Microstructural features and conditions such as cracking and crack progression, pit morphology, selective leaching, surface features, and other characteristics can all be examined. These features provide key evidence regarding the cause of failure and extent of damage. As discussed earlier, in some cases it is useful to examine a cross section of the corrosion product with the base material. There are various techniques to retain the scale without pullout during the polishing process. In some cases, plating over the scale prior to mounting the sample or impregnating the mount with resin after preparation can retain the scale for examination. Additional information is provided in the articles “Practices in Failure Analysis” and “Metallographic Techniques in Failure Analysis” in this Volume. Corrosion testing includes several categories of tests. Normally, corrosion testing is considered a long-term approach to investigations regarding material selection. Simulated testing or in situ testing is often performed when a given environment may be unique or where materials may experience unique flow conditions. For example, piping containing fluid may experience unique flow conditions that may result in erosion-corrosion failures. In that case, sections of pipe of different alloys may be placed on line and monitored. Other common methods of testing include accelerated tests, simulated or pilot testing, and electrochemical tests. Typically, a variety of materials are selected for evaluation. A standard, such as carbon steel, may be used to verify the corrosiveness of the environment. Multiple samples of the same material should be tested to verify reproducibility. Sources for corrosion test methods are ASTM International, NACE, and internationally, ISO, EN, and JIS standards. Other testing methods and inspection procedures have been developed on an industryspecific basis by industrial and government organizations, such as in the pulp and paper and the electric power industries, to specifically address corrosion issues relating to their specific operating conditions and environments. Accelerated tests are performed when an expedited answer is needed to solve a serious corrosion problem and to demonstrate an acceptable corrosion-resisting lifetime, provided that correlations can be established. Accelerated testing usually involves the exposure to extreme conditions, relative to the actual service environment, to accelerate the corrosion process. The process may be accelerated with high-stress conditions, elevated temperatures, or highly aggressive solutions. While accelerated testing does provide test data sooner, extrapolation of results may be misleading. The acceleration of certain factors during the testing may cause different corrosion mechanisms to become active, and thus the results may not reflect those that would be obtained under actual service conditions. Care must be taken when evaluating results so that economical and acceptable choices are not eliminated due to their performance in unrealistic environments. Simulated-use tests are primarily set up in the laboratory with an environment as close as possible to the environment of interest. In these cases, sufficient time must be allowed for an appropriate response; a short time
exposure likely will not produce enough corrosion activity to evaluate the corrosion behavior of the material. Nevertheless, by simulating the exact conditions of operation, an accurate assessment of the material can generally be produced with an adequate exposure time ASTM publishes standard test methods and analytical procedures for corrosion and wear testing (Ref 2) Electrochemical testing is performed for general information regarding the passivity or anodic protection of material against corrosion and to determine the critical breakdown or pitting potential. This type of testing is performed by two methods: controlling the current or controlling the potential. Standard methods for electrochemical testing are published in Ref 2 As the name suggests in the controlled-current test method the current is controlled and the resulting corrosion potential is measured. Polarization curves are generated. Galvanostatic and galvanodynamic polarization measurements are used to plot anodic and cathodic polarization curves. The assumption that corrosion rates remain constant with time can produce inaccurate results with this test method In the controlled-potential method, instrumentation is available for both constant-potential(potentiostatic)and variable-potential(potentiodynamic) testing to determine overall corrosion-rate profiles for metal-electrolyte stems over a range of potentials Corrosion Rates and Types. Knowing and understanding corrosion rates and the types of corrosion is essential to the evaluation of corrosion failures