Clearly it is advantageous to know when corrosion is occurring in operating equipment to avoid future failure. There is extensive literature on this topic. This post is a brief summary of information selected from the four references cited below. Each source includes several other references.

It is useful to think about monitoring in two ways. One is measuring operating conditions that are indirect indicators of corrosion. The other is measuring direct damage in a probe that includes the same metal as the in-service metal. The first category includes evaluating parameters such as electrolyte temperature, dissolved oxygen levels, solution pH, the concentration of metal ions in the electrolyte from the in-service metal, and the electrochemical corrosion potential of the exposed metal in the electrolyte. Measurements in the second category may provide average attack rates over the period of exposure or instantaneous rates.

Indirect indicators of corrosion often are measured by instrumentation that's already in place (or could be) because these instruments are used during normal operations. If records are kept, significant changes in "normal" values may be used to indicate active corrosion. One often used indirect parameter is electrochemical corrosion potential. These values are used to indicate proper operation of cathodic protection (CP) on underground structures or on rebar in concrete. Measuring these potentials will also indicate whether an active-passive alloy such as a stainless steel is in its passive range or not.

A traditional example of a direct method is the periodic use of ultrasonic thickness measurements on a metal in service. This is most effective when general corrosion is occurring, and therefore fairly uniform metal loss occurs. These measurements may provide only "hit and miss" results for the localized forms of attack. There are many other NDE-based detection methods.

Another widely used direct indication example is coupons of the metal of interest that are placed in the corrosive medium. These can be simple weight-loss coupons, i.e., individual coupons that are weighed on precision scales, before and then after a period of exposure so that average rates of attack over the period can be calculated. Specialized coupons can also be used that are configured to indicate if crevice corrosion, galvanic corrosion or SCC occurs. Rates are not defined with these types of specialized coupons. However, the incidence of localized forms of attack can be determined after exposure by careful visual examinations. Sometimes a lengthy exposure is required for corrosion to initiate. Summaries of other direct monitoring methods follow:

Electric resistance (ER) probes consist of two wires, strips or tubes fabricated of the alloy being monitored. One of these forms is exposed to the corrosive environment and the other is not. The principle is that corrosion will decrease the cross-sectional area of the exposed component and then its electrical resistance increases. Multiple resistance measurements overtime are made on the exposed component and from the slope of the curve generated (plus comparison to the resistance of the non-exposed component) a corrosion rate is calculated. Electrical resistance is very temperature sensitive. Compensation for temperature is achieved using the non-exposed component. Average corrosion rates of general corrosion are defined. Localized corrosion cannot be reliability measured.

Polarization resistance (PR) is a DC electrochemical technique that provides approximate, instantaneous general corrosion rate data. It is based on the fact that the corrosion rate of an exposed specimen of the metal of interest is inversely proportional to the slope of a plot of the electrochemical potential of that metal and its DC current density for small polarizations away from its free corrosion potential. Accurate rates depend on attaining multiple ideal conditions. However, PR has the major advantage of providing approximate, relative corrosion rates that can be obtained within minutes. Then the effects on corrosion of short-term but significant changes in service conditions can be closely monitored. Relatively high conductivity electrolytes are required. Localized forms of corrosion are not measured.

Electrochemical Impedance Spectroscopy (EIS) is another electrochemical-based technique that depends on the use of AC currents employed at different frequencies. The method requires sophisticated instruments and skilled interpretation of results. At present it is primarily for laboratory and not field use. Unlike the PR method, EIS can be used in low conductivity electrolytes.

Electrochemical Noise (EN) is a method that is based on detecting the small electrical fluctuations (not audible changes) that occur in potential or current when corrosion occurs on a metal surface. Sophisticated instrumentation is required and analysis of data generated can be complex. However, if used by knowledgeable personnel EN has the major advantage of being able to distinguish the initiation of pitting or other forms of localized corrosion from general attack.

There are other practical aspects of corrosion monitoring. One is where to place coupons or probes to generate the most meaningful data. Typically it is locations where the most severe corrosion is likely. Another valid practice is use of different methods simultaneously so that results can substantiate one another. It may be desirable to electrically connect probes to a central location via wiring. However, the cost of running the wiring may be much more than the cost of the probe and related instruments. Also, wire breakage may be an issue.

References -
Dean, S.D., "Corrosion Monitoring for Industrial Processes", ASM International Handbook Volume 13A, 2003, pages 533-541.

Krebs, L.A., "A Brief History of Corrosion Sensing Methods", NACE International Conference 2003, Paper 03419.

Roberge, P.R. and Klassen, R.D., "Corrosion Monitoring Techniques", ASM International Handbook Volume 13A, 2003, pages 514-518.

Applicable ASTM Standards: G4, G96, G71, G78 and G48.