Subscribe to our newsletter to receive the latest news and events from TWI:

Subscribe >
Skip to content

Postweld heat treatment to avoid intergranular stress corrosion cracking of supermartensitic stainless steels (November 2005)

   

P Woollin
TWI Ltd

Paper presented at Stainless Steel World 2005 Conference, Maastricht, The Netherlands, 8-10 November 2005.

Abstract

Supermartensitic stainless steels (SMSS) are attractive materials for flowlines transporting produced fluids with high levels of CO 2 and low levels of H 2 S. However, recent cracking of lean grade material in service and both lean and high-alloy grades during qualification testing have revealed sensitivity to intergranular stress corrosion cracking (IGSCC) at some girth welds [1-11] , although all flowlines in high alloy SMSS have apparently had no such problems in service. One potential solution is to use a brief postweld heat treatment, typically at around 630-650°C for five minutes, which has been shown to overcome susceptibility to IGSCC in laboratory tests. [4-11]

The paper considers existing information on the effects of brief PWHT on welded SMSS, presents additional data for a range of pipes and weld types and discusses the likely mechanism by which PWHT may be effective in preventing IGSCC. It is concluded that a microstructural effect is probably dominant. Based on this preliminary conclusion and a consideration of the potential detrimental effects of an inappropriate PWHT cycle, the necessary control of the PWHT process is addressed and recommendations are made with respect to application of PWHT, highlighting best practice based on current knowledge.

1. Introduction

Intergranular SCC of SMSS pipe girth welds represents an obstacle to the exploitation of these materials in flowlines for some applications, although they are still being used extensively by some operators. It is recognised that a reliable way to prevent susceptibility to IGSCC of as-welded SMSS via control of welding parameters and without PWHT may still take some time to develop, assuming that it is possible. However, experimental evidence to date suggests that the use of brief PWHT at around 650°C will eliminate sensitivity to IGSCC. [4-11] No examples of IGSCC have been reported after PWHT at around 650°C for 5 minutes. Such PWHT is therefore an attractive interim solution to the problem, albeit one that will add to the cost of producing welded fabrications. Nevertheless some flowlines are operating successfully without PWHT, although the ability to do this will probably depend on the operating environment. It is noted that welded SMSS is also susceptible to cracking in sour environments but this is by a different mechanism to the IGSCC discussed here and PWHT does not prevent cracking in sour environments, although it may improve resistance.

Several authors have previously examined the effects of PWHT on the properties of SMSS welds, [12,13] although only more recently has its effect on sensitivity to IGSCC been explored. [4-11] The range of PWHT treatments studied on actual welds is from 600-700°C for 3-5 minutes, [6-11] although simulated HAZ studies have suggested that a wider range of thermal cycles, from 550-700°C for 1 to 17 minutes may also be effective at eliminating sensitivity to IGSCC. [5] However, it should be noted that the heat treatment required to eliminate sensitivity to IGSCC will presumably depend its severity and the composition of the steel, notably C and perhaps N content, and the levels of other carbide/nitride forming elements such as Ti, Nb and Mo.

When specifying a PWHT cycle in practice, it is essential that it should not only provide acceptable IGSCC properties but other mechanical and corrosion properties must also be acceptable after PWHT. Therefore this study also examines effects of PWHT on microstructure, hardness and toughness.

2. Experimental work

2.1 Materials

Six low carbon martensitic stainless steel pipes with 10.9-13.5%Cr were selected, all of which could broadly be considered as 'supermartensitic' but not all representing currently commercially available grades, Table 1. Two of the steels were variants of the same grade, with very similar composition (C1 and C2). Nickel content varied from 1.55-6.4% and Mo ranged from 0-2.5% for the steels examined.

Table 1 Chemical analyses of the materials used

Pipe codeElement, wt%
CNSiMnCrNiMoCuTi
A (12Cr5Ni2Mo) 0.013 0.009 0.12 0.54 11.8 5.1 2.03 0.04 <0.005
B (12Cr6Ni2Mo) 0.010 0.011 0.17 0.18 12.4 5.8 2.18 0.03 0.020
C1 (12Cr6Ni2.5MoTi) 0.009 0.005 0.20 0.43 12.2 6.4 2.51 0.03 0.12
C2 (12Cr6Ni2.5MoTi) 0.010 0.007 0.26 0.46 12.2 6.5 2.48 0.03 0.09
D (13Cr5Ni1Mo) 0.013 0.006 0.16 0.65 13.5 5.1 0.78 0.03 0.088
E (11Cr1.5Ni0.5Cu) 0.010 0.006 0.18 1.14 10.9 1.6 <0.01 0.49 0.01
NA = not analysed

2.2 Welding

Three types of girth weld were examined, (i) three automatic pulsed MIG welds made with superduplex solid filler wire throughout (W1-W3), (ii) two automatic pulsed MIG welds made with approximately matching composition metal cored filler wires (W4 and W5) and (iii) two manual welds made using the TIG process for the root and second pass and the MMA process for the fill and cap passes, using superduplex consumables (W6 and W7).