and to the communication of the results to others the articles that follow in this Section discuss the forms mechanisms and relative rates of corrosion. Corrosion Volume 13 of the ASM Handbook, provides detailed information regarding types of corrosion, corrosion testing, corrosion failures, and industry- and alloy-Specific corrosion considerations. Volume 13 also provides information regarding the use of specific alloys in given environments, corrosion prevention, and the use of nonmetallic materials. The Handbook of Corrosion Data, 2nd edition(ASM International, 1995)is a compilation of corrosion data from published sources. Corrosion rates of various alloys are provided with a general discussion of the corrosion resistance of alloy groups in particular environments Internet web sites published by ASM International, ASTM, NACE, the Nickel Development Institute, and the Copper Development Association provide the ability to search libraries of data for a given request Analysis of Incomplete Data. Incomplete or inconsistent data may occur in certain instances when the failure piece has been contaminated by an unknown source. Improper handling of the failure sample can introduce contamination on the sample. Testing of contaminated samples may produce misleading data and erroneous results. For example, the sampling of a deposit removed from a fracture surface that experienced stress corrosion may not reveal the corrodent that caused the scc. often the fracture surface is flushed with water or cleaned prior to testing. Liquid penetrants, cleaning fluids, cutting fluids, and solvents may alter the chemical composition of the surface deposits Results from laboratory testing may provide extraneous results. The tests may not model the service conditions Laboratory testing cannot easily model flow conditions such as turbulence, erosion, and localized attack. Care must be taken when evaluating the laboratory data to be certain that the conclusions drawn are an accurate assessment of the operating environment ASTMG 16, "Applying Statistics to Analysis of Corrosion Data"(Ref 3), provides a guide for handling data; it refers to ASTM E 178, Practice for Dealing with Outlying Observations"(Ref 4) for the treatment of data that appear inconsistent with the bulk of the findings References cited in this section 2. Wear and Corrosion, Vol 03.02, Annual Book of AsTM Standards, ASTM 3.Applying Statistics to Analysis of Corrosion Data, G 16, Annual Book of ASTM Standards, ASTM 4."Practice for Dealing with Outlying Observations, " E 178, Annual Book of ASTM Standards, ASTM Thefileisdownloadedfromwww.bzfxw.com
exposure likely will not produce enough corrosion activity to evaluate the corrosion behavior of the material. Nevertheless, by simulating the exact conditions of operation, an accurate assessment of the material can generally be produced with an adequate exposure time. ASTM publishes standard test methods and analytical procedures for corrosion and wear testing (Ref 2). Electrochemical testing is performed for general information regarding the passivity or anodic protection of a material against corrosion and to determine the critical breakdown or pitting potential. This type of testing is performed by two methods: controlling the current or controlling the potential. Standard methods for electrochemical testing are published in Ref 2. As the name suggests, in the controlled-current test method, the current is controlled and the resulting corrosion potential is measured. Polarization curves are generated. Galvanostatic and galvanodynamic polarization measurements are used to plot anodic and cathodic polarization curves. The assumption that corrosion rates remain constant with time can produce inaccurate results with this test method. In the controlled-potential method, instrumentation is available for both constant-potential (potentiostatic) and variable-potential (potentiodynamic) testing to determine overall corrosion-rate profiles for metal-electrolyte systems over a range of potentials. Corrosion Rates and Types. Knowing and understanding corrosion rates and the types of corrosion is essential to the evaluation of corrosion failures and to the communication of the results to others. The articles that follow in this Section discuss the forms, mechanisms, and relative rates of corrosion. Corrosion, Volume 13 of the ASM Handbook, provides detailed information regarding types of corrosion, corrosion testing, corrosion failures, and industry- and alloy-specific corrosion considerations. Volume 13 also provides information regarding the use of specific alloys in given environments, corrosion prevention, and the use of nonmetallic materials. The Handbook of Corrosion Data, 2nd edition (ASM International, 1995) is a compilation of corrosion data from published