4 Introduction 4.1 Importance of corrasion prevention and control Corrosion is serious point of con- cetn in many (petro)chemical process plants. It represents the difference between trouble-free operation and costly downtime, But corrosion is an unknown quantity only to those who do not attempt to understand it. Since corrosion is impractical to eliminate Im al cases, the secret of effective ‘engineering generally les in control- ling rather than preventing it. Prefer ably this has already to be done in the design stage of a process. To avoid corrosion totally one should choose sometimes very expensive materials of construction resulting in high investment costs. Low invest- ment can result in high maintenance costs, To realize very low maintenance costs generally one should choose expensive materials. All metals and alloys are susceptible to corrosion. No single material is suitable for all ap- plications. For example, gold, known for its excellent resistance to the atmosphere, will corrade if exposed to mercury at amblent temperature. On the other hand, iron does not corrode in mercury but “rusts” readily in the atmosphere, In most cases howev ‘more corrosion resistant materials are also more expensive. The choice of materials of construc: ton is alway$ a compromise between, investment and maintenance costs Fortunately, usually there are several ‘materials that will perform satisfac torily in a given environment. Also, there are corrosion control methods to minimize the problem. C problems do arise after the plant has, been built and is operating. It is up to the process engineer and/or the ‘mechanical engineer (in co-operation with a corrosion engineer) to control corrosion. He should know where to start and have a basic knowledge of: + How to recognize corrosion; in which form (morphology) it may appear + How it is caused; mechanism of cor rosion failure modes, + How to measure its severity, * Inspection techniques, * What tools ace available to measure it. + Design variables affecting corrosion. * Materials selection. * Cortosion prevention methods, + How to interpret and apply corro- sion information. + Where to obtain help, ‘The objective of this cerrosion ‘engineering guide is to provide this knowledge. By definition, corrosion is the (unde- sirable) deterioration of materials of construction due to chemical or elec- trochemical influences, initiating from the interface material / environment. Corrosion occurs in several, widely di. fering forms, Classification is usually based on one of three factors: 1. Nature of the corrosive agent - Cor. rosion can be classified as "wet" or “dry”. A liquid or moisture is neces- sary for the former and dry corro: sion usually involves reaction with high temperature gases. 2, Appearance of the corroded metal ~The morphology of the corrosion phenomenon. Corrosion is eitherIntRoDucTION uniform and the metal corrodes at the same rate over its entire surface; ‘or (mostly) itis localized, in which ‘ase only small areas are affected, 3. Mechanism of corrosion ~ This involves either electrochemical or direct chemical reactions or in some ‘cases physical-metallurgical deterio- ration of the metal, Plant shutdowns can and do occur as 2 result of corrosion. This is just one of its many direct and indirect economic consequences. Some are economic, ‘other consequences are social or SHE. (Gafety, health and environment) related, Economic consequences are + Repair or replacement of corroded equipment. + Over-design to allow for corrosion, + Preventive (time based) or predictive (condition based) maintenance, e.g painting, + Shutdown of equipment, due to cor- rosion failure. + Contamination of a product. + Catalyst poisoning, * Loss of efficiency, e.g, over-design and corrosion products decrease the heat-transfer rate in heat exchangers, + Loss of valuable products, eg. ftom a reactor that has corraded through, + Damage of equipment adjacent to that in which corrosion failure oc- curs. Its difficult, and impossible when the Joss of life is considered, to accurately calculate the costs of corrosion ane protection, but it is estimated that in industrialized countries these costs amount up to 2 to 3.5% of the gross national product (according to the English Committee Corrosion and Pro. tection, chairman Dr. TP. Hoar). The sgtoss national product of the Nether- lands in 2007 was about 500 billion Euro, indicating costs due to corro- sion of 10 to 18 billion Euro, Some Of these direct costs are unavoidable such as those due to painting or other protective measures. However, it has become clear that taking preventive ‘measures is in most cases cheaper than. just letting it run its course. It has also been estimated that about 25% of the corrosion costs could be saved by the application of existing knowledge of corrosion prevention and control, 4.< by using better protective systems, by improving design and materials selec- tion. Beside the direct costs, ie. those due to higher investment, inspection, ‘monitoring, maintenance and repait | costs in many cases indirect costs are at stake lke costs due to production or product loss. These costs are some- times even higher than direct costs | and more difficult to estimate. According to the before mentioned English committee in a (petro)chemi- cal industry the (direct + indirect) costs of corrosion control and protec- tion can rise up to about 5 % of the turnover. ‘The turnover of DSM in 2007 globally ‘was near to 9 billion Euto, indicating direct and indirect costs of corrosion and corrosion control of about 450 million Euro, In chemical plants the maintenance | budget is about 2 to 4% of the re- investment value of the process plant. According to the DSM maintenance ‘organizations up to about 10 to 15% of this maintenance budget is applied for comrosion prevention and control | OF course also in other branches of industry, ike marine, power and transport high corrosion costs ace at stake. One can imagine that replace ment of exhaust systems of cars (about once in every 3 or 4 years) do cost an extremely lot of money, notwithstand- ing the fact that we have the techni- = cal possibility to extend the IMetime of exhaust systems extensively by ‘minor additional costs. An anecdote out of the USA in the early seventies is illustrative; this anecdote tells that the annual costs of replacing exhaust = systems of cars in the US does cost more than did cost the digging of the | Panama canal Apart from the immediate economic or technical consequences, corrosion ‘can also have serious SHE consequenc- es. Perforation of an underground pipeline, used for transport of e.g, organic product, may lead to extensive damage to the environment. Corro- sion is also a serious problem because it contributes to the depletion of our ‘natural resources. All over the world every 90 seconds one thousand of kg steel is transformed to rust. It should >be mentioned that corrosion controt also contributes to the conservation of energy. The energy costs to produce one thousand kg steel equals to the | energy consumption of a family of 4 Persons in a period of 3 months. SHE and social consequences are: + Safety, e.g. sudden failure may cause fire, explosion, release of toxic prod- uct, construction collapse, Health, e.g. pollution due to escap- ing product from corroded equip: ment, | + Depletion of natural resources -met als and the fuels to manufacture them, * Appearance (aesthetic or cosmetic as- pects) ~ corroded material is usually unpleasing to the eye. Of course, all the preceding social items do have economic aspects also, | Clearly, there are many reasons for wanting to avoid corrosion. However, sometimes weathering steels or other alloys are applied to be rusted or cor roded on purpose for aesthetic reasons like front panels, statues, cathedral | toofs (copper alloys) Discussing these economical, SHE and social consequences of corrosion itis, evident that corrosion engineering is a very important aspect in the control- © lability of a process plant. Corrosion engineering comprises the following activities: +» Advice regarding corrosion resistant ‘materials of construction, design of equipment and design of corrosion, protection systems. + Inspection of equipment / Risk Based Inspection (RBD, + Trouble shooting. + Study of corrosion phenomena, + Study of inhubitor systems, * Draw up of material specifications and standards, = Activation and evaluation of development of new materials of construction and protection systems (inhibitors, coatings, ete). * Development and evaluation of cor- rosion measuring and monitoring techniques. * Presentation of corrosion courses and training. + Improve equipment reliability (IER) by means of equipment condition ‘monitoring (ECM) and life time extension (LTE) studies, ‘The ECM philosophy is gaining more | and mote acceptance as an important tool to IER and is discussed in next chapter,3s plants normally are designed ra particular service life, mostly for wut twenty years. A plant is expected “offer 100 percent rellablity without ‘maintenance other than of those ts where wear or corrosion is anti fated and accepted. In practice, how- , the life of a plant and its compo- lents can be quite checkered, While lure mechanisms may be identified accepted in the design phase, they ‘may in practice take a different course ‘than originally anticipated. Indeed, ‘plant management may be confronted ‘with unforeseen failure mechanisms ‘within the expected service life. Also, as a plant nears the end of its de- sigh life, management often wonders if its mechanical integrity might allow ‘operation to be continued for another fifteen years or so. For these reasons, plant asset management is becoming Increasingly important for profitable and competitive plant operation. High plant integrity and availability and low costs are of paramount impor- tance. These three parameters are interdependent and can be controlled only if the underlying processes, i. the failure mechanisms and preven- tive measures, are fully controlled and predictable (Figure 1). One has to be aware of potential casualties, due to failure modes and an inventory has to be made of the performance killers and the cost drivers in the plant, The costs of maintenance and inspection should be balanced with the profits ‘of maintenance. [tis necessary to per- forma life cycle cost (LCC) analysis Life cycle costs are the integral costs of: + maintenance + non-destructive testing + visual inspections + condition monitoring + (un)planned shutdowns + repair costs + warehousing of spare parts , Improved equipment reliability by means of equipment condition monitoring Availability Integrity Controllability = Reliability = Predictability Costs Figure 1. Asset management of process plant. = Perlodic inspections are useful for equipment condition monitoring and detecting trends in failures, so en- abling proper asset management. The ‘crucial questions here are: what needs to be inspected, how and when? How ccan the greatest possible benefits be achieved for the