There are many kinds of corrosion which can attack metals. They all can cause leaks, explosions, fire, in equipment like pipelines and pressure vessels, but also can cause cracks and embrittlement of metals used for construction. Corrosion is often identified with atmospheric rusting of iron base alloys, but that is only a subclass (although important) of one of the many possible mechanisms of corrosive attack.
Many metals and alloys have their own individual corrosion problem, but almost all of them are caused by liquid water (excepting high temperature corrosion of e.g. turbine blades) and the chemicals dissolved in it.
Each corrosion problem has its own cure (if any). Some corrosion can be controlled by electrical means (cathodic protection) or the application of a non-metallic barrier e.g. a coating. Such barriers should be 100% pore free when used in highly aggressive liquid, and since this is rarely the case, they do not form a universal cure for metallic corrosion. Sometimes the metal generates its own barrier from corrosion products (e.g. metal oxide); the metal is then in the "passive condition".
Another control method consists of adding certain chemicals to the corrosive liquid (corrosion inhibitors).
One of the most important methods is to find an economic alloy/environment combination where the alloy is in the passive condition and corrodes much less than carbon steel.
Non-metals like plastics (with and without fibre reinforment) and ceramics are mostly immune to corrosion. They are likely to replace metals in the future in many instances, but still are inferior to metals in many respects e.g. mechanical properties, joining techniques etc.
It is known that more than half of all corrosion problems could be avoided by direct application of established knowledge.
This term is reserved for straightforward dissolution of a metal in corrosive water. Theoretically, corrosion is evenly spread over the surface of the metal. Also referred to as "weightloss corrosion" or wall thinning. Example: dissolution of steel in HCl. Simple systems like this serve to demonstrate the electrochemical basis of corrosion reactions, e.g.:
anodic reaction: Fe ----> Fe++ + 2 e- (e = electron)
cathodic reaction: 2 H+ + 2 e- ----> H2 (hydrogen gas)
Part of the exposed surface (normally one half) supports the anodic reaction; the remainder supports the cathodic reaction. The rates of these reactions adjust themselves until electrical equilibrium is obtained. Each reaction has an electrochemical potential associated with it; at equilibrium these are equal. Evenly distributed corrosion is uncommon in practice.
When different metals are exposed to a corrosive medium while in metallic contact with each other, an electric cell is formed. The current increases the corrosion rate of the least noble (base) metal (e.g. zinc), while the more noble one (e.g. copper) corrodes less, and supports most of the cathodic reaction. This is the basis of corrosion protection with sacrificial anodes. The effect depends on how far apart the metals are in the electrochemical series.
Occurs when two pieces of metal, which are in metallic contact, are in different environments e.g. different concentrations of corrosive agent. Also sets up an electrical cell current which increases corrosion of electrode immersed in smallest concentration; a differential concentration cell. Happens often under deposits of sand, tar, debris etc., in combination with the presence of dissolved oxygen.
Very important in natural gas industry. CO2 causes severe corrosion of carbon steel, often in the form of sharp-edged holes. This is sometimes called "mesa" corrosion (mesa = table, also name of sharp-edged mountains of this shape).
This nomogram for predicting CO2 corrosion of carbon steel is based on the deWaard-Milliams equation. Although it often gives surprisingly good results,there are many cases where it only reflects part of the story. For example, the influence of pH and liquid velocity on CO2 corrosion rate are interrelated in a rather complex manner.
This term is reserved for stainless steels, where welding and subsequent heat-treatment can lead to depletion of Chromium, which is scavenged in the form of carbides and in this form is no longer effective in reducing the corrosion rate.
A special form of CO2 corrosion, caused by partial annealing of carbon steel pipe after upsetting the ends to enable thread cutting.