sources. Corrosion rates of various alloys are provided with a general discussion of the corrosion resistance of alloy groups in particular environments. Internet web sites published by ASM International, ASTM, NACE, the Nickel Development Institute, and the Copper Development Association provide the ability to search libraries of data for a given request. Analysis of Incomplete Data. Incomplete or inconsistent data may occur in certain instances when the failure piece has been contaminated by an unknown source. Improper handling of the failure sample can introduce contamination on the sample. Testing of contaminated samples may produce misleading data and erroneous results. For example, the sampling of a deposit removed from a fracture surface that experienced stress corrosion may not reveal the corrodent that caused the SCC. Often the fracture surface is flushed with water or cleaned prior to testing. Liquid penetrants, cleaning fluids, cutting fluids, and solvents may alter the chemical composition of the surface deposits. Results from laboratory testing may provide extraneous results. The tests may not model the service conditions. Laboratory testing cannot easily model flow conditions such as turbulence, erosion, and localized attack. Care must be taken when evaluating the laboratory data to be certain that the conclusions drawn are an accurate assessment of the operating environment. ASTM G 16, “Applying Statistics to Analysis of Corrosion Data” (Ref 3), provides a guide for handling data; it refers to ASTM E 178, “Practice for Dealing with Outlying Observations” (Ref 4) for the treatment of data that appear inconsistent with the bulk of the findings. References cited in this section 2. Wear and Corrosion, Vol 03.02, Annual Book of ASTM Standards, ASTM 3. “Applying Statistics to Analysis of Corrosion Data,” G 16, Annual Book of ASTM Standards, ASTM 4. “Practice for Dealing with Outlying Observations,” E 178, Annual Book of ASTM Standards, ASTM The file is downloaded from www.bzfxw.com
Analysis and Prevention of Corrosion-Related Failures S.R. Freeman, Millennium Metallurgy Ltd Examples of Corrosion Failure Analy Example 1: Analysis of Pitting and MIC of Stainless Steel Piping. Type S31603)austenitic stainless steel piping was installed as part of a collection system for a storm nt system used in a manufacturing facility. Within six months of start-up, leaks were discovered On-Site Examination. A of tests were performed on site to eliminate stray currents as a possible cause of failure in the piping system. The piping system was inspected for possible sources of ac or dc current flow.An external battery was used to impress a voltage. Voltage and current were recorded with the external power supply applied and removed There was no evidence of stray currents or electrical discharge It was noted that the ambient temperature and slow or stagnant flow conditions present in the piping were ideal for bacteria growth On-Site Sampling. Water samples were obtained for corrosivity and MIC testing. Commercially available field MIC kits were used. Samples of the damaged pipe were removed total dissolved solids level of 10,000 ppm was also high. The MIC water samples were shipped to the laboratory within 24 h of removal for viable culture testing. The testing showed high levels of aerobic, acid producing, and low-nutrient bacteria Laboratory Examination. Perforated pipe samples were provided for metallurgical evaluation. Figure 2(a) shows the leak area as viewed from the outside-diameter surface. The sample was cut dry to avoid contamination. The inside-diameter surface is shown in Fig. 2(b). The pit appeared larger on the inside diameter surface, indicating pit initiation occurred at the inside surface and at the bottom of the pipe. a rusty discoloration was apparent along the bottom length of the pipe. This discoloration corresponded to the area of the pitting. There was no corrosion deposit associated with the pitting. The discolored areas and other areas were evaluated using EDs. The EDS revealed contaminants consisting of chlorine, sulfur, sodium, silicon, and potassium. The area of discoloration revealed iron and oxygen only Fig. 2 Pitting corrosion of 316L stainless steel pipe. (a) view of pitting on the outside- diameter surface at the leak location.(b) View of the inside-diameter surface where the pit size was larger at the leak location. There was a rusty discoloration along the bottom of the pipe. (c) Cross section of pipe wall through the perforation. The sample was etched in AstM 89 reagent to delineate the microstructure. Uniform wall thickness is approximately 2.9 mm(0.1l in. ) 5x Courtesy of s.R. Freeman, Millennium Metallurgy, Ltd A metallurgical cross section was prepared through the pitted region. Figure 2(c)shows a 10X magnification of the cross section through the pitted region after etching with AStM 89 reagent. The pit was not associated with a welded region, and the microstructure appeared normal. There was no evidence of general wall loss(uniform corrosion)