least possible effort? Inspections are looked at more and ‘mote critically now that low main- tenance costs are increasingly key to plant profitability. The aim is to © cut back on both the direct costs of Inspections and the indirect costs such ‘as production outage. Today, Risk Based Inspection (RBI is hatled as one © way of dealing more selectively with Inspections. RBIs a technique for optimizing in- spection activities based on a critical- lty rating of items of equipment. It determines the criticality of each item, iticality is a combination of the failure probability and the effects of failure, The effects of failure may refer to both the economic consequences, such as plant outage and repairs, as well as safety, health and environmen- tal effects, such as injuries and pollu- 5 tion, Of course, items of equipment ‘with a high criticality ranking call for | more attention than those with a low © criticality ranking. Thus, maintenance 5 can be prioritized on the basis of the rankings. At fist sight, the rankirg technique may seem quite straightforward, but is not so easy in actual practice. It is not | easy to accurately estimate the prob- ability of each potential failure mecha- nism, Things become even more © complex when plant management {wants not only relative rankings but also concrete figures, for instance for © the optimum time inteival between inspections, These intervals must be neither too short nor sc long as to allow the condition to deteriorate to the point where corrective action is rno longer justified or were the item falls. Consequently, the evaluation requires personnel with thorough practical know-how and experience. DSM‘: former Corrosion Engineering department has developed an evalu- ation technique in which this know- how and experience is embedded and updated to include developments as they take place. This technique is, ‘ealled ECM. It is schematically shown, in Figure 2. Basically, ECM Isa *Plan-Do-Study- “Act” (PDSA) cycle as commonly used in Total Quality Management, The Plan phase concems the preparation of inspection and monioring pro ‘grammes on the basis oa criticality ranking depending on the failure mechanisms identified. The Do phase covers the implementation of theINTRODUCTION Figure 2. Equipment Condition Monitoring programmes and recording of observa- tions and failures. The Study phase comprises the evaluation and analysis Of the results of the programmes. The ‘Act phase consists in adding fresh insights to the inspection and mont- toring programmes and to the docu- ‘ments, or databases, on which they are based. At the plant level, these documents are the Plant Integrity Handbook or Plant integrity Database and at corporate level the Corporate Standards and Practices, ECM isa dynamic learning system allowing experience to be retained, documented and utilized also else- where in the organization, 1 Literature ‘Matarals and corrasion expertise at DSM; Mechanical Pant Services offers @ ming of information. Interview with Gil Noten. Sjef Roymens Stainlogs Stoel Europe. Vol lume 8, Issue 4, May 1993: poga 44.29, 2. OSM TechnoPartners; Corrosion consul taney with @ ene Iouerviw wit Gil Notten and Jacko ‘Aerts. Michael van Wingoarden Stainless Steel World, Valume 15, Octo- ber 2003, page 32.36, Equipment Condition Monitoring (ECM) ‘System 2s an Inspectan Toc! for (Petro) Chemical Process Pl, 6. Noten, J. Boogaard, A. Krigsman, 1140 inernasional Corasion Congress; (Cooperation in Corresian Cantro. Cape Tonn, South Africa: 26 Sepuembar 10 01 October 1999,juction 1 Introduction Before discussing the several forms of yesion more in detail itis neces- sary to introduce the modern electro~ ‘Chemical theory of aqueous corrosion. laithiough the presentation is necessar- ily simplified and brief itis thought this theory will contribute to a better ‘understanding of corrosion phenom. cena, First will be approached ther- ‘modynamic principles and electrode kinetics. Afterwards will be discussed applications of these principles. 2.2 Thermodynamics (the science of energy change) Change of free energy SG ‘Most metals exist in their natural state as compounds (oxides, sulfides ‘etc.). This is their thermodynamically stable state. We dig these compounds out of the ground, extract the metal (by lifting it to a higher energy state) and then make it into useful articles. Unless we have substantially changed the metal (eg. by alloying), the ther- ‘modynamics will not have changed, ‘When we expose this metal article to a natural environment, it will (If not protected) revert to its natural state by corrosion, Of course, metals that exist In nature as pure elements (e.g. gold) have excellent corrosion resistance in natural environments. Corrosion is only possible if its products are more stable than the reactants, Most corro- sion reactions are electrochemical and we will restrict our discussion to these. ‘The change in free energy AG is a Electrochemistry direct measure of the work capacity or maximum electric energy available from the system. If the change in free energy during transition of a system from one state to another is negative, this indicates a loss in free energy and indicates also the spontaneous reaction direction of the system. That means, if no external forces act on the system, the system will tend to trans- orm to its lowest energy state. Ifthe ‘change In free energy is positive, this Indicates that the transition represents! © an increase in energy, and this requires that additional energy be added to the system. This principle can be illustrat- ed by a mechanical analogy as shown in figure 1a, Ifthe ball moves from po- sition 1 to position 2, this represents a ‘decrease in free energy. The transition from position 1 to position 2 is the spontaneous direction of this particu Jar system. The reverse transformation, © from postion 2to postion 1, isnot a spontaneous direction and requires, = the application of energy. ‘The change in free energy is indepen- dent of the reaction path. This is also illustrated in figure Tb In this figure ‘two possible reaction paths A and B are indicated. For either path, free energy change for the transition from state 2A or 2B is exactly the same. It is obvious, however, that the transfor- mation along path B will sequire more time and will be slower than along path A. Chemical and electrochemt- cal corrosion reactions behave in the same way, It is not possible to predict the velocity of a reaction from the change in free energy. This thermo- | Spontaneous As rection ? Ac ‘Spontaneous direction Figure 1a (left) ‘Mechanical analogy of fee energy change. Figure 1b. right) Effect of reaction path on reaction rate Fontans & Greene ‘dynamic parameter only indicates the direction ofa reactior by its sign, and any predictions of velocity may be wrong, ‘The free energy change accompany- ing an electrochemicel reaction can be calculated by the following equa AG = -nFE AG = free energy change. n number of electrons involved in the reaction. E Faraday constant (8,65.10' Coulomb per (g)mo) E Ee- Ex AG = -nE(Ec- Es) F represents the total electric charge of Nelectrons. Nis the number of Avogadro (number (of atoms per (g)mol cf element) = 6.03, x 10% (ghmol'. Since the charge of 1 electron e = 1.6 x 10 coulomb, F=Nxe=9.65 x 10°C (g)mol" "The mass of 1 atom H is 1,008/N = 1.67 x 10% gram = I dalton. Cell potentials and EMF-series For a good understanding of electro- chemical corrosion ore should have a correct comprehension of the elec- trochemical phenomena acting in a galvanic cell. Ifa solid piece of metal {s placed in a solution containing its ‘metal ions, a potential difference, E | [Me/Me'] exists between the metal ‘atoms in the solid and the metal ions in solution, the half cell potential (Figure 2), At the interface of metal / electrolyte | following processes occur A: Me-> Me® + ne; at certain points fon the metal surface, metal atoms are oxidized to metal ions leaving nega- tively charged electrons in the metal electrode. 1B Me™ + ne-> Me;at other points on the metal surface metal ions are reduced to metal atom, picking up electrons out of the metal electrode and depositing on the metal surface. © In the first process the metal electrode ‘will be charged negatively opposite | the electrolyte; in the second process the metal electrode will be charged positively opposite the electrolyte. Af-EvectrocHemistry ter some time an equilibrium situation ‘will exist; this equilibrium condition dictates that the rate of both reactions Aaand B be equal: Me © Me" + ne ‘The potential of the metal electrode to the electrotyte (the real electrode po- tential) is called equilibrium potential ‘The extent of this equilibrium poten- ‘ial is determined by two parameters + the type of the metal in contact with the electrolyte; + the concentration of the metal ions (patticipating in the reaction) in the electrolyte. ‘The different metals have a differ- ent tendency to send ions into the electrolyte. Noble metals like eg. gold have a low tendency to go into solu tion while zinc has a high tendency to dissolve, ‘A metal electrode in a solution is called a half-cell. When the concentra: ° tions of the reactants are maintained at unit activity (IN = Tkmol/m®), they are termed standard half-cells. ‘The potential difference between the solid piece of metal (a half-cell) and, solution cannot be measured disectly. Itis possible to measure the potential difference between two half-cells as F electrod Figure 2. Meta electrode Me in electrolyte Solution containing Me™ ions in dynamic equilibrium cu sty |) 2 et cy Figure 3. Reversible cel containing copper and zinc in equilibrium with ther ions. © shown in Figure 3. In this figure an electrochemical cell is constructed. containing copper and zinc electrodes in equilibrium with their ions sepa- 5 rated by a porous membrane to retard ‘mixing. For purposes of simplicity, the concentration of metal ions is main- tained at unit activity. Both electrodes are at equilibrium, The reactions in ‘each compartment are represented by the following equations Cu +2e6Cu 20. Zn +20 ‘The rates of metal dissolution and de- position are the same; there is no net change in the system. Ifa high resis- tance voltmeter is connected between, | the copper and the zine electrodes, a Potential difference of about 1.1 volt is | “observed. This isthe cell potential (or EME) which is used in.determining the free energy of the overall electrochem: [cal reaction. The positive terminal of the voltmeter must be connected to the copper electrode, and the negative terminal must be connected to the | zine electrode to have the voltmeter ead on scale. ‘The cell potential isthe difference be- tween the electrode potentials of both Bewpouze ~ Ezaynae ‘An electrode can be an electron con- ducting metal, which takes part in the electrode reactions. It can also be an inert metal at which the electrochemi- cal reactions take place while itself does not take part in the reaction as fot instance the hydrogen electrode. A platinum electrode