Welding Supermartensitic Stainless Steels for Corrosive Service
Paper presented at Stainless Steel World Conference, 6-8 November 2007, Maastricht, The Netherlands.
Academic Education and Degrees:
1990 PhD on fatigue in nickel superalloys, Cambridge University
1984 BA in Metallurgy and Materials Science, Cambridge University
Present professional position:
Manager of Metallurgy, Corrosion & Surfacing Group, TWI
Supermartensitic stainless steels (SMSS) are attractive materials for flowlines requiring resistance to CO 2 corrosion but service failures resulting from intergranular stress corrosion cracking of girth weld HAZs have occurred, leading to uncertainty over their widespread applicability. This paper presents information on the corrosion resistance of welds in SMSS and gives advice on choice of welding procedures to produce welds that are not sensitive to intergranular corrosion and have good pitting corrosion and sulphide stress cracking resistance.
SMSS pipes have been selected by the oil and gas industry for their resistance to deoxygenated brines acidified by CO 2 . They can tolerate a low level of H 2 S but are not generally resistant to highly sour service.
Welded SMSS can suffer from intergranular stress corrosion cracking (IGSCC) of the girth weld HAZ when exposed to sweet or mildly sour environments at around 70-140°C and in-service failures of lean grade steel have occurred  , in additional to laboratory test failures of lean and high grades.  Parent steel in the normal supply condition is apparently not susceptible. If SMSS pipes are to be used successfully at high temperatures in acidic conditions, the girth welds must not be susceptible to IGSCC.
Chloride ions encourage localised corrosion, i.e. pitting and crevice corrosion, as does H 2 S.  Pitting susceptibility is also dependent on oxygen levels although, for pipelines, interest is in oxygen-free media. The pitting resistance of girth weld roots exposed to acidified brine with or without H 2 S is therefore of concern and welds should be designed to have adequate resistance for the intended service.
Hydrogen sulphide can be tolerated only at fairly low levels by SMSS, as it encourages sulphide stress cracking (SSC). Concern with respect to SSC of welded SMSS pipes has been primarily in relation to resistance at ambient temperature.  Weldment performance is significantly worse than that of the parent steel and the areas of greatest susceptibility are HAZ regions where hard, untempered martensite may exist.  It is essential that the welds are designed also to give adequate SSC resistance.
When considering how to weld SMSS, these three potential failure mechanisms from the internal fluid need to be considered to ensure reliable service, with susceptibility to IGSCC being unacceptable. Seawater corrosion resistance may also be of concern when exposure cannot be avoided, although these materials are generally not exposed to seawater without cathodic protection, as their resistance to seawater is very low. This paper considers choice of welding procedure to control corrosion resistance relevant to the internal pipe environment. The various corrosion/stress corrosion mechanisms are considered in turn, with initial attention paid to intergranular SCC, which is a particular area of concern, as it has caused in-service failures.
In general, there are four main factors that control the corrosion resistance of welds in a corrosion resistant alloy: (i) the composition of the parent steel and weld metal, (ii) the microstructure of HAZ, weld metal and fusion boundary, which is controlled by the weld thermal cycle  and choice of welding consumable, (iii) the level of weld oxidation and (iv) the presence of geometric features that act to encourage initiation of corrosion, e.g. crevices. All of these can be controlled to some extent by the welding procedure.
2 Avoiding intergranular SCC
2.1 As-welded condition
The likelihood of IGSCC occurring in as-welded SMSS in service will depend on a combination of the degree of sensitisation of the HAZ and the severity of the environment. Very aggressive environments will tend to cause general corrosion rather than intergranular SCC and more benign environments may cause neither.
The most commonly observed macroscopic cracking mechanism follows prior austenite boundaries, suggesting that formation of chromium-depleted zones adjacent to M 23 C 6 carbides on prior austenite grain boundary provides the main mechanism for sensitisation, Fig.1. At least two thermal cycles are required, with the first thermal cycle forming fresh martensite and the second and subsequent cycles forming the damaging precipitation. Low heat input and low interpass temperatures appear to be detrimental for this particular mechanism of sensitisation. However, conclusive proof of damaging precipitation and chromium-depleted zones has proved elusive in high grade steel, especially when titanium is added.  A second mechanism of sensitisation has been proposed in high grade steel with Ti addition, involving sensitisation due to phosphorus segregation to grain boundaries.  Also, several welds have shown IGSCC coincident with the band of retained delta ferrite in the HAZ running parallel to the fusion boundary. This suggests that a third sensitisation mechanism might exist associated with carbide precipitation on delta ferrite/martensite phase boundaries. It is not clear whether this mechanism also requires two or more thermal cycles although it is apparently favoured by longer thermal cycles and greater numbers of welding passes.  This mechanism does not necessarily give corrosion/cracking following the prior austenite boundaries, Fig.2.