H2S can cause pitting corrosion on several alloys. In most cases a protective sulphide is formed; weak spots in this layer can lead to enhanced corrosion. With ferritic steels, dissolved H2S causes H-atoms to diffuse into the metal. The reason for this is that sulphides interfere with the reaction of H-atoms, originating from corrosion reactions, to H2 molecules. This causes the metal to become brittle. There are two forms of damage:
HIC is caused by high pressure H2 gas bubbles which grow at intermetallic impurities like manganese sulphide stringers. Also referred to as "stepwise cracking". SSCC is controlled by keeping the hardness of the steel below 22 Rockwell. For SSCC reference is often made to a guidance (as opposed to solid advice!) document from NACE (MR0175) which also specifies the amount of H2S which can be harmfull. HIC is controlled by specifying low sulphur steel.
Caused by erosion of a protective layer (oxide, carbonate, sulphide) by erosive liquid flow. The metal then tries to grow a new layer via corrosion, which is removed again etc. Has little to do with erosion of the metal itself.
The mechanical process of fatigue crack propagation can be enhanced when the metal is immersed in a corrosive medium e.g. water. This effect can be reduced by application of cathodic protection, although by overdoing it, it becomes worse because of Hydrogen Embrittlement.
A mode of corrosion of stainless steels (or better: alloys in passive condition). Caused by attack of chlorides on a protective (passive) layer, which is enhanced by presence of oxidising chemicals. Anodic reaction occurs at weak spots, remaining surface supports cathodic reaction. Remedial measures: more molybdenum in alloy, more Cr and Ni, cathodic protection.
Corrosion of stainless steels at spots where two surfaces are pressed together, e.g. steel to plastic, or to the same steel (bolts and nuts!). Caused by loss of passivity in the resulting water film at these crevices by a differential concentration cell.
In some alloys one of the components can corrode selectively. Example is brass, where the zinc sometimes dissolves preferentially, leaving behind the red copper matrix which is weak and brittle. For badly understood reasons, the remedy is alloying with 0.1% arsenic.
Microbes (bacteria) can cause corrosion, even on stainless steels. The bacteria do not directly eat the metal, but their waste products are corrosive. They also can cause the development of differential concentration cells, leading to pitting. The most common type are the sulphate reducing bacteria or SRB's, which convert sulphates to sulphides and H2S. Counter measures are: not giving them anything to eat (sulphates, nitrates, sources for carbon which they also need), and killing them by injection of bactericides or chlorination. This is difficult because they hide behind slime deposits which they create.
As the result of heating, an alloy (e.g. stainless steel) can suffer local (normally the grain boundaries) depletion of an essential element for corrosion protection (e.g. chromium). This is called sensitisation. When such an alloy is exposed to oxidising media, the corrosion proceeds along the grain boundaries and the alloy disintegrates into grains.
Specific combinations of alloy and environment can lead to SSC, when the metal is mechanically stressed while being exposed to this environment (the stresses can also result from the fabrication process!). The metal then fails at a load far below its nominal mechanical strength. Almost every alloy and metal has its specific enemies in this respect!
A common example is that of stainless steel in hot chloride containing water. When there is a sufficient supply of dissolved oxygen, a network of branching transgranular cracks results. More oxidising conditions i.e. more anodic potentials enhance this type of stress corrosion, which is therefore called "anodic cracking". Intergranular cracking where cracks follow grain bounaries is also possible. Cathodic SCC also exists, and is caused by hydrogen embrittlement.
A generic name for all kinds of problems caused by the dissolution of hydrogen in metals, which causes loss of ductility. One of the most important examples is the SSCC, caused by the presence of H2S.
Low alloy carbon manganese construction steels are always among the first candidates to be considered for a project, but the corrosion resistance of these steels is limited, and more expensive corrosion resistant metals and alloys (CRA's) may be needed. A few examples:
An effective way of improving the corrosion resistance is to alloy the steel with chromium. Depending on the corrosive environment, even a few percent of Cr can give a substantial improvement and at levels above around 12%, corrosion rates can drop to virtually zero. This "passivation" results from the formation of a very thin, stable layer of chromium oxides at the steel surfaces.
The 12-13% Cr alloys (e.g. AISI 420) are the lowest alloyed and consequently, the least expensive in the family of stainless steels. Their structure can be ferritic or martensitic, depending on the carbon content and they have high strength.