acts as an inert substrate for the electrochemical reaction. At different points on the platinum electrode, hydzogen ions are reduced to hydrogen gas and hhyclrogen gas is oxidized to hydrogen = tons, with electron transfer occurring between these points; 2H? + 2e > Hy Its important to note that the Pt | electrode does not take part In this = reaction but merely serves as a solid Interface at whch this reaction can occur. Many metals function as revers- ‘ible hydrogen electrodes; Pt is usually preferred cue to its inertness and the | ease with which electron transfer oc- ‘curs on its surface. If concentration of hydrogen tons is maintained at 1 gram atomic weight per liter (unit acti ity) and the solution is saturated with Figure 4. Hydrogen half-cel. hydrogen gas at 1 atmosphere pres- sure we speak of the Normal Hydro- gen Flectrode, This half-cell reaction of NHE is arbitrary chosen as zero potential. ‘The hydrogen half-cell is shown in figure 4. ‘Combining a NHE with any metal clectrode to an electrochemical cel, © the measured potential sper dcfini- tion the electrode potential of the ‘metal electrode. {In figure 5 an electrochemical cell con- taining reversible zinc and hydrogen electrodes is shown. ‘The cell as shown in figure 5 has a potential of 0.76 and zinc is negative with respect to the hydrogen elec- trode. Using the above mentioned convention, the hydrogen electrode is defined as having the potential of zero, and it follows that the potential of the zine half-cell is -0.76 volt. In = this fashion, other half-cell potentials © can be determined. Im Table 1 the haif-cell potentials for some electrochemical reactions are | listed. This table is frequently called the EMF series, half-cell, or reduction- | oxidation potentials (abbreviated redox potentials) In all cases these potentials refer to | electrodes in which all reactants are at unit activity and at 25°C, For measurements as mentioned be- fore it is not necessary to use the NHE. Fach well defined, constant half-cell © can be used as a reference electrode ifigure 5, Cell containing reversible zinc and dagen electrode. Figure 6. Saturated Copper-Coppersulphate Electrode the potential difference with the NHE is known. Generally more practical reference electrodes are used lke: * Saturated Calomel Electrode (SCI +242 mV versus NHE + Silver Silverchloride Electrode: +288 mV versus NHE + Copper-Coppersulphate Electrode: 4335 mV versus NHE (Figure 6), ‘The EMF series as mentioned before Is not very Important in studying cor- rosion phenomena since this series is related to circumstances which are not likely to occur in practical situations. Next to the theoretical EMF series other “galvanic series” are determined, In some of these series the potentials ‘of metals as well as alloys aze mea. sured in practical or standard solutions e.g, sea water or 2% NaCl. Examples of these series arc: Series of Von Zeerleder in 2% NaCl INCO series in seawater. ‘The galvanic series show quite some ifferences with the EMF series. They can give important information in ‘comparing the differences in corro- sion behaviour of metals and alloys in ‘concerning environment. A practical galvanic series is given in Table 2 Electrode reactions Of essential importance in the reac- ton at an electrode surface in an electrochemical cell isthe transfer of charged particles through the interface clectrode/solution. These reactions are called electrode reactions. ‘When there is only one reaction at an electrode then itis called a single electrode. In case more reactions take place at one electrode then itis called ‘a mixed electrode, When in electrode reactions a transfer of metal ions takes place we speak of a metal ions elec- trode. The electrode itself does take part in the reaction; eg. Fe and Zn, If the electrode does not take part in the reaction but merely serves as a solid interface at which a transfer of electrons takes place, we speak of a redox- clectrode; e.g. Pt and C. The ‘general formula for redox-reactions is: Red Ox tne ‘The electrode potentials of redox reac- tions can be fixed in the same way as. ‘metal ions-reactions. These electrode potentials are called redox-potentials. In Table 3 some redox reactions with the standard potentials ae listed. In case of corrosion there is more than one electrode-reaction at the corrod- ing metal sucface. One of the reactions is the transfer of metal to metal ions which is the anodic or oxidation reac- tion. The other reactions are redox- reactions or other metal ions reactions cr both beside each other; these are the cathodic or the reduction reac- tions. “ectate reseien own | [wear ‘emit | | Hetod reastin | eseek “mages az) |orseeor oat Oe +26 4900 + a aly zac 7200 a | | otreon 095 Ne secre ane 278 | | FeONe? +20 Fela 038 Mgrs 26 6 Ma suminun $9.5 067 | | Oy+2H0+4e «+ 40H 40 B+ 3 ot = mid steel 040 | | ip Beeoar 053 Bi + Be 20 -casimn 65-22 035 | | rah 20- Hos 53 OF +3269 00 1 Cro feb) wor 000 | | Fase cre ov Fe + Bos Fa sex crawnisced ectiel | aepr-030 | | Na+ adeeono~ako | oat e+ 2a stent 98.2 92 | | aeeeeane 108 NBs 2e oth tas 6040 007 | | +4 +s cann 123 S42 e950 ~coneer 2010 | | onoeeaies arco s 20 | 1.24 Pee 20-0 Po -rareiK wie | | oos2eeen 1.98 Bt De eoH2 7A.SDeuronctel 10m | | mar eBH+ secon sano | 152 OF 420-600 = throm and comer Hide + 2 « Be 6s BHO wn fave orig ikl fesse) aor s040 | | rancor 28s Hah 26 oy aes {120 | Teble2. Practical galvanic cris fora Table 3 Redox actions with standard Bcc A! 7188 | numberof alloys in air saturated, neutral potentials. Table 1 Electrochemical series of metas seawater (INCO series).ELectrocHemistry Anodic reaction (oxidation): Me Me" +ne In case an alloy corrodes, several an- dic reactions will take place simulta- neously. Cathodic reactions (reduction): Various types of reduction can occur; ‘common reduction reactions in corro- sion processes are: + Hydrogen evolution in acid solu- tion: 2H +206 Hy + Oxygen reduction in acid solutions: Op 4H" + 4e «9 2H,0 ‘+ Oxygen reduction in neutral or alkaline solutions: 0,4 2H,0 + deo 4 OFF + Metal-ion reduction: Me +ee>Me* + Metal plating: Me" +2¢ 6 Me Nenysr reuarion ‘The electrode potential is influenced by the concentration of the species involved in the electrode reaction, If the concentration of all compo- nents is 1 gram atomic weight per litre or kmol per m! (unit activity) the clectrode-potential is mentioned stan- dard half-cell potential. To determine the potential of a system in which the reactants are not at unit activity, the Nernst equation can be employed. RY fou Bo12.3 AT log Sout half-ell potential standard half-cell potential {a8 constant of Joule absolute temperature in degree Kelvin number of electrodes transferred Faraday constant qua aNd ae are the activities (concen- trations) of the oxidized and reduced species. If the temperature is 25°C the Nernst ‘equation is as follows: B= Bor 2999 Jog 2004 ae aed For the anodic reaction Me ¢» Me™ + ne, the Nernst equation is as follows: 0. Extent™*= BOwere™* + 059 og [Me™] 3 sien] For the cathodic reaction 2U* + 2e © Fi the equation is: Burg = Bony, + 2-989 Jog (HT 2 Pha 059. pH (Pyp= 1 atm.) For the cathodic reduction reaction of oxygen in acid solution equation is 02+ 4H" + 4e © 2180, the Nemst equation is Bode = Bos /uy + 2959 tog pon (HI* ‘Note: The coefficients in the reaction equation are exponents in the Nemst equation, Pournars piacrans As discussed before, there is a defi- nite relation between the free-energy change and the cell potential of an electrochemical reaction, In most instances, the actual magnitude of the free-energy change is relatively unim- | portant in corrosion applications. ‘The most important factor is the sign of the free-energy change for a given 5 reaction, since this indicates whether ‘or not the reaction Is spontaneous. Hence, although the equation | AG=- nFE forms the basic for the thermodynamic calculations itis rarely used in studying corrosion phenomena. However a simple rule derived from the equation AG = - FE is used to predict the | spontaneous direction of any electro- chemical reaction, This rule can be simply stated as: In any electrochemi- | eal reaction, the most negative or ac- tive half-cell tends to be oxidized, and the most positive or noble half-cell tends to be reduced, From the above rule it follows that all | metals with reversible potentials more active (negative) than hydrogen will tend to be corroded by acid solutions Copper and silver, which have more noble potentials, are not corroded in acid solutions. However, if dissolved ‘oxygen is present there is a possibility of oxygen reduction, indicating that In presence of oxygen, copper anid = silver tend to corrode spontaneously. As the reversible potential of a metal {becomes more noble its tendency to corrode in the presence of oxidizing agents decreases. Metals such as platinum and gold —— Figure 7. Simpliied Pourbaix diagram for water at 25°C are very Inert since there will be no tendency to corrode except in the presence of extremely powerful oxidiz~ ing agents, ‘Thermodynamics, or more specifi cally, half-cell potentials, can be used © to state a criterion for corrosion: corrosion will not occur unless the © spontaneous direction of the reaction indicates metal oxidation, = Itis important to note that although the spontaneous direction of a reac- tion may be in the direction of metal Corrosion, this does not necessarily indicate that serious corrosion will occur. Ifthe reaction proceeds at a negligible rate, then the metal will be essentially inert. Hence, the major use of thermodynamic calculations is negative one. That is, thermodynam- © Ics can indicate cleatly that corrosion will not occur ‘The applications of thermodynamics to corrosion phenomena have been further generalized by means of po- = tential - pH plots. These are frequently called Pourbaix diagrams after Dr. M. Pourbaix who first suggested their use, | Such diagrams are constructed from calculations based on the Nernst equa- tion and solubility data for various ‘metal compounds AA simplified form of the E - pH dia- gram of water is shown in figure 7 Line a in the diagram represents the equilibrium of the reaction: 024+ 4H + 42.6 2H,0 with the Nernst equation: Fog logPon {Ht Foon + 0.059 4 © orat poo Latm, Foohivo = 1.228 - 0,059 pHatm, the region marked H,O between ines a and b water is stable with atm. In other words: in that region the jetion 2H2 + 02 ¢>2H,0 has a AG < \d tends to proceed spontaneously If water is brought Into the region marked O, the reaction O, + 4H1* + 4e {¢2H,0 will proceed spontaneously to the left with oxygen evolution. In a closed system this means that the pH sill decrease and the oxygen pressure ‘wlll increase. Similarly in the region marked H, re- action 2H + 2e © Hz proceeds spon- taneously to the right with hydrogen evolution and increasing pH, The line c represents the equilibrium: H,0.