Fig.1. Intergranular crack in HAZ of welded 12Cr6Ni2Mo steel, tested at 110°C, in 25%NaCl, 10bar CO 2
Fig.2. Shallow intergranular crack in band of ferrite in HAZ of welded 12Cr6Ni2Mo steel, tested at 110°C, in 25%NaCl, 10bar CO 2
One consequence of the proposed main mechanism of sensitisation of prior austenite boundaries is that although higher heat input appears to be beneficial, particularly for high grade steel, as it apparently allows 'healing', this beneficial effect might be removed if the heat input of any subsequent pass was sufficiently high to re-transform the root surface to austenite, and hence form fresh martensite on cooling. In this case, the material would then havebeen returned to a condition in which it may be sensitised by further passes. Significant variation of heat input between passes may therefore contribute to sensitisation. For this reason, manual welding is not ideal for producing welds that are reliably free from sensitisation to IGSCC. In this respect, it is noted that when lean grade pipe failed by IGSCC in service, it typically coincided with the side of the weld on which the final capping pass was applied.  The capping pass probably reheated material into the critical range for sensitisation, whilst several if not all of the previous passes may have reheated the HAZ to >Ac 1 .
In principle, it should be possible to select welding parameters that produce welds that are not sensitive to IGSCC, based on knowledge of the critical thermal cycles that must be avoided. Published data provide evidence of the damaging thermal cycles but the allowable range is not well defined for the various SMSS grades. [7,9] Also, the wide range of thermal cycles experienced in a multi-pass weld HAZ, and the difficulty of accurately measuring or predicting them, imply that reliably avoiding IGSCC by control of the weld thermal cycle will be very difficult, if not impossible. An empirical approach, based on trying to avoid rapid cooling of the root during second and subsequent passes, superficially offers the best chance of producing as-welded welds that are not sensitive to IGSCC although sensitisation of the ferrite seems to occur for longer thermal cycles.  However, even if such welds were produced and qualification corrosion tests showed acceptable performance, the lack of understanding of the range of applicability of the result, i.e. what range of heat input and interpass temperature would be acceptable in production welding, make this approach unreliable.
A more innovative approach would be to adopt a 'single-shot' welding process, such as laser, electron-beam, or friction welding, so that only one thermal cycle is experienced. All available data suggest that two weld passes are required for any of the proposed sensitisation mechanisms. Nevertheless power beam welds involve a region of overlap at the start/stop location where two thermal cycles will be experienced and more data are required to demonstrate whether such processes could produce welds that are not sensitised.
Surface condition and weld geometry probably contribute to initiation of IGSCC and cracks have been observed to initiate at laps formed at the weld toes during GMA welding onto a copper backing bar. In practice, efforts to achieve low oxidation levels and a smooth weld toe are considered prudent, to reduce the likelihood of initiating IGSCC.
Despite the apparent difficulty in producing welds that are not sensitised, efforts to recreate sensitisation and IGSCC in the laboratory have met with mixed success  and indeed there have been several applications where girth welded SMSS has been successfully qualified without PWHT and in-service failures have not occurred. However, it may be noted that the environmental conditions required for IGSCC have not been defined, so it is not clear whether these cases have been successful because the weld was not sensitised or whether the environment was not sufficient to cause IGSCC.
2.2 Use of Postweld Heat Treatment
Several published studies have shown that PWHT at around 650°C for five minutes, applied by induction heating, eliminates intergranular SCC of SMSS girth welds.  Therefore, if any uncertainty exists over the suitability of a SMSS weld for service in the as-welded condition with respect to IGSCC, it is recommended that such PWHT should be applied, provided that an appropriate qualification process has been undertaken to ensure it does not have unacceptable detrimental effects on toughness, e.g. of superduplex weld metal, or on other aspect of corrosion resistance. Due to the limited information available, the use of welded SMSS in the PWHT condition will require qualification on a case by case basis. There are insufficient data to conclude whether 90-day corrosion testing is required but it is advisable. The qualification process should consider the extremes of the range of PWHT thermal cycles that may be experienced, as the acceptable range has not been established and there is evidence that some sensitivity to IGSCC can remain after PWHT at 650°C for five minutes if the weld is initially heavily sensitised  , Fig.3.