Analysis and Prevention of Corrosion-Related Failures S.R. Freeman, Millennium Metallurgy, Ltd. Examples of Corrosion Failure Analysis Example 1: Analysis of Pitting and MIC of Stainless Steel Piping. Type 316L (UNS S31603) austenitic stainless steel piping was installed as part of a collection system for a storm sewer treatment system used in a manufacturing facility. Within six months of start-up, leaks were discovered. On-Site Examination. A series of tests were performed on site to eliminate stray currents as a possible cause of failure in the piping system. The piping system was inspected for possible sources of ac or dc current flow. An external battery was used to impress a voltage. Voltage and current were recorded with the external power supply applied and removed. There was no evidence of stray currents or electrical discharge. It was noted that the ambient temperature and slow or stagnant flow conditions present in the piping were ideal for bacteria growth. On-Site Sampling. Water samples were obtained for corrosivity and MIC testing. Commercially available field MIC kits were used. Samples of the damaged pipe were removed. Laboratory Testing. High levels of chlorides, as high as 20,000 ppm, were reported in the water sample. The total dissolved solids level of 10,000 ppm was also high. The MIC water samples were shipped to the laboratory within 24 h of removal for viable culture testing. The testing showed high levels of aerobic, acidproducing, and low-nutrient bacteria. Laboratory Examination. Perforated pipe samples were provided for metallurgical evaluation. Figure 2(a) shows the leak area as viewed from the outside-diameter surface. The sample was cut dry to avoid contamination. The inside-diameter surface is shown in Fig. 2(b). The pit appeared larger on the insidediameter surface, indicating pit initiation occurred at the inside surface and at the bottom of the pipe. A rusty discoloration was apparent along the bottom length of the pipe. This discoloration corresponded to the area of the pitting. There was no corrosion deposit associated with the pitting. The discolored areas and other areas were evaluated using EDS. The EDS revealed contaminants consisting of chlorine, sulfur, sodium, silicon, and potassium. The area of discoloration revealed iron and oxygen only. Fig. 2 Pitting corrosion of 316L stainless steel pipe. (a) View of pitting on the outsidediameter surface at the leak location. (b) View of the inside-diameter surface, where the pit size was larger at the leak location. There was a rusty discoloration along the bottom of the pipe. (c) Cross section of pipe wall through the perforation. The sample was etched in ASTM 89 reagent to delineate the microstructure. Uniform wall thickness is approximately 2.9 mm (0.11 in.). 5×. Courtesy of S.R. Freeman, Millennium Metallurgy, Ltd. A metallurgical cross section was prepared through the pitted region. Figure 2(c) shows a 10× magnification of the cross section through the pitted region after etching with ASTM 89 reagent. The pit was not associated with a welded region, and the microstructure appeared normal. There was no evidence of general wall loss (uniform corrosion)
Conclusion. The pitting in the austenitic stainless steel pipe is believed to be caused by damage to the passive layer brought about by a combination of MIC, high chloride levels, and high total dissolved solids. The low flow and stagnant conditions present in the piping are primary contributors to the pit progression. This type of material will perform significantly better when there is more flow of water Retesting of the water indicated similar high levels of aerobic bacteria, so these must be considered a part of th design basis Recommendations. Due to the extensive amount of piping, the pinhole size of the leaks, and environmental consequences of leaks, repair of the existing pitted pipe was impractical. Replacement of the pipe was recommended Several alloys, nonmetallic materials, and lined materials were proposed for coupon testing to determine which is the best in this particular environment Example 2: Analysis of a Corrosion Failure of an Aboveground Storage Tank. A failure of an aboveground storage tank occurred due to external corrosion of the tank floor. The liquid asphalt tank operated at elevated temperatures(approximately 177C, or 350F)and had been in service for six years. Cathodic protection (rectifiers) had been installed since start-up of the tank operation. It was noted, however, that some operational problems with the rectifier may have interrupted its protection On-Site