@ Ht +OH In this reaction electrons do not play a part, This means that itis not influ. enced by the potential and thus runs vertically. ‘The simplified potential - pH diagram. for tite Fe-H,O system is shown in figure 8. All lines for the iron equilibria are drawn for (Fe) = 10° mole/l. This concentration is often chosen in cor- rosion work as being so low that this corresponds to “natural” conditions ‘We will now discuss the application of this type of diagram to the corrosion of iron. We see that in acid solutions the line Lis below line a of the hydro- {gen evolution reaction. This means ‘that reaction { cannot be in equilib- rium in acidic water because the total reaction: Fes 2H ¢> Fe? + Hy will proceed spontaneously to the right and iron corrodes. Only when, With the help of an external voltage source, iron is brought to a potential below line I the dissolution of iron will stop. In this region iron is sald to be immune. The hydrogen evolution will of course continue, If the pHi of the solution is ralsed we come to line I Again this is below the hydrogen evo: lution line and the total reaction 3Fe-+ AHO «9 FeO, + 4H will tend to proceed spontaneously to the right. However, it now depends on the form in which the solid iron oxide (magnetite) is produced whether this, reaction continues or not. Ifthe oxide forms a closed, adherent layer the reaction, although remaining sponta- neous, will nevertheless (almost) stop because the direct contact between ‘metal and solution Is broken. When this occurs we say that iron has be- ccome passive. If the oxide is porous or non-adherent no passivity occurs and the dissolution reaction continues ‘The E - pH diagram shows the regions in which passivity may be expected to occur, but not if it will indeed so, because that depends upon the form in which the oxide is produced. At stil higher pH the iron starts to ‘corrode again according to the reac. ton: Fe+OH +H,0 © HPeOs + Hy Al iron-dissolution reactions also lie under the oxygen reduction line b. In oxygen containing solution we will thus simultaneously have reactions such as: 2Fe +O, + 4H" «> 2Fe* + 21,0, pro- ‘ceeding spontaneously to the right. As shown, itis possible to delineate (volt) 4 —> pH EE immunity EE passivity RE corrosion areas in which iron, iron hydroxide, ferrous ions etc. are thermodynamical- ly stable. That is, these forms represent states of lowest free energy. ‘The main uses of these diagrams are: + predicting the spontaneous direction of reaction; + estimating the composition of corto- sion products; + predicting environmental changes, which will prevent or reduce cor- rosive attack. For example the large region in labeled Fe indicates that iron is inert under these conditions of potential and pH. Potential - pH diagrams are subject to the same limitations as any thermo- ‘dynamic calculation. They represent equilibrium conditions and should never be used to predict the velocity of a reaction. ‘Thermodynamic arguments only lead to conclusions about the dizection of possible reactions and not about the rates with which they take place, Summarizing we can say that a metal tends to corrosion when it is in con- tact with a solution where a cathodic reaction is possible with an equi "ium potential higher than that of the metal dissolution reaction in that solution, -————-$]$]$)>— ser FeO, + BH + Be e9 3Fe + 40 W HeDy + 24+ 2e 4 Fe +240 IV. Shady + BH + ee 2F ey + Hed VFO + BH” 9 2009 Bre + 340 Vi Felts eeore VI Fada» 240+ 20> OHFeDS + HY 12 Figure 8. Simplified potential - pH diagram for the Fe - HO system.ELectaocHemistry 2.3 Electrode Kinetics Introduction From an engineering standpoint, the ‘major interest is in the kinetic s or rate of corrosion. It is not sufficient to know whether a metal can corrode reaction. Corroding systems are not at equilibrium, and therefore thermody. namic calculations cannot be applied. Essentially we are Interested in what happens when cells such as have been described before are short-circuited as illustrated in figure 1 In this instance a vigorous reaction oc- curs, the zine clectrode dissolves in the solution and simultaneously a rapid evolution of hydrogen is observed at the platinum electrode. Electrons released from the zincdissolution reaction are transferred through the connecting wire to the platinum electrode where they are consumed in the hydrogen-reduction reaction, ‘The process which occurs is the same process that occurs when zine metals, immersed in an acid solution. In the divided cell as shown in figure 1 the reactions occur on separate electrodes, while at the immersed piece of zinc in an acid solution th oon the same metal surfas Before discussing the electrode-ki- netics, several useful terms should be defined, Anode refers to an electrode at which a net oxidation process cathode refers to an electrode at which a net reduction reaction occurs, In fig: lure 1 the zinc electrode is the anode, and the platinum or hydrogen elec~ trode is the cathode in this particular cell with short-circuited electrodes. An reactions occur urs, and Figure 1. short-circuited cell containing zine and hydrogen electrodes 10 an oxidation reaction and cathodic reaction is synonymous with a reduc ton reaction, Faraday's law Electrode reactions ate running with. transfer of charged particles like fons and electrons. There isa relationship between the current density in the electrochemical cell and the amount s per time unit (I sulomb per sec) By means of an Ampere-meter, in- stalled in the connecting wire betwe the two electrodes of the electrochem- 1 cell as shown in figure 1, w measure the current density. The cur- rent density is related to the amount ‘of metal which is dissolving, so related with the corrosion rate, ‘This relationship is glyen by Faraday’s law: Ian electric charge of F coulomb is passed, of each species X, which, takes part in the reaction, Mx/nx has reacted, F = Faraday constant Coulomb per (g)mol X = symbol for reacting species Mg = Molecule weight ‘nx = number of electrons transferred, ‘The reaction rate of a species X (vs) is defined as the amount of species X reacting per surface unit and per time unit, The relationship between the reaction rate vy and the current density iy, de fined as strength of current per surf uit, is given by: ig = Evy 9.68. 108 Exchenge-current density Let us consider again for instance the 20 cell in equilibrium situation own in figure 3 in chapter 2, 2.2 Electrochemistry, Thermodynamics, In equilibrium situation at the copper Figure 2. Short-circulted cell containing zine and hydrogen electrodes. (LL. Shree) Cu Cu +26 is as fast as the cathodic reaction Cum +2 Cu ‘The reaction rate of both reactions is even in equilibrium situation nat 2 The absolute value of the cathodic cur- rent density Tox, is equal to the anodic current density Ige,: The direction of both custents is opposite which means that there is no net current, In equilibrium situation the following equation isv Tat lc ine magnitude isis called exchange- it density and Is related to t rate at which charged particles are passing through the Interface metal/ electrolyte ‘The relationship between the ex- ‘change reaction rate and current density can be directly derived from where Vooa and Ya are the equilibrium ‘oxidation and reduction rates and ig is the exchange-current density; n and F have been defined previously. ‘The exchange-current density depends for instance on kind of metal electrode and on the surface roughness of this metal electrode. The exchange-current density of hy: drogen reaction 2H" +2e ¢ H, at platinum is extremely igh. For this reason platinum is chosen for the NHE (normal hydrogen reference electrode), Exchange current densities for the H* H, system are markedly reduced by the presence of trace impurities such as.arsenic, sulphur and antimony- containing fons. Polarization and overvoltage When a cell as shown in figure 1 is circuited, and net oxidation \ 2H Hh \nCHEeFoe call at 3 Esl ue Ra Figure 3. Diagram illustrating polarization and ove voltage of anodic and cathodic eletrade reactions.sduction processes occur at the fe interfaces, the potentials electrodes will no longer be ‘equilibrium potential. This jation from equilibrium potential js called polarization. Polarization an be defined as the displacement Fhe electrode potential resulting ‘net current: The magnitude of ‘polarization is frequently measured terms of overvoltage. Overvoltage, wally abbreviated as n, Is a measure of polarization with respect to the equilibrium potential of an electrode. ‘That is, the equilibrium potential of an clectroxe is considered as zero, and the overvoltage is stated In terms of volts or millivolts plus or minus with respect (0 this zero reference. ‘This polarization phenomenon can be considered by means of the idealized equipment as illustrated in figure 2, ‘The simplest way of visualizing polar- ization and overvoltage is to consider the idealized experiment shown in ‘igure 2, in which zinc, in a solution of its own fons, is corroded when an elec- trical connection is made to platinum Jn an acid solution. ‘The electroche: cal cell may be presented as: Zn] Zn {Hf Hy Pe In this corrosion cell, Zn is the anode and its potential E, Is measured by means of a high impedance voltmeter and a reference electrode (RE). The potential of the Pt electrode, Ee, 1s similarly monitored. The corrosion rate is controlled by the resistance Ry between Zn and Pt and measured by ‘means of a low impedance ammeter. Figure 3 shows how Es, Fe and i (cor- rosion current) change as Rx varies from « to 0, When Ry = = no current can flow from the anode to the cathode and so. B= Egg, the open circuit potential of the anode. Ee is similarly Ec. © situation the polarization As Ry is reduced, Zn passes into solu- tion as Zn’ and at the cathode H’ is reduced to Hp gas. However, the flow of current has increased E, from Eoy toa higher potential, whilst Ec has, + fallen to a more negative value. The ifference between E and Fy, the open circuit potential, is called the overpo- tential, ‘The anodic overpotential nA is posi- tive: A = Eq- ox ‘The cathodic overpotential nC is nega- tive: nC = Eo- Boe The current flowing through the am: imeter jis directly related to the rate at which Zn goes into solution. When Rx =0, the Zn and Pt are short circuited, (the usual condition during | corrosion because the anodic and ca- thodic reactions are usually occurring | ona metal surface in close proximity), and $0 Ee = Eq = Bean Fens is the corto sion potential, and the corresponding current is the corrosion current. Polarisation curves; Evans diagrams Let us consider an electrode reaction, consisting of an anodic and a cathodic reaction, occurring at one single elec- trode, If we plot the current density j Of this reaction against the overvolt- age 1) we get the polarization curve (figure 4). ‘The line of net current density is the algebraic sum of the partial current densities fa and ic. In the equilibrium 0. This dicated as equilibrium potential is, © point zero on the vertical axis. On the horizontal axis the linear current density is plotted = In case the overpotential y = 0, the net current density is 0. in this