Fig.3. Shallow intergranular crack in HAZ of welded 12Cr6Ni2Mo steel, given PWHT at 650°C for five minutes and tested at 110°C, in 25%NaCl, 10bar CO 2
Based on current data, a PWHT temperature of 620-660°C at the root is recommended and the heat treated zone should encompass the whole of the weld metal and HAZ. The maximum allowable cap temperature has not been established but an upper limit of about 670°C is suggested.  If the PWHT temperature is much higher, retransformation to austenite and consequently fresh martensite is likely. Heating and cooling should be fairly rapid. The most appropriate PWHT duration has not been established but there is fairly common agreement that five minutes is an appropriate duration. The most suitable PWHT condition is likely to be dependent on the grade in question.
It is noted that very shallow intergranular corrosion is sometimes observed during corrosion testing of welded SMSS. This has been linked to chromium depletion on prior austenite boundaries close to the oxidised surface, due tograin boundary diffusion of chromium to the surface.  The consequences of this are not clear, but in the absence of PWHT it might act to encourage initiation of IGSCC and hence it should be minimised via use of inert gas back purging.
3 Maximising pitting resistance
Figure 4 compares the pitting corrosion resistance of various SMSS welds in 12Cr6Ni2Mo (UNS S41426) steel including ones made by the GMA, CO 2 laser and reduced pressure electron beam (RPEB) processes. Details of the materials and welds are given in Tables 1 and 2. Potentiodynamic scans were undertaken to measure pitting potential in 3.5%NaCl at 20°C. The solution was deoxygenated with (i) N 2 , (ii) 1 bar CO 2 and (iii) 0.99 bar CO 2 with 10mbar H 2 S.
Table 1 Material analyses
NA = not applicable
Table 2 Welding details for 12Cr6Ni2Mo SMSS pipe (UNS S41426)
|Heat input/power/voltage, current
|38He 2CO 2 60Ar
|38He 2CO 2 60Ar
|38He 2CO 2 60Ar
|38He 2CO 2 60Ar
|38He 2CO 2 60Ar
|CO 2 laser
|He trailing shield
The root side pitting corrosion resistance of the welds, in the as-welded condition only showed fairly small variations and results were within a range of 140mV at most for each individual solution. The use of superduplex fillerwire gave a weld with a little higher pitting resistance than ones made with matching composition 12Cr6Ni2Mo wire. Pickling did not improve the pitting performance of the welds made with matching wire, whilst it did improve the pitting resistance of the weld made with superduplex wire, indicating that preferential pitting of the HAZ probably occurs in the as-welded condition but the matching composition weld metal limits pitting resistance when pickled.
Fig.4. Pitting potentials for a variety of SMSS welds in 3.5%NaCla) deoxygenated with
(ii) CO 2 and
(iii) CO 2 with 10mbar H 2 S,
b) showing the effect of pickling (3.5%NaCl deoxygenated with N2)
For the SMSS wire, use of Ar root shielding gas or a copper backing bar was beneficial, due to reduction of the detrimental effect of oxidation. Laser welding, which had a helium trailing shroud but no inert gas root protection, gave lower pitting resistance than GMA welding but reduced pressure RPEB welding gave the best pitting resistance, as a result of the associated low level of oxidation, arising from production in a partial vacuum. The effect of oxide was emphasised by testing of pickled samples, which generally showed improved pitting resistance compared to the as-welded condition. No significant effect of heat input in GMA welds was noted over the range studied (0.55-0.97kJ/mm).It was noted that localised corrosion occurred at the weld toes in several cases, where small crevices/laps often form on welding Fig.5, and the oxidised HAZs.
Fig.5. Pit developed at weld toe in GMA weld made with 25Cr wire, low heat input, tested in CO 2 acidified 3.5%NaCl solution
No published data could be found to show the effect of PWHT on pitting resistance. However, one implication of the data showing a beneficial effect of use of inert purge gas during welding and pickling of welds is that oxidationarising from PWHT is likely to be detrimental to pitting resistance. Therefore use of some form of protection of the root, if possible, should be considered during PWHT.