An increase in the chromium content to 15-20%, combined with the addition of nickel in the range 8-10%, has the effect of stabilising the austenitic structure down to room temperature. This is very beneficial in terms of weldability since it means that the steel no longer has to undergo a phase transformation on cooling and hardenability is no longer a concern.
The austenitic alloys are non-magnetic and are characterised by excellent ductility and toughness. Molybdenum is commonly added in the range of 2-3% to improve the resistance to pitting corrosion (AISI 316).
Alloys with less nickel (~5%) than the austenitic stainless steels results in the austenite phase not being completely stabilised, leading to the formation of a microstructure with approximately equal amounts of austenite (non-magnetic, fcc crystal structure) and ferrite (magnetic, bcc structure) grains. The proportion of each phase depends on the exact composition and on the cooling rate during production or heat treatment. These steels are stronger, but they have some drawbacks with respect to welding and stress corrosion cracking.
Copper-base alloys are often used for piping water, and for marine environments, because of their corrosion resistance and resistance to biological fouling. Brass (with various amounts of zinc) has a problem with erosion corrosion at liquid velocities above about 1 m/s. Another problem can be de-alloying. Copper alloys can exhibit severe Stress Corrosion Cracking in the presence of ammonia. For more severe service copper-nickel alloys are often a good choice e.g. 90/10 Cu/Ni.
Nickel base alloys are used when superior strength and corrosion resistance are needed. Examples are 254SMO (19Cr 18Ni 6Mo), Incoloy 825 (23Cr 46Ni 3.5Mo) and for absolute top corrosion resistance e.g. Hastelloy C-276 (16Cr 55Ni 17Mo 4.5W). These alloys are very expensive!
Nickel base coatings can be applied by electrodeposition or by electroless bath processes. These coatings can be hardened for wear resistance.
Titanium is an extremely reactive metal, which for this reason passivates in most environments and is virtually inert unless conditions are very reducing. Ti alloys are attractive for use in seawater, but cathodic protection embrittles them. Subject to various forms of stress corrosion cracking. For overview, see link given below.
Various chemicals can be injected to reduce the corrosivity of an environment. Problem is to prove that they work under practical conditions, and that the injection is not stopped for any logistic reason. Many of them have side effects like emulsion forming. How they work is badly understood; most of them are supposed to form a protective water-repellant film with or without the aid of the inhibitor solvent. Some inorganic inhibitors induce passivity. CO2 corrosion in wet gas pipelines can be controlled by adding certain glycol compounds, which also has the advantage of controlling "top-of-line corrosion" in two-phase (gas/liquid) situations.
By lowering the potential between metal and water, the corrosion reaction rate can be greatly reduced. To this end an electrical current of the correct polarity is injected into the metal. This can be done by electrical connection to sacrificial anodes made of a more anodic metal, setting up a galvanic cell (virtually shorted by the conductivity of the water) which increases the corrosion of the sacrificial anode and reducing the corrosion of the metal to be conserved. Examples: steel/magnesium, steel/aluminum, copper/zinc. Alternatively, the current can also be provided from a battery or rectifier. This is referred to as cathodic protection by impressed current.
For the protection of offshore structures the use of sacrificial anodes is the preferred method. Although there are exceptions, these structures are normally made from bare (not painted) steel, and protection relies completely on cathodic protection. The design of such a system involves size, shape, material and quantity of sacrificial anodes.
MATERIALS SELECTION TAKES A STEP FURTHER IN TERMS OF QUANTIFICATION!
The "Electronic Corrosion Engineer" program ("ECE") can be your first line adviser for the evaluation of corrosion risks for a flowline or for well production tubing handling wet corrosive oil or natural gas - two key expenditure items for the oil and gas production. It advises on the possibilities of corrosion control for carbon steel, on the suitability and availability of Corrosion Resistant Alloys, and on the economics of the various options by means of Life Cycle Cost calculation. It has been developed as a joint effort of CorCon Corrosion Consultancy and Intetech.Ltd. Click for more information
|Dr. C. de Waard           (who is he?)|