Even limiting consideration to seawater is not straightforward as this varies geographically and is also affected by biological activity. The large oil companies invest significantly in metallurgy and corrosion control – it is interesting to observe the effect of company culture on metallurgical decisions: there is a line in the North Sea, to the left of which bronze is suitable for hydrants, but to the right only titanium will do! This shows that there are many solutions to similar problems.
What follows is a broad generalisation and is intended to be helpful as an overview. It is drawn from a variety of sources, most of which are listed in the bibliography (Appendix Section). It is frustrating that although much has been written and researched on comparative corrosion, it is inevitable that the comparison that you are searching for is not available directly, so a certain degree of extrapolation is necessary. Inevitably the conditions compared are different – often it is difficult to tell if this is significant or not. The following summary aims to be helpful by being giving a general picture, however the specifics in any particular situation should always be considered.
General corrosion
Most of the materials considered do not have a problem with general corrosion – except for carbon steel and cast iron where protection in the form of coating is required.Pitting and Crevice Corrosion
Pitting is a localised form of attack in quiet seawater, resulting from non-uniformities in the environment. This is a significant differentiating factor between the materials considered. Tests giving a critical pitting temperature (CPT) or a calculation to assess the pitting resistance to chloride pitting and crevice corrosion for stainless steels, the pitting resistance equivalent number (PREN= %Cr +3.3x%Mo + {16 or 30}x%N) can be used. These will give an indication of pit initiation, rather than pit propagation. Worst affected are the steels, as the alloys become “higher” the problem disappears. It is worth noting that pit propagation on duplexes can be more severe than on austenitic materials, and that pitting and crevice corrosion resistance of 22Cr duplex in waters containing high levels of chloride is poor. Precautions are advised in the case of contact with untreated seawater (Smith, Celant & Pourbaix, 2000; HSE Safety Notice 3/2003, Norsok M-001).
NAB is not considered to be affected by chloride pitting, attack in crevices is minimal and is reported to show no tendency to chloride stress corrosion cracking. (Oldfield and Masters, 1996)
Velocity effects
The types of corrosion covered here are erosion-corrosion, cavitation and impingement at higher velocities and fouling from marine organisms at low flows.
Fouling occurs when marine organisms attach themselves to the material. Due to differential aeration effects, this sets up a corrosion cell. Copper alloys are considered to be good in this situation as copper is inhospitable to many of the organisms. Ni-Cu has some fouling resistance and the remaining alloys have a very low resistance. (Tuthill & Shillmoller, 1965)
As velocity increases, the flow of oxygen to the metal surface is increased and this can has a significant effect on corrosion rates. Galvanising extends life only by about 6 months on carbon steel (Todd). Higher flow rates, particularly if the flow medium contains abrasive particles can also strip off protective oxide films. This is the case with copper alloys where there are velocity limitations. Sources vary in what these limits are, but 4.3 m/s is frequently quoted as the limit for NAB with a recommendation of 10 m/s as an intermittent maximum, (Norsok M-001) whereas 23 m/s is a guideline for the peripheral velocity of pumps and propellers (Tuthill, 1987).
Temperature
In general, temperature will accelerate the rate of any chemical reaction, so corrosion processes take place more quickly. However, raised temperatures also leads to a lower oxygen content which can have the opposite effect. The effective corrosion rate is therefore a balance between these two factors. This is significant for the crevice corrosion of stainless and duplex stainless steels where a limit of 20 oC is recommended for 6 Mo, 22Cr & 25Cr materials in seawater applications with crevices. (Norsok M-001, with maximum free chlorine of 1.5 ppm). Other temperature recommendations are minimum temperatures of –46 oC for 22Cr, -30 oC for 25Cr and in seawater maxima of 15 oC with crevices and 30 oC without crevices. (Tystad, 1997).
As can be expected, the corrosion processes are accelerated in NAB, but no particular adverse effects are noted.
Galvanic Considerations
Galvanic corrosion results from the connection of two different metals by an electrolyte, such as seawater. The corrosive effect is in proportion to the distance apart on the galvanic series, or the difference of the potentials of the two materials. This effect is significant and the reason why valve material is frequently determined by the selection of the piping material. Relative exposed areas are highly relevant.
In general NAB is more anodic than the other materials considered (except for carbon steel, cast iron and bronze) and is therefore more likely to corrode in contact with the other materials. This is emphasised by the recommendation that NAB should not be coupled with 25Cr in natural seawater (Francis, 1999).
Dissimilar materials can also be chosen to maximise protection. One such case is the use of NiCu alloy trim with NAB body material. The NiCu is protected and the large surface area of the NAB ensures that the corrosion goes unnoticed.
Other Corrosive Conditions
NAB should not be used in polluted seawater conditions due to the presence of hydrogen suphide. Precautions are also advised for duplex materials (Smith et al, 2000; Norsok M-001).
In general, NAB is good in acidic environments, but as strongly alkaline environments remove the protective film, corrosion rates in this case can be high.
Corrosion Summary
Building on earlier work (Oldfield and Masters, 1996), the tables below are a summary of the relative corrosion of the materials considered. The scale is arbitrary and is intended to convey the overall performance of the various materials under the appropriate headings. It is useful if the scale is used as intended, a ranking system, rather than a detailed comparison (10 does not mean twice as good as 5). Although not corrosion, wear and galling performance has been added. The excellent NAB properties in this area make the life of valve designers and users significantly easier.
Rather than identify an overall winner, the strengths and weaknesses of the various materials are identified. None of these properties can be properly considered without a view of the costs and intended life as well
Arbitrary scale, higher is better |
General Corrosion |
Pitting Corrosion |
Crevice Corrosion |
Erosion Corrosion |
Cavitation |
Stress Corrosion |
Bronze |
8 |
9 |
9 |
7 |
5 |
|
NiAlBrnze |
9 |
10 |
8 |
8 |
8 |
10 |
Ni-Cu Alloy |
10 |
5 |
2 |
10 |
8 |
? |
Carbon |
3 |
3 |
|
|
2 |
|
Stainless |
10 |
4 |
3 |
10 |
7 |
8 |
6Mo |
10 |
9 |
8 |
10 |
8 |
8 |
Duplex |
10 |
5 |
4 |
10 |
8 |
9 |
Superduplex |
10 |
9 |
8 |
10 |
8 |
9 |
Ni Alloy 625 |
10 |
13 |
12 |
|
13 |
|
Ni Alloy C22 |
10 |
14 |
|
|
10 |
|
Titanium |
10 |
15 |
10 |
|
9 |
|
Arbitrary scale, higher is better |
Polluted Seawater |
Corrosion Fatigue |
Fouling Resistance |
Galvanic |
Wear & Galling |
Bronze |
|
|
10 |
5 |
10 |
NiAlBrnze |
4 |
9 |
8 |
6 |
10 |
Ni-Cu Alloy |
? |
? |
4 |
8 |
5 |
Carbon |
|
|
|
1 |
8 |
Stainless |
4 |
6 |
1 |
4/7 |
6 |
6Mo |
9 |
6 |
1 |
9 |
5 |
Duplex |
5 |
9 |
1 |
8 |
4 |
Superduplex |
9 |
9 |
1 |
10 |
3 |
Ni Alloy 625 |
|
12 |
1 |
10 |
3 |
Ni Alloy C22 |
|
|
1 |
10 |
3 |
Titanium |
|
|
1 |
9 |
2 |