Examination. The underside of the floor plates showed extensive localized wall thinning. Figure 3 shows a pit in one of the tank floor plates. While it was expected that the ground under the tank was near operational temperatures, it was determined that the temperatures were below 104C(220F)at most areas evaluated in the ground. Thus, it was possible for moisture to accumulate at the tank floor, producing intermittent wetting and drying conditions. This provided an explanation of the circumferential corrosion attern observed several feet in from the periphery of the tank. Heating was provided near the center of the tank and water in the backfill below the tank floor should be minimal. Perforations of the tank were large, and thinning was apparent in the adjacent areas PIT Fig3 View of an isolated pit on the outside of the floor tank of a liquid asphalt tank. the asphalt had been leaking for some time as the surrounding area was covered with deposits and the wall thinning was significant. Courtesy of s.R. Freeman, Millennium Metallurgy, Ltd On-Site Sampling Soil samples were taken and tested, but did not indicate high corrosivity. Sample of the scale deposits from the outer surface of the tank floor were taken Laboratory Examination. Scale samples analyzed by EDS showed the deposits consisted of primarily Fe203 (iron oxide). The rapid exfoliation, scale buildup, and pitting of the tank floor from the external (underside) surface also prompted MIC testing of the soil and of the deposits in the area of wall thinning. Although the temperatures were somewhat high for MIC to thrive, the results of the samples confirmed high levels of various Thefileisdownloadedfromwww.bzfxw.com
Conclusion. The pitting in the austenitic stainless steel pipe is believed to be caused by damage to the passive layer brought about by a combination of MIC, high chloride levels, and high total dissolved solids. The lowflow and stagnant conditions present in the piping are primary contributors to the pit progression. This type of material will perform significantly better when there is more flow of water. Retesting of the water indicated similar high levels of aerobic bacteria, so these must be considered a part of the design basis. Recommendations. Due to the extensive amount of piping, the pinhole size of the leaks, and environmental consequences of leaks, repair of the existing pitted pipe was impractical. Replacement of the pipe was recommended. Several alloys, nonmetallic materials, and lined materials were proposed for coupon testing to determine which is the best in this particular environment. Example 2: Analysis of a Corrosion Failure of an Aboveground Storage Tank. A failure of an aboveground storage tank occurred due to external corrosion of the tank floor. The liquid asphalt tank operated at elevated temperatures (approximately 177 °C, or 350 °F) and had been in service for six years. Cathodic protection (rectifiers) had been installed since start-up of the tank operation. It was noted, however, that some operational problems with the rectifier may have interrupted its protection. On-Site Examination. The underside of the floor plates showed extensive localized wall thinning. Figure 3 shows a pit in one of the tank floor plates. While it was expected that the ground under the tank was near operational temperatures, it was determined that the temperatures were below 104 °C (220 °F) at most areas evaluated in the ground. Thus, it was possible for moisture to accumulate at the tank floor, producing intermittent wetting and drying conditions. This provided an explanation of the circumferential corrosion pattern observed several feet in from the periphery of the tank. Heating was provided near the center of the tank and water in the backfill below the tank floor should be minimal. Perforations of the tank were large, and thinning was apparent in the adjacent areas. Fig. 3 View of an isolated pit on the outside of the floor tank of a liquid asphalt tank. The asphalt had been leaking for some time as the surrounding area was covered with deposits and the wall thinning was significant. Courtesy of S.R. Freeman, Millennium Metallurgy, Ltd. On-Site Sampling. Soil samples were taken and tested, but did not indicate high corrosivity. Sample of the scale deposits from the outer surface of the tank floor were taken. Laboratory Examination. Scale samples analyzed by EDS showed the deposits consisted of primarily Fe2O3 (iron oxide). The rapid exfoliation, scale buildup, and pitting of the tank floor from the external (underside) surface also prompted MIC testing of the soil and of the deposits in the area of wall thinning. Although the temperatures were somewhat high for MIC to thrive, the results of the samples confirmed high levels of various The file is downloaded from www.bzfxw.com