situation the anodic and cathodic current densi- ties are equal and opposite in direc- tion. Both current densities equal the © We have to discuss now more deeply exchange-current density ig. In case of 2 large overpotential the net current density is almost equal to one of the partial current densities Beside the linear polarization curve fone uses the logarithmic polarization curve (Evans diagram), in which the logacithmic current density is plotted horizontally (figure 5) ‘The anodic and cathocie branch of the polarization curve are lacated at the same side of the potential axe, which simplifies determining of the corrosion potential and corrosion cur- rent density at the intersection of the ‘two polarization curves ‘The logarithmic polarization curve also shows that there isa linear rela- tionship between » anc log |i Polarisation processes, determining the corrosion rate the polarization theory. As mentioned before polarization can be defined as, the displacement of electrode poten- ‘al resulting from a net current. We have to notice that, the greater is the polatization (the larger the slope of the Ei curve) the lowe: is the cor- rosion current. Hence polarization is beneficial in reducing the corro- = sion rate. The most common forms of electrochemical polarization are charge-transfer (chemical ot activa- = tion) polarization and concentration Polarization. A third possibility of polarization Is resistance polarization, occurring if the electrolyte has a high resistance, so that migration Is slow. ‘The charge-transfer and concentra- = tion polarization processes will be discussed in this corrosion engineering guide, : os a I ss Soret Charge-transfer polarization Charge-transfer polarization or activa a= a+ Blog hi] a= -B log ie t a= charge transfer coefficient B= Tafel slope Eq- Aon = 6 log Mio na > leg Figure 4. The linear polarization curve of single electrode reaction. Figure 5. Logarithmic polarization curve of a single electrode reaction (Evans diagram). Figure 6. Charge-transfer polazation atthe anode,E.ectrocHemistry tion polarization refers to electzo- chemical reactions which are con tolled by a slow step in the reaction sequence. Ifthe taking-up or giving- off of electrons is rate determining we speak of charge-transfer polarization, AAs discussed before this slow step dur- Ing hydrogen evolution might be the electron transfer step or the forma- tion of hydrogen molecules. In figure 6 charge-transfer polarization of an anodic reaction is illustrated. This po- larization mechanism is also possible ‘with cathodic reactions. if charge transfer i the rate determin- ing tep, the overvoltage is indicated as charge transfer overvoltage ny. In case [n> 0.1 V the following rela- ‘ionship is valid between the rate of the anodic electrode reaction i, the exchange-curtent density igand the overvoltage n¢ nu= 0 +8 lg li “6. logiy wcplone in which a is the charge transfer coet- ficient and ps the Tae slope or Tafel constant. ‘This equation is called the Tafel equa tion. If logarithmic curent sale is used, the relationship between overvoltage 0 potential and current density isa linear function as shoven already in figure 5. The value for f for eectro- chemical reactions ranges between 0.05 and 0.15 volt and is usually 0.1 volt ‘At a very small overvoltage || < 0.01 V there isa linear relationship between n, and! (ee figure 4) tu=Kify (Ke RT/nf) The slope ofthe polarization curve at equilibrium potential m= 0s indi- ‘ated as polarization resistance Rp. So, previous equation can be written Rp Concentration polarization When the rate of the charge transfer process is very high (high exchange- current density and/or high overvolt age and high corrosion rate), itis possible for the transport of ions or = molecules to or from the electrode to become rate-determining, © If transport of reacting species is limit- 5 Ing the reaction rate, overvoltage is indicated as concentration overvolt- age. There are three main mechanisms by which transport can take pla ‘+ migration in an electric field, + convection due to movement of the liquid, © + diffusion caused by concentration differences. ‘Migration can nearly always be neglected in our case because there usually is an excess of “neutral electro: lyte", Le. ions not taking part in the electrode reaction.” © Convection causes homogenization of © the solution but, depending on veloc- = ity and geometcical factors, concen- tration differences stil exist in a thin boundary layer 6 of 0.01 0 0.5 mm thickness. There diffusion will take place. Nemst introduced a model to calcu- late the effect of this diffusion. In this model the concentration is assumed to be constant at C® in the bulk of the solution. In a boundary layer of thick ress 8 next to the electrode Nernst assumed a linear concentration gradi= cent from the value C? in the bulk of the solution to Cat the surface. This is shown in figure 7. According to Fick's lw the flux of the substance to the electrode is propor- tional to the concentration gradient with the diffusion coefficient D as the proportionality constant. Combined with Faraday’s law this gives for the current density: i= -nFD. (C*.C)/6 ‘The rate of transport of a certain spe- les through this layer is maximum if the concentration at the electrode sur- face is zero. In that case the maximum possible current density is reached; usually called the limiting custent ly, which is equal to: i,=-nFD.C%8 for C=0. Especially at high current densities diffusion is rate determining for the electrode reaction, The relationship | between diffusion or concentration, overvoltage 46, the limiting current density il and the current density is: ye= 2.3. RT/nF .Log(t - ify) ‘The diffusion polarization of a cathod. ic reaction is illustrated in figure 8 It Is obvious that the limiting diffu- sion current density increases with © increasing electrolyte velocity. © Combined polarization Usually, at an electrode charge transfer polarization (at low reaction rates) a well as ditfusion polarization (at higher reaction rates) may occur. Especially at reduction processes like ‘oaygen reduction concentration po- larization Is combined with activation polarization. This combined concen- tration and activation polarization Gr=ns + 1) is illustrated in figure 9 Using only three basic parameters, namely, 8, ip and ig, the kinetics of virtually every corrosion reaction can be precisely described. Previously ‘mentioned equations represent a sim- plification of the complex phenomena observed during corrosion reactions. ‘The use and application of some equa- tions will be discussed more in detail In chapter 5, “Corrosion Examination, Inspection and Monitoring’, Figure 7. Concentration profile in difusion layer according tothe model of Nemst (PJ Gellings) 12 tion of cathodic reaction Figure 9. Combined activation ana diffusion polarization of 2 cathodic reaction= mixed: potential theory has fen to be useful in corrosion stud- because: it permits prediction of rosion behaviour, ithas been used to develop new cor- ‘osion prevention methods, ‘it has been used as a basis for new "apid corrosion-rate measurement ~ techniques. Corrosion reactions are, as mentioned before, electrochemical reactions. Blec- trochemical corrosion is only possible if there exist potential differences at ‘the interface metal/electrolyte. ‘These potential differences may originate from the metal or from the electrolyte Potential differences originating from the metal: ‘contact of different metals or alloys; + presence of precipitations, (chromi- lum) depleted zones, inclusions; * differences in microstructure; + roughness of metal surface; * differences in stress level (tensile stresses are anodic compared to com- pressive stresses) Potential differences originating from the corrosive environment: * differences in concentration; * differences in oxygen content; + temperature difference. We will discuss a number of corro- sion phenomena using the theoret cal aspects regarding dynamics and Kinetics of electrode-reactions. The phenomenon of passivity will be discussed as well. The effect of several different environmental and metallur- gical variables are considered from an electrochemical point of view, mmplex cor 2420 Hy Figure 1. Schematic illustration of corrosion of ion in acid solution, Corrosion in acid environment. consider the corrosion process of a pi on. There are two different electrode reactions: [eo Fe" +20, the anodic reaction and. 2H"+2e cH, the cathodle reaction. The equilibrium potential of these reactions can be calculated with the Nemst equation. ean = Foren: + 2959 tog (Fe) {Fe"[o10% mole D + 0.059 log [H] 0.059 pH The corroding iron constitutes a poly- electrode and the corrosion potential will settle in between the equilibrium potentials of the anodic and cathodic reaction. At this mixed potential or corrosion potential the Fe/Fe~ reaction is polarized anodically and the H,/H* reaction is polarized cathodically. This polarization is in such a way that the anodic current density is equal in ‘magnitude and opposite the cathodic current density. The metal will be in trical neutrality. ‘The stationary-corrosion situation is illustrated by means of the logarithmic polarization curve in figure 1. At the horizontal axis the logarithmic current density is plotted and at the vertical axis the electrode potential plotted. Figure 2. Schematic illustration of corosion ‘of zinc and iron in acid solution, ‘The steady state of this particulars tem occurs at the intersection between the polarization curves for iron disso- lution and hydrogen evolution. ‘The corrosion potential Fy, and the corrosion current density 10g lll ‘can be determined by means of the Intersection of the two curves Figure 1 also indicates that as higher is the potential difference between the two electrode reactions as higher Is the corrosion current density. The corro- sion current density is also influenced by kinetic parameters like exchange ‘current density and the Tafel slope, Out of before mentioned example we can understand the mixed-potential theory of Wagner and Teaud, This theory consists of two hypotheses: 1. Any electrochemical reaction can be divided into two ot more partial oxidation and reduction reactions. ‘There can be no net accumula- tion of electrical charge during an electrochemical reaction, From this it follows that during the corrosion, of an electrically isolated metal sample, the total rate of oxidation ‘must equal the total rate of reduc: ‘The mixed-potential theory, together with the kinetic equations described above, constitute the basis of modes electrorle-kineties theory orrosion of metals in oxygen-free and non-oxidizing acids is thermody- ‘amically just possible if the equilibri- tum potential of the Me/Me" reaction is below the equilibrium potential of the H°/H reactio In oxidizing acids, however, like for Instance nitric acid, even noble metals like copper will corrode, In case of metal e.g. copper corrosion in nitric acid evolution of nitrous will occur Anodic reaction: Cue