Table 2 lists the welding consumable analyses, which were either direct analyses of the solid wires or were from all-weld metal pads deposited using the coated electrodes and the metal cored wires. Analyses were obtained by OES and inert gas fusion for O and N.

Table 2 Welding consumable analyses

Consumable codeDia (mm)Element, wt%
C Si Mn Cr Mo Ni Cu W N
C1 (SMSS, MCW) 1.2 0.009 0.67 1.22 11.9 1.49 6.6 0.48 NA 0.009
C2 (SMSS, MCW) 1.2 0.008 0.39 1.77 12.1 2.51 6.4 0.58 NA 0.009
C3 (SDSS, SW) 1.2 0.027 0.40 0.41 26.1 3.90 9.3 0.12 <0.05 0.23
C4 (SDSS, SW) 1.2 0.015 0.30 0.40 25.0 4.00 9.5 NA NA 0.24
C5 (SDSS, SW) 2.4 0.018 0.39 0.69 24.8 3.80 9.3 0.60 0.61 0.22
C6 (SDSS, CE) 2.5 0.030 0.32 0.90 24.9 3.65 9.4 0.79 0.68 0.24
C7 (SDSS, CE) 3.2 0.030 0.34 0.90 25.4 3.61 9.0 0.75 0.67 0.21
SMSS = supermartensitic stainless steel
SDSS = superduplex stainless steel
MCW = metal cored wire
SW = solid wire
CE = coated electrode
NA - not analysed

Table 3 summarises the welding matrix. All welding was in the 5G position (pipe horizontal, fixed). For the automatic pulsed MIG welding, a copper backing shoe and Ar backing gas were used and interpass temperature was restricted to <150°C, whilst pre-heat was just sufficient to remove moisture. Travel speed was in the range 350-500mm/min and heat input approximately 0.5kJ/mm. An Ar/He/CO 2 /N 2 shielding gas mixture was used for welding with the superduplex wires and Ar+0.5%CO 2 was used for the matching composition wires. A narrow gap J preparation was used. For the manual welding, interpass temperature was again <150°C, the heat input for the root and second pass was 1.1-1.5kJ/mm and for the fill and cap passes it was 0.5-1.4kJ/mm. Argon shielding gas was used for TIG welding and Ar back purge gas was used throughout. A 30° bevel was used with no root face and a 4mm root gap.

Table 3 Girth welding matrix

Weld codePipe codeWelding processWelding consumableShielding gasPWHT
RootFill and capRootFill and cap
W1 A (12Cr5Ni2Mo) Pulsed MIG Pulsed MIG 25%Cr wire C3 Ar/He/CO 2 /N 2 None
W1P A (12Cr5Ni2Mo) Pulsed MIG Pulsed MIG 25%Cr wire C3 Ar/He/CO 2 /N 2 650°C/5min*
W2 B (12Cr6Ni2Mo) Pulsed MIG Pulsed MIG 25%Cr wire C4 Ar/He/CO 2 /N 2 None
W2P B (12Cr6Ni2Mo) Pulsed MIG Pulsed MIG 25%Cr wire C4 Ar/He/CO 2 /N 2 650°C/5min*
W3 C1 (12Cr6Ni2.5MoTi) Pulsed MIG Pulsed MIG 25%Cr wire C3 Ar/He/CO 2 /N 2 None
W3P C1 (12Cr6Ni2.5MoTi) Pulsed MIG Pulsed MIG 25%Cr wire C3 Ar/He/CO 2 /N 2 650°C/5min*
W4 D (13Cr5Ni1Mo) Pulsed MIG Pulsed MIG 1.5%Mo SMSS wire (C1) Ar+0.5%CO 2 None
W4P D (13Cr5Ni1Mo) Pulsed MIG Pulsed MIG 1.5%Mo SMSS wire (C1) Ar+0.5%CO 2 650°C/5min*
W5 B (12Cr6Ni2Mo) Pulsed MIG Pulsed MIG 2.5%Mo SMSS wire (C2) Ar+0.5%CO 2 None
W5P B (12Cr6Ni2Mo) Pulsed MIG Pulsed MIG 2.5%Mo SMSS wire (C2) Ar+0.5%CO 2 650°C/5min*
W6 E (11Cr1.5Ni0.5Cu) Manual TIG MMA SDSS wire (C5) SDSS CE (C6,C7) Ar None
W6P E (11Cr1.5Ni0.5Cu) Manual TIG MMA SDSS wire (C5) SDSS CE (C6,C7) Ar 650°C/5min **
W7 C2 (12Cr6Ni2.5MoTi) Manual TIG MMA SDSS wire (C5) SDSS CE (C6,C7) Ar None
W7P C2 (12Cr6Ni2.5MoTi) Manual TIG MMA SDSS wire (C5) SDSS CE (C6,C7) Ar 650°C/5min **
* induction heat treatment of whole pipe girth weld
** furnace heat treatment of piece cut from pipe girth weld