4 Maximising SSC resistance
Studies have been undertaken to derive limiting conditions for SSC of welded SMSS in 5%NaCl in high and low grade SMSS, i.e. 12Cr6Ni2Mo, with and without Ti, and lean grade 11Cr1.5Ni steels.  For each parent steel, two types of welds were made (i) with superduplex (Zeron 100X) and alloy 625 (AWS E/ERNiCrMo-3) consumables and (ii) with ER 2209 duplex and 12Cr4Ni (ER 410NiMo) consumables respectively. Pipe compositions are given in Table 3. Girth welds were produced in each steel using TIG or MMA processes for the root pass and either GTA, SMA or FCA for fill passes. All welding was with the pipe fixed and horizontal, with welding vertically up. Nopre-heat was used and heat input was in the range 0.5-1.5kJ/mm. Four point bend tests were undertaken on cross-weld specimens, with the root intact, in a range of mildly sour environments at 25°C to define environmental limits.
Table 3 Material analyses
The results highlight the increased sensitivity of the weld region to SSC in a range of weldable martensitic stainless steel pipe materials Fig.6 to 9). Comparison with published data for parent steel [12,13] indicates a marked reduction, after welding, in the acceptable H 2 S level for SMSS. Highest resistance to SSC for as-welded joints was found in the steels with highest Ni and Mo content, even though these also tended to give higher HAZ hardness than the lower alloy grades.
Fig.6. Proposed SSC limits for as-welded 12Cr6Ni2Mo steel in 5%NaCl, 25°C
Fig.7. Proposed SSC limits for as-welded 11Cr1.5Ni steel in 5%NaCl, 25°C
Fig.8. Proposed SSC limits in 5%NaCl, 25°C for welded 12Cr6Ni2Mo steel after PWHT at 650°C for five minutes
Fig.9. Proposed SSC limits in 5%NaCl, 25°C for welded 11Cr1.5Ni steel after PWHT at 650°C for five minutes
Definition of limits was complicated, in particular for the high grade steel, by the observations of very short 'microcracks', typically 50-150µm deep, in many of the specimens. The origin of these cracks was not clear. Examination of sections from the welds, which had not been exposed to a test environment, indicated that small weld toe flaws (typically <50µm in depth) may form on welding. The limiting conditions for SSC proposed here arebased on the assumption that any crack <50µm deep was a result of SSC. However, it may be noted that for steel with 2-2.5%Mo and 6-6.5%Ni, cracking to a depth of >0.5mm was only observed in this test series after exposure ata pH of 3.2, with 0.016 bar H 2 S, so the proposed limits may be conservative.
For 12Cr6Ni2Mo pipe welded with superduplex or alloy 625 consumables, the data suggest H 2 S limits in 5%NaCl of around 0.035 bar at a pH of 5 and less than 0.01 bar at a pH of 4 for the as-welded condition. Limits for the 11Cr1.5Ni steel in the as-welded condition were not established, as all specimens cracked in all the tests performed. After brief PWHT (5 minutes at 650°C), the 11Cr1.5Ni steel cross-weld specimens gave very similar results to the as-welded specimens from the 12Cr6Ni2Mo steel. The effect of brief PWHTon the SSC resistance of the 12Cr6Ni2Mo steel was minor in the tests performed. Crack-depths for the 12Cr6Ni2Mo steel were typically less after PWHT but H 2 S limits based on the criterion described previously were not shifted significantly. This may reflect on the criterion used and the test conditions selected here, as previously published work on a 12Cr6Ni2Mosteel has shown a beneficial effect of a similar PWHT for a particular practical application with 0.04 bar H 2 S.  Nevertheless, PWHT seems to have a more significant effect on SSC resistance for the 11Cr1.5Ni grade, which will have fairly high Ac 1 temperature and hence will not tend to austenitise during PWHT at around 650°C, in contrast to the 12Cr6Ni2Mo grade. This was reflected in a greater hardness reduction for the lean grade. PWHT at 650°C reduced the maximum root HAZ hardness in each case, by up to 90HV for the lean grade and by up to 40HV for the high grade.