Cut 4 20 Cathodic reaction: AH’ + NO + 3e ++ NO + 2,0 Kinetic parameters Kinetic factors can be demonstrated by for instance the behaviour of Fe and 2n in diluted sulphuric acid. Attack of both metals here is about equal, although the equilibrium potential of zinc is much more negative than the equilibrium potential of iron. From14 | LECTROCHEMISTRY this thermodynamic point of view one expects a much higher corrosion rate of zinc. The lower equilibrium poten- tial of zinc is however compensated by the fact that the exchange-current density of the hydrogen reaction (liz) at the zinc surface is much lower than at the iron surface Beside the exchange current density also the Tafel slope is very important. ‘The corrosion of jron and zine is schematically illustrated in figure 2. In this figure the intersection of the two curves representing the electrode reactions occurring at the iron surface Is Indicated A; the intersection of the two curves representing the electrode reactions at the zinc surface is indicat- ed B. Both corrosion current densities are almost equal, despite the much lower equilibrium potential of the Zn/ Zn* reaction. This is caused by the lower exchange-current density of the H*/He reaction at the zine electrode. If the exchange-current density of the hydrogen evolution reaction at the zinc electrode would be equal to the cexchange-current density of the hy- drogen reduction reaction at the iron ‘electrode the corrosion rate of zinc would be much higher, indicated by intersection point C in figure 2. Corrosion in neutral and alkaline environment (oxygen-corrosion) 1n neutral solutions (pH=7) the equi- librium potential of the hydrogen reaction has been decreased to low values compared to acid environment, Also the exchange current density io decreases at increasing pH. This reduc- tion reaction is of minor importance at neutral and higher pH values. ‘The corrosion rate will be determined in increasing measure by another redox-reaction: the reduction of oxy- gen. In most cases more or less oxygen, is present in electrolyte solutions. ‘This kind of corrosion with oxygen reduction as cathodic reaction of the corrosion process is indicated as oxy- gen corrosion. Oxygen reduction takes place according one of the following equations: + in acid environment: On 4H + de © 2H;0 ‘+n neutral and alkaline environment: 02+ 1,0 + de 6 40H ‘The equilibrium potential of the oxy- en reduction reaction is so high that even noble metals like copper and sil ver, which do not corrode in (oxygen free) acid media, can corrode if oxygen 5 is present in the solution. he reacting oxygen is transported to the metal interface by diffsion, Asa ‘consequence polarization according fre Provided that the potential can be ‘maintained in the passive area of the anodic polarization curve, the current required to maintain passivity is very small. This current, Ig, isthe current Just necessary to plug the “weak spots” where passivity has broken down, ‘There are several processes which lead to break-down of passivity: + Self activation by local cathodic reduction of oxide. * Adsorption of chloride. + Oxidative dissolution (in transpassive area), * Chemical breakdown: dissolution of, oxide. + Undermining: rapid attack of substrate at breaks in the film. ‘+ Mechanical: applied steesses will ‘cause brittle films to fall, Active dis. solution then occurs at breaks in the film until repassivaticn occurs. The presence of chlorides has a delete- tous effect on the passivation process. A number of mechanisms may be operative: * Chloride Is strongly adsorbed at ‘the metal surface and forms strong soluble complexes with most metals thus favouring active dissolution. * Chloride can “dope” existing oxide scales leading to a detective oxide with increased conductivity, *= The presence of chlorides favours conditioning for pitting. ‘The relationship between an Evans diagram and a Pourbaix diagram is Eon r os e ‘eal —> ee eee = on Teal al ooh a Figure 5. Polarization curve ofa metal with active-passive behaviour, Figure 6, Schematic illustration of Pourbaix and Evans diagram to illustrate thelr telationship.ELectnocHemistry shown in figure 6. The Evans diagram. in figure 6 corresponds to a pH of 5. ‘The two diagrams drawn together help to emphasize the relationships between the zones marked on the Pourbaix diagram, Influence of electrolyte resistance Corrosion generally takes place in solutions containing ions, part pating in the electrode reaction well as “neutral” electrolytes. These “neutral” electrolytes do not inlu- ence the pH and do not take part in the electrode reactions. They allow however that the charged particles be ‘transported through the solution from the anode to the cathode. Ifthe liquid hhas an electrical resistance above zero, a certain potential difference AE is necessary between the anodic and cathodic areas at the surface to allow the current to flow through the electrolyte, This has a direct influence ‘on the corrosion rate, as schematically illustrated in figure 7. If the total electrolyte resistance is zero, the corrosion current density has the highest possible value (ion) and the metal surface has at all places the same potential Eeoy. Ata certain resistance R, a potential difference AE: iy R exists between the anodic and cathodic area's (eo < hon) With decreasing electrolyte resis: tance, that means with increasing salt concentration, generally the corrosion rate increases. In some cases the corrosion rate decreases with increasing salt concen- tration. As an example the behaviour of iron, In NaCl-solutions, saturated with ‘oxygen, is discussed. Initially the cortosion rate of iron increases with {increasing NaCl concentration. Above about 0.5 mol/l (approximately the salt content of seawater), the corto- sion rate decreases with increasing NaCl-concentration, This is caused by decreasing solubility of oxygen as shown in figure 8, Influence of velocity Ina diffusion controlled cathodie pro- cess at increasing velocity the diff sion layer decreases and consequently the limit current density increases. Let us consider a system containing @ ‘metal M immersed in a corrosive sys- tem in which the reduction reaction is under diffusion control as illustrated in figure 9, Curves 1 to 6 correspond to limiting diffusion current densities at relative solution velocities ranging from 1 to 6. The velocity effects may be deter- ‘mined by plotting corrosion rate ver- sus the relative velocity of the electeo- Iyte. As velocity is increased from 1 10 3, corrosion rate increases from A to C. However, as velocity increases further, the reduction reaction becomes activa tion controlled. As a consequence, the corrosion rate becomes independent of velocity. As tthe velocity is increased from 4 to 6, the corrosion rate remains fixed at point D. From this discussion, two general rules concerning velacity effects are evident: + Solution velocity affects the corro- slon rate of a diffusion-controlled system. Velocity has no effect on activation-controlled systems. * The corrosion rate of a metal in a iffusion-controlled system becomes independent of solution velocity at very high velocities ‘These statements assume that no ero- ace fn Me Me ene (ea [RE eH lone Wotan i2a4se Inereating veleety C1 aa 4. Retawvo velocty Figure 7. Influence of electrolyte resistance on the ‘corrosion cate. Figure 9. Effect of velocity on the electrochemical bahaviour and corrosion rate of a metal (ith active behaviour) corroding with difusion-controlled cathodie reaction. TRS, Corrosion rate —+ ° ° increasing vebeiy*® —® Figure 8. Influence of NaCl-concentration on corrosion rate of iron in aerated solutions at 25°C and the solubility of oxygen (dotted line). 16 Figure 10. Effect of velocity onthe electrochemical Bahaviour and corrosion rate of a metal corroding under difusion contol‘velocity. As velocity is increased, the corrosion rate increases along path (C. When velocity Is Increased be- 143, there is a rapid transition from int C in the active region to point D the passive region. These results are shown in terms of velocity versus corrosion rate. The difference in veloc- ‘dependence between a metal with active behaviour and one dem- ‘onstrating active-passive behaviour is ‘the result of the unusual dissolution ‘behaviour of active-passive metals ‘and alloys. This behaviour is typical of all active-passive metals which are corroding under diffusion control Tt isalso obvious that the ease with ‘which an active metal is passivated by increasing velocity Is determined by the critical anodic current density. The smaller the critical anodic current density, the easier a metal will be pas- sivated by an increase in velocity, Influence of oxidizers fan active metal Is corroding in an acid environment in absence of oxidizers the corrosion rate is deter- mined by the intersection point of the polarization curves of the hydrogen evolution reaction and the metal dis- solution reaction. This is indicated as, Fon in igure 11. By addition of an oxidizer like Fe'ions, the cathodic polarization | corrosion rate incr on of electrochemical corrosion theory on corrosion reactions curve will have another appearance, © Actually there are two cathodic reac: tions. To determine the corrosion potential the two reactions have to be added together resulting in cathodic sum-polarization curve. The inter- section point ofthis curve with the anodic polarization curve indicates the new corrosion potential Fey. The 865 frOM i to {gq The hydrogen evolution however decreases from i‘, tO Iya. However, the total cathodic current density Increases and as a result also the cozto- sion current density ‘The phenomenon of decrease in hydrogen evolution by addition of conygen or an oxidizing agent in acid solutions is called depolarization. Figure 11 indicates that this depolar- ization is a direct consequence of the change of the corrosion potential to higher values and is independent of the chemical character of the oxidiz- Ing substance. ‘The magnitude of these effets is dependent on the redox potential of the oxidizer and its kinetic parameters (exchange-current density and beta slope). ‘The peculiar anodic dissolution behaviour of active-passive metals in presence of oxidizers is illustated Jn figure 12, which shows a typical active-passive metal M immersed in an ° 5 oxidizer necessary to achieve passiva- electrolyte containing a redox agent © R We consider the behaviour of this system as the amount of oxidizer is increased. For simplicity, it is assumed that the exchange-current density for this redox system remains constant | and that the only effect produced by Increasing the amount of oxidizer is to shift the reversible potential in the positive direction according to the ‘Nernst equation. The increase in oxi- izing agent from conceatration 1 t0 7 5 is represented by curves 1 to 7. Initially, metal M corrodes in the active state ata rate corresponding | to point A. As the concentration of oxidizer Is increased from 1 t0 3, cor. | tosion rate increases from A to C. Ata © concentration corresponding to curve 4, there is a rapid transition in cor- rosion potential from point D in the active state to G in the passive state 4s oxidizer concentration Is increased from 4 to 5, the metal remains passive and its corrosion rate remains low and constant. As the concentration of the oxidizer Is increased further, the transpassive region is intersected and. corrosion rate increases again, Let us consider now the effects of reducing oxidizer concentration in an active-passive system. Ifthe concentra tion of oxidizer is reduced by dilution or replacement of the electrolyte, the corrosion rate will decrease from point Jo I. Further reductions in the con- centration of oxidizer will cause the system to shift into the passive region following points H,G, and F. Further reduction corresponding to curve 2, a transition from point F to B occurs. As the concentration is reduced from 2 to 1, the corrosion rate will decrease from point B to point A. The hysteresis in figure 12 indicates that the amount of tion is greater than the ariount of oxi- dizer necessary to maintain passivity. Figure 12 may be used to explain the unusual active-passive behaviour © observed during immersion of iron in 5 nitric acid as shown in figure 4. Ifthe concentration of nitric acid is diluted so that both active and passive states are stable (oxidizer concentration between 2 and 4), then scratching or Peery ferrere Log corrosion tate + Figure 1. tnflence of oxidizing components on the electrochemical behaviour of active metals, Figure 12. Effect of oxidizer concentration on the electrochemical behaviour and corrosion rate ‘of an active-passive metal.ELectrocHemistry damaging the surface causes a transi- tion from the passive to the active state ‘To maintain passivity in a safe man- ner, oxidizer concentration should be equal or greater than the minimum amount necessary to produce sponta- neous passivation, Influence of metallurgical aspects Potential differences and as a conse- quence corrosion reactions are also Influenced by metallurgical aspects. Materials composed out of two or ‘more elements are called alloys. We distinguish two different alloys: ho- -mogenleous or heterogeneous alloys, In homogeneous alloys the different elements are completely soluble; the material consists out of one phase (example: austenitic stainless stee! X2CrNi19-11). In heterogeneous alloys the different components are not completely soluble, resulting in different phases, As an example can be ‘mentioned carbon steel in which car- bon with iron forms carbides resulting in lamellar phases. Generally homogeneous alloys are ‘more corrosion resistant than hetero geneous alloys because there are no galvanic coupling effects. However, there are exceptions; in many environ- ‘ments duplex stainless steels offer con- siderable advantages above austenitic stainless steels For the same reason very clean metals. are generally more corrosion resistant than commercial alloys. In these com- ‘mercial alloys always impurities, inclu- sions, precipitations etc. are present, A striking example is the corrosion re- sistance of extremely clean aluminum (99.998 % Al) in hydrochloric acid ‘compared to aluminum with impuri- ties (99.2.9 Al). The corrosion rate of the aluminum with impurities is some 30000 times higher then the corrosion tate of extremely pure aluminum, It has to be emphasized that duplex stainless steels form an exception to the rule that homogeneous alloys are ‘more corrosion resistant than hetero- ‘geneous alloys, 18 iterature - Chapter 2 1.LL. Shret, RA, Jarman, GT. Burstein, Corrosion, Volume 1, Metal / Environment Reactions 2, Mats G. Fontana Corrosion Engineering 3. Herbert H. Uhlig Corrosion and Corrosion Control An Introduction to Corrosion Science and Engineering (third edition) 4. Ulick R. Evans ‘The Corrosion and Oxidation of Metals 5. PJ. Gellings Introduction to Cocrosion Preven- tion and Control for Engineers © 6,Norman E. Hamner Corrosion Data Survey; ‘Metals Section (fifth edition) National Association of Corrosion Engineers,} Forms of Introduction sion occurs In several, widely dif- 1g forms. Classification is usually sed on factors like: ature of the corrosive wwironment jppeatance of the corroded ‘material of construction. Mechanism of corrosion, [Nature of the corrosive environment “Corrosion can be classified as “wet or sry’. In *wet” corrosion the oxida ton of the metal and reduction of a species in solution (electron acceptor ‘or oxidising agent) occur at different areas on the metal surface with con- sequent electron transfer through the ‘metal from the anode (metal oxidised) tothe cathode (electron acceptor re- duced). The thermodynamically stable corrosion products formed at the ‘metal/solution interface may be solid ‘compounds or hydrated ions (ations fr anions) that may be transported aay from the interface by processes such as migration, diffusion and con- vection. “Dry" corrosion reactions are ‘generally metal/ gas or metalvapour reactions involving non-metals such as oxygen, halogens, hydrogen sul- phide, sulphur vapour, etc. Oxidation, scaling and tarnishing are the more corrosion Bectrachenicst on) Coe = Sirs Corasion Cracking SCC + corrosonfovave = Exsencxasion lntergondar Selective tack ce coe Cee eo + Hiogen age Lau Matt Elmar ME) + Oven sbnitaton oD mek + Metal dusting + yogen Hel as nadng +bep + Cranston “able 1. Survey of corosion phenomena classified according mechanism important forms. A characteristic of ‘these reactions is that the initial oxi- dation of the metal, reduction of the non-metal, and formation of corrosion compound must occur at one and the same place at the metal/non-metal interface. Appearance of the corroded metal Classification by appearance (uniform 7 localized) 1s helpful fora prelimi- nary discussion. For completeness, ‘we should make a further distinction between macroscopically localized cor- rosion and microscopic local attack. In cease of local attack like stress corrosion cracking, corrosion-fatigue, selec- ai ontegmcc one gta =a (EEE, = enw tno ‘rove coon aa = SS Irgarieereson teat coraon Trae cpont fa (Cuz) & z £ t Seve sorosen Graphneaton vont rr ‘ress corouon eractng (SCC) ‘Coron aque Figure 1. Morphology of éiferent forms of corrasion tive attack etc,, the amount of metal dissolved is minute, and considerable damage can occur before the problem becomes visible to the naked eye. A schematic presentation of the dif- ferent forms of corrosion according “appearance or morphelogy is given in figure 1. Mechanism of corrosion Another possibility of classification of, the different forms of corrosion Is the mechanism of the corrosion failure mode. This involves either electro- chemical or direct chernical reactions. In electrochemical corrosion phenom- ena we distinguish phenomena in which also mechanical aspects play an important role. ‘Typical electrochemical corrosion phenomena are: uniform corrosion, galvanic corrosion, pitting, crevice corrosion, intergranular corrosion and selective attack. Electrochemical cor- rosion phenomena with mechanical interference are: stress corrosion crack- ing, corrosion-fatigue and erosion- corrosion. ‘The corrosion phenomena like oxidation, sulphidation, CO-attack © hydrogen (Nelson) attack, metal dust- ing, and nitriding can be classified as ‘chemical corrosion or high tempera ture attack, The failure mode creep will also be discussed, Beside the failure modes based on. electrochemical and chemical phe- nomena we will discuss in this course also phenomena based on physical/ ‘metallurgical aspects such as hydrogen 19TT Forms oF Corrosion 19 33% Uniform Corrosion 1@ 11% Corrosion Fatigue 1@ 19% Transgranular SCC 196% ntergranuler SOC 2% Hombritioment 1% H2- attack 195% Piting 18 4% intergranular corosion 13.8% Mechanical (wear, erosion, cavitation 19 3% High temperature 1B 10% other corrosion forms Figure 2. Frequency of fllure modes in a German chemical plant ‘damage and liquid metal damage. In these latter failure modes generally also mechanical aspects are involved. A survey of the corrosion phenomena Classified according mechanism is presented in Table 1 Corrosion under insulation (atmo- spheric corrosion) is a serious threat to the safe operation of chemical process plants, The different forms in which atmospheric corrosion can occur and the measure to combat this fallure mode will be discussed more in detail Also soll corrosion forms like stray- ‘current corrosion and MIC (mictobio- logically induced corrosion) will be liscussed separately. It is of course of interest to have Union carosin insight in the frequency of occurrence of the most common fallure modes. Several large chemical industries like DuPont, BASF and Bayer did perform Statistic analysis on occurrence of the most common corrosion failure modes. Figure 2 shows the results of this examination performed by Bayer on failure modes occurring in equip- ‘ment of chemical plants. In Table 2 the results of the investiga- tion on frequency af occurrence of corrosion failure modes in 3 different European plants are compazed to each other. Lees ack + peg + gavancoracion * ergata eeroson 1 soc lacing Eecrostomica meohenea + sums cormsion erasing *covosunfetque hysicsretaturgical ‘+ Hemttement ‘lid meta erritiement Dir caroson rms Pood Total rumbor falas Table 2. Frequency of corrosion fllure modes n several chemical industries 20Eira corson Bey chartered by a chemi Peach ection, which Fe rcnormly oe the ene ex Be slocomcs tuner and evenualy fae Pred cuted nytoslon acd Me overs entre surice Obvious, terug surface change eal o a Teper te geste destocton of Baion tonnage als Tis fom of Conca on te cha stand poi because tee of equipment fog secuntcly stated on te Photo 1, Uniform corrosion at 304L heat exchanger tube in acidic am- ‘onium bisuiphate environment. Photo 3. Ruptured tube (316L) out of «falling film evaporator of fertilizer plant. basis of comparatively simple tests and/or means of non-destructive test- ing methods. In case of a corrosion allowance of eg. 3mm a uniform corrosion rate of 0.15 ‘mm indicates life prediction of about 20 years. Coirosion rate of uniform corrosion can be expressed in several ways ‘* mm/year. + g/m?shr. = wm/48he. + inchiyear In Table 1 several ways of expression of corrosion rates with conversion fac- tors are presented. Preventive measures Uniform corrosion can be prevented cor reduced by: * Application of proper material + Application of a corrosion allowance. + Use of coating systems. + Addition of inhibitor systems. Photo 4. Top and bottom section out of carbon ste! reboler tube. + Application of cathodic (or anodic) protection. Case history of uniform corrosion. Photo I shows a wrong application of 3041 (X2CrNi19-11) heat exchanger tube severely corroded in acidic am- montium bisulphate envionment at 90°C. Corrosion did occur within 3 ‘weeks on-stream time, Original tube wall thickness was 1.6 mm. Photo 2 shows a uniformly corroded overflow plate out of the NOx absorp- tion column ofa nitric acid