Most cracking occurred in the high temperature supercritically reheated HAZ region but some cracks were found also in 12Cr 4NI weld metal and at the fusion boundary. Most cracking was transgranular in nature but some areas of intergranular HAZ cracking were observed in lean grade steel specimens. The intergranular path was presumably with respect to the prior austenite grain structure. Fusionline cracking may be significant, as it suggests increased sensitivity to SSC in this area when a superduplex consumable is used, Fig.10.
Fig.10. Fusionline crack in SSR specimen from 12Cr6Ni2Mo pipe welded with superduplex wire, tested in 3.5wt%NaCl, 0.69mg/l NaHCO 3 with 0.01bar H 2 S, 0.99bar CO 2 , pH 4.6 at 25°C
A second test series, involving slow strain rate (SSR) tests on cross-weld specimens from welds in high grade steel welded with superduplex and matching composition wires, Table 3, showed better performance with the use of matching composition 12Cr6Ni2Mo wire, compared to a weld made with 25Cr superduplex wire, Table 4. This might be related at least in part to the lower strength of 25Cr weld metal compared to matching composition weld metal and the SMSS parent steel. Hence, yielding may have occurred preferentially in the 25Cr weld metal and this may have tended to encourage cracking in the fusionline and immediately adjacent HAZ. However, the fusionline structure is expected to consist of a high alloy martensite when superduplex consumables are used and locally high microhardness was measured in this area both before and after PWHT. Maximum microhardness was around 400HV0.05 as-welded and 370HV0.05 after PWHT for the 12Cr6Ni2Mo steel and around 420HV0.05 as-welded and 330HV0.05 after PWHT for11Cr1.5Ni steel. Also, the fusionline and HAZ material immediately adjacent to it may be enriched in nitrogen due to diffusion across the interface from the superduplex weld metal. As the parent steel is deliberately produced with low carbon and nitrogen, local enrichment of nitrogen may enhance hardness locally, possibly with an associated loss of SSC resistance.
Table 4 Results of slow strain rate tensile test performed in 3.5wt%NaCl, 0.69mg/l NaHCO 3 solution with 0.01bar H 2 S and 0.99bar CO 2 , pH 4.6 at 25°C
|Ratio of properties in air and sour solution
|Failure location in air
|Failure location in H 2 S solution
|W1 (12Cr6Ni2Mo wire, low heat input)
|W2 (12Cr6Ni2Mo wire, high heat input)
|W3 (25Cr wire, low heat input)
|El = Elongation
RA = Reduction in cross-sectional area
Hence, if the toughness and hydrogen cracking concerns associated with matching composition SMSS consumables, which caused particular problems on one subsea flowline  , can be overcome and if PWHT is to be adopted to eliminate IGSCC, there may be merit in use of matching consumables for SSC resistance in view of (i) the potential concerns relating to the fusion boundary area when dissimilar consumables are used and (ii) the potential for precipitation of deleterious phases in high alloy weld metal if PWHT is used.
An effect of the small HAZ hardness difference between two welds made with 12Cr filler was noted also in SSR tests. The weld with lower hardness, produced with higher heat input, had slightly better resistance to SSC than the harder weld.
- The control of the welding process to reliably avoid sensitisation of SMSS to IGSCC is not straightforward. Each steel will have its own range of critical thermal cycles for causing sensitisation. Mechanised welding giving reproducible thermal cycles for each weld pass provides a higher likelihood of avoiding sensitisation to IGSCC than manual welding, where a wider range of thermal cycles will inevitably result.
- Welds in SMSS that experience more than one thermal cycle during welding should be given a brief PWHT at about 650°C for five minutes at the root if the environment that they will be exposed to is sufficiently aggressive, i.e. hot and acidic, to cause IGSCC of a sensitised HAZ microstructure.
- The use of single shot welding processes such as power beam and friction welding may be considered for supermartensitic stainless steel pipes to overcome IGSCC problems associated with multiple thermal cycles.
- When welding and postweld heat treating SMSS pipes for corrosion resistant applications, the root should be protected from oxidation and small crevices at the weld toe or in the weld metal should be avoided or minimised where possible to maximise pitting corrosion resistance.
- Welding substantially reduces the SSC resistance of weldable martensitic stainless steel pipe materials, with cracking in tests occurring in HAZ and at the fusionline with superduplex weld metal.