plant. Environment: 55% nitricacid con- taminated with NOs at temperature of 60°C; on-stream time 10 years. Photo 3 shows a ruptured tube (3161) ‘out of a falling film evaporator of a fertilizer plant. Rupture ozcurred after serious internal uniform corrosion. Passive corrosion rate in this example ‘was about 0.2 mm/year. The tube rup. tured after about 10 years of service. | Photo 2. Uniformly corroded overflow plate cut of the NOx absorp~ tion column, aForms or Corrosion plant. ‘These kind of catastrophic failures can be avoided by plugging the tube before rupture, based on wall thick= hess measurements by means of eddy current technique. Uniform corrosion will generally not occur uniformly over the entire surface of an equipment or even a heat exchanger tube. Photo 4 shows ‘two pleces of a carbon steel reboiler tube: one not corroded piece out of the lower section of the tube and the plece with serious thinning out of the upper section of the tube. In the upper section of the tube the wall tempera- ture as well as the liquid/gas velocity is Photo 5. Suction line (3041) inthe ammonia absorber ofa fertilizer fertilizer plant. higher, resulting in a higher corro- sion (erosion) rate, Environment cyclohexanone? cyclohexanol/water contaminated with salts of organic acids at temperature of about 130°C. (On-stream time 12 to 15 years Photo 5 shows 2 304L suction line in the ammonia absorber of low pressure section ofa fertilizer plant, Environment: high concentration of nitric acid in ammonium nitrate, The (nternal) ine was found in this condi- tion during inspection after about 15 years on stream time. Sometimes uniform corrosion is occur- ting locally while other areas are not ney Ce sinh pintzan gfdn24h rmoldne2ah need rratent2an Ibyne2sh baer mmhesr saith sarvtsh incest ty inchesiemt om) isiyea (ps sniginont (rem co ooo Table 1. Expressions of corrosion rates with conversion factors, 22 adensity (specif gravity) ofthe metal Examples: AIS! 304() stainless steel = 7.9 blanium = 45 aluminium =2.7 Photo 6, Local uniform corrosion in AIS! 347 wall ofa reactor in @ corroding as result of formation of 4 protective layer. Photo 6 shows an example of local uniform corrosion, occurring in the wall of a precipitation reactor In NP plant, Material of construction: AISI 347 (X6CININDI8-10). Corrosion did oc- © cur in a couple of days and was due to presence of amido-sulphonic acid in ammonium sulphate originating from a caprolactam plant. Environ- ‘ment: acidic nitrate/ phosphate/sul- phate solution at 65°C. (FPO, HNOs, (NH,);$04, CaS09),.2 GaLVANIC CORROSION ‘METAL CORROSION) sroduction jpotentaldiference usually exists tween two dissimilar metals when “they are immersed in a corrosive or ‘conductive solution. If these met- fais are placed in contact (electrically connected), this potential difference ‘produces electron flow between them. ‘Co:rosion of the less noble or less ‘corrosion-resistant metal is usually {increased and attack of the more resistant metal is decreased, Thhe less fesistant metal becomes anodic and ‘the more resistant metal cathodic. Usually the cathodic part corrodes ‘yery little or not at all in this type of galvanic couple. ‘The so-called dry-cell battery as de- picted in figure 1 is a good example of galvanic “corrosion”. The zinc case ‘ets as the anode at which the anodic reaction occurs and as a consequence corrodes. ‘The carbon electrode (center post) acts as a noble of corrosion-resistant metal, the cathode. The moist paste, MnO; ina NH,Cl-solution, between the elec- trodes is the conductive and corrosive environment that carries the current. ‘The following reactions take place: Anode: Zn € Zn +2e Cathode: MnO; + 4H* + 2e © Mn‘ + 2H,0 During current delivery in an accu mulator the following reactions take place: ‘Anode: Pb © Pb" + 2¢ Cathode: PbO; + 4H" + 2e 62 Pb"? + 2H,0 Mechanism of galvanic corrosion ‘The mechanism of galvanic corrosion ‘can be ilustrated strikingly by means of polarization curves. In figure 2a galvanic cell consisting of an iron. electrode and a zinc electrode in acid solution is schematically illustrated. In this figure the anodic polarization curves of iron and zinc are presented as occurring separately in the solution. ‘The cathodic curves of the hydrogen reactions at the iron and zinc elec- | trodes are also indicated: branches A and B respectively. At the intersection points of the cathodic curves and the anodic curves we find the corrosion potentials and the belonging corrosion current densities. In the figure is also sndicated the situation if the iron and zinc are coupled together. It appears that a new situation will establish, In this new situation the equilibrium potential of the couple iron + zinc is ‘more negative than the equilibrium, potential of only iron. ‘The dissolution reaction of iron: Fee Fe" + 2e cannot develop anymore and corro- sion of iron will stop. However comro- sion of zine increases. Actually iron behaves as a cathode and zinc behaves as an anode. lron is cathodically pro- neds: tne Zo +28 camo: MnO, «4462869 Mee #2 carson Zoe carter pet Cane (catnoce) (ened) tected by means of zine, Zinc is called the sacrificial anode, The driving force for current and corrosion is the potential difference between the two metals or alloys. The potential differences between metals under reversible, or non-corroding, conditions form the basis for predict- ing corrosion tendencies, Since most engineering materials are alloys, galvanic couples usually includes ‘metallic alloys. The galvanic series of some commercial metals and alloys as presented in Table 1, yields a more or less accurate prediction of galvar mages 19 tin olay zamek 7400, 038 tine 278 umm 38.598 087 rid ste 00 caren G22 0s 1396 Crstal ae) a 030 10¥6Gr9%6 Nestea lot) | epee 090 loot 900% 085 ress 60-40, oor copper +910 oak 4012 700 eprnihe +034 heomien an hromium neta steals ipsstel axon 40.40, Figure 1. Drycell battery (after Fontana) “Table 1. Practical galvanicserles for a number ‘of metas and alloys inal saturated, neutral seawater Eg, va Fear. cathodic addition current densities Ea, Fare ‘anodic addition Ear, 2nsre current densities Fasr.2n 109 eon Fe 109 cor, zn couptes with Fe Figure 2. Explanation ofthe behaviour of he galvanic coupling of iron and zn In aie sol tion by means of schematic polarization curves. 2aForms of Corrosion corrosion. In most practical applica- tons there is little danger of corro- sion if the metals and alloys are close together in the galvanic series and the potential difference generated by these couples is not great. The further apart in the series, the greater the potential generated and the greater the risk of galvanic corrosion. One should be aware of course that in different cor- rosive environments the relative post= tions of metals and alloys can change. In absence of actual tests in a given environment, the galvanic series gives ‘a good indication of possible galvanic effects Environmental effects ‘The nature and aggressiveness of the environment determine to a large ex tent the degree of galvanic corrosion. Generally the galvanic corrosion rate will increase more extensively with Inceeasing aggressiveness of the em ronment. In case of low aggressiveness of the environment the galvanic effect fon the corrosion rate may be very low. In “dead, oxygen free recirculating hhot water of a heating system a brass coupling in carbon steel pipelines will hardly influence the corrosion rate of the carbon steel. Galvanic corrosion also occurs in the atmosphere. The severity depends largely on the type and amount of ‘moisture present. For example, cotro- sion Is greater near the seashore and in industrial atmosphere than in dry rural atmosphere. Condensate near a seashore and in industrial atmosphere contains corrosive compounds like salt and fs therefore more conductive and corrosive and a better electrolyte than condensate in an inland loca- Photo 1. Galvanic corrosion in carbon stel Tjint next to brass ft. ting. 24 ton, even under equal humidity and temperature conditions. Galvanic corrosion does not occur when the metals or alloys are com- pletely dry since there is no electrolyte to carry the current between the two electrode areas. Sometimes the potential reverses for ‘given couple in different environ- ‘ments. Usually both zinc and steel corrode by themselves, but when they are coupled, the zinc corrodes and the steel Is protected. In the exceptional case, such as some domestic waters at temperatures over 70°C, the couple reverses and the steel becomes anodic. Apparently the corrosion products on the zine, in this case, make it act as a noble surface to steel. Distance effect Accelerated corrosion due to galvanic affects is usually greatest near the junction, with attack decreasing with Increasing distance from that point. The distance affected by galvanic corrosion depends on the conductiv- lay of the solution, In high-resistance environments, e.g. pure water, the at tack may be a sharp groove. Galvanic corrosion is most likely to occur near the junction of the two different metal or alloys. A weld junction is generally more suspected than a flange connec- tion. Area effect Another important factor in galvanic coftosion is the area effect or the ratio of the cathodic to the anodic areas. An unfavourable area ratio consists ofa large cathode and a small anode. For a given current flow in the corrosion cell, the current density is greater for a small electrode than for a larger one. The greater the current density at the anodic area the greater the corrosion rate, This area effect can be demonstrated by on the one side steel plates riveted with copper rivets and on the other side copper plates with steel rivets Doth immersed in seawater. After 2 years of exposure we will see some Corrosion at the steel plates but in the copper plates the steel rivets are com- pletely corroded. ‘This kind of corrosion may occur in carbon steel vessels locally lined or cladded with stainless steel. It bap- pens that bottom or top portions of carbon steel columns are lined with, stainless steel. In the transition area = of carbon steel stainless steel gal vanic corrosion may occur. If we try to solve this problem by painting the carbon steel in this transition area the result is expected to be disastrous, In general coatings will have some porous defects. This will result in very small anodic areas in good electrical contact with the large stainless steel © surface. The high current density at the small anodic area will lead to rapid. perforation of the carbon steel vessel ‘This example helps to understand the following statement related to coat- ings: “If one of two dissimilar metals in contact is to be coated, the more noble or more corrosion-resistant metal should be coated.” Beneficial applications Galvanic corrosion has several benefi- ial applications. Dry cells and other primary batteries derive their electric power by galvanic corrosion of an electrode. Photo 2. Galvanic coresion at stainless stel blind plate due to carbon seposits