- High grade SMSS steels showed higher weld SSC resistance than lean grades, especially in the as-welded condition. In 5%NaCl solution, approximate limiting H 2 S partial pressures of 0.035bar at a pH of 5 and 0.01bar at a pH of 4 are suggested for 12Cr6Ni2Mo steel in the as-welded condition, at 25°C. Similar H 2 S limits were found for 11Cr1.5Ni steel after PWHT at 650°C for five minutes but, in the as-welded condition, the limits were lower, with cracking noted at the lowest H 2 S levels examined, i.e. 0.03 bar/pH=5 and 0.01bar/pH=4.
- There is some evidence that the use of matching composition filler metal may have advantages over superduplex filler for SMSS, with respect to SSC resistance, although further work would be required to establish the acceptability of such an approach, particularly for subsea applications, given the concerns over toughness and hydrogen embrittlement of matching composition weld metal.
- Use of low heat input may be slightly detrimental with respect to SSC, as it increases HAZ hardness a little.
- Brief PWHT at 650°C for five minutes gave a marked improvement of SSC performance for 11Cr1.5Ni steel but had much less effect for 12Cr6Ni2Mo steel in the tests performed.
- Gregori A, Woollin P, van Gestel W, 'Analysis of in-service failures in girth-welded supermartensitic pipes', 'Stainless Steel World', 15, December 2003, 17-24.
- Rogne T, Lange H I, Svenning M, Aldstedt S, Solberg J K, Ladanova E, Olsen S, Howard R D and Leturno R, 'Intergranular corrosion/cracking of weldable 13%Cr steel at elevated temperature', Proc conf 'Corrosion 2002', NACE International, paper 02428.
- Enerhaug J, Grong O, Steinsmo U M: 'Factors affecting initiation of pitting corrosion in supermartensitic stainless steel weldments', Science and Technology of Welding and Joining, 2001, Vol 6, No 5.
- Enerhaug J, Kvaale P E, Bjordal M, Drugli J, Rogne T: 'Qualification of welded super 13%Cr martensitic stainless steel for the Asgard field', Proc conf Corrosion 1999, Paper 0587.
- Woollin P: 'Sulphide Stress Cracking of Welded Joints in Weldable Martensitic Stainless Steels', NACE Middle East Corrosion Conference, 2001, Bahrain.
- Woollin P, Carrouge D, 'Heat affected zone microstructures in supermartensitic stainless steel', Proc conf Supermartensitics 2002, KCI Publishing BV, paper 027.
- Woollin P, 'Understanding and avoiding intergranular stress corrosion cracking of welded supermartensitic stainless steel', Proc conf Corrosion 2007, NACE International, paper 07094.
- Amaya H, Ogawa K, Taniyama A, Ueda M, 'Effect of alloying element on intergranular stress corrosion cracking susceptibility of supermartensitic stainless steel welded joints in sweet environment', Proc conf Corrosion 2007, NACE International, paper 07091.
- Hisashi A, Kondo K, Taniyama A, Sagara M, Ogawa K, Murase T, Hirata H, Takabe H and Ueda M, 'Stress corrosion cracking sensitivity of super martensitic stainless steels in high chloride concentration environment', Proc conf Corrosion '04, NACE, paper 04124.
- Amaya H, Kondo K, Taniyama A, Sagara M, Ogawa K, Murase T, Hirata H, Takabe H and Ueda, M: 'Stress corrosion cracking sensitivity of supermartensitic stainless steels in high chloride concentration environment', Proc conf Corrosion 2004, USA, NACE International, paper 04124.
- Lee C M, 2006: 'The corrosion resistance of supermartensitic and duplex stainless steel pipes welded by arc and power beam processes' Eurocorr 2006, Maastricht, The Netherlands.
- Ueda M, Amaya H, Kondo K, Ogawa K and Mori T: 'Corrosion resistance of weldable super 13Cr stainless steel in H 2 S-containing CO 2 , environments', Proc conf. 'Corrosion 96', NACE International, Paper 58.
- Amaya H, Kondo K and Hirata H: 'Effect of chromium and molybdenum on corrosion resistance of super 13Cr martensitic stainless steel in CO 2 environments', Proc Conf. 'Corrosion 98', NACE International, Paper 113.
- Mollan R, 'Experience with 13Cr supermartensitic stainless steel in the Tune submarine flowlines', Proc Conf Corrosion 2005, NACE International, paper 05092.