2.3 PWHT

Examples of each weld type were subjected to brief PWHT. For the pulsed MIG welds W1-W5, the PWHT was applied on the whole weld by induction heating, whilst pieces from welds W6 and W7 were heat treated in a furnace. The specified heat treatment cycle was rapid heating to 650°C, followed by holding for 5 minutes and air cooling. A volume of metal around 40mm wide, including the weld metal and approximately 15-20mm of pipe either side of the root and 10-15mm either side of the weld cap was heated by the induction coil. Temperature was controlled via thermocouples on the weld metal cap and measurements were made also on the root. Heating was fairly rapid to 600°C and then temperature rose to 650°C over about two minutes. During the five minute hold period, the cap temperature remained between 640 and 657°C. The root temperature was typically 15-35°C less than the cap temperature, i.e. 620-640°C, depending on the pipe wall thickness. For furnace heat treatment, heating was fairly slow taking about 10 minutes to reach 650°C. Temperature was again monitored by thermocouples, this time on the root weld metal. The welds were given a 'P' designation after PWHT.

2.4 Weld characterisation

Sections were taken through the welds for microstructural examination and Vickers hardness measurement (HV10) in the weld metal and HAZ both before and after PWHT.

2.5 Toughness testing

Charpy V-notch impact tests were performed on through-thickness notched specimens from weld W7 (12Cr6Ni2.5MoTi pipe welded with superduplex consumables) before and after PWHT, with the notch on the weld metal centreline or at the fusionline mid-thickness position. Tests were performed over the temperature range -80 to +40°C.

In addition, fracture toughness tests were performed to BS7448 part 1 at -20°C on Bx2B (11.5x23mm) specimens from weld W7 both before and after PWHT, both given 1% local compression to reduce the effects of residual stress. Specimens were through-thickness notched on the weld metal centreline or on the fusion line mid-thickness position. A loading rate of 0.4mm/min was employed.

2.6 Corrosion testing

Four point bend tests were performed in two environments (i.e. 25%NaCl solutions acidified to calculated pH = 3.3 and 4.5 respectively, Table 4) on 100x15x3mm specimens from each of the girth weld types W1-W6, in both as-welded and PWHT conditions. Two specimen types were examined (i) with the root machined flush and ground to a 600 grit finish and (ii) with the root in the as-welded condition. Both specimen types had strain gauges applied (i) on the test face for flush ground specimens and (ii) on the non-test face for specimens with the profile intact. Specimens were deflected to give a strain equivalent to 100% of the parent material 0.2% proof stress in the HAZ. After deflection the specimens were placed in a nitrogen-blanketed autoclave filled with deoxygenated test solution. The vessel was then heated to test temperature and finally pressurised with the test gas. Test exposure was for 30 days. After test, specimens were examined visually under a binocular microscope, photographed and, if cracking was not observed, they were sectioned transverse to the weld at the mid-width position, to look for small cracks.

Table 4 Corrosion test environments

CodeTotal pressure
(bar)
ppH 2 S
(bar)
ppCO 2
(bar)
NaCl
(%)
NaHCO 3
(ppm)
Temp
(°C)
Calculated pH
Env. A 21.5 0 10 25 0 110 3.3
Env. B 21.5 0 10 25 500 120 4.5

3. Results

3.1 Weld characterisation

The HAZs were generally visibly tempered by PWHT, i.e. showed slightly greater etching response, particularly for the lean grades D and E, but no other microstructural changes were observed optically. In some areas, precipitation on HAZ prior austenite boundaries in steel C1 was visible at high magnification after PWHT. Under a light microscope, this leads to a clear definition of the HAZ prior austenite grain boundaries in the high temperature HAZ within about 150µm of the fusionline, Fig.1. The superduplex weld metals showed evidence of precipitation of very fine secondary austenite after PWHT but no intermetallic phases were observed, Fig.2.

Fig.1. HAZ of weld W3P (pipe C1, 12Cr6Ni2.5MoTi) after PWHT
Fig.1. HAZ of weld W3P (pipe C1, 12Cr6Ni2.5MoTi) after PWHT
Fig.2. Superduplex weld metal of weld W7P after PWHT. The secondary austenite appears as very fine particles between the larger primary austenite units
Fig.2. Superduplex weld metal of weld W7P after PWHT. The secondary austenite appears as very fine particles between the larger primary austenite units

In general, the weld root/mid thickness gave higher maximum HAZ hardness than the weld cap, reflecting effects of reheating/straining, Table 5. Maximum HAZ values, as-welded, were in the range 332-351 HV5, with the highest hardness always being about 2mm from the fusion line. After brief induction PWHT at 620-660°C for 5 minutes, peak HAZ hardness was typically reduced at the root position, by up to 54 HV5 but more typically by 10-15 HV5 for the higher alloy grades. The weld cap HAZs showed a mixed response with hardening observed in some cases (up to +12 HV5) and softening (up to -26 HV5) in others.

Table 5 Maximum HAZ hardness change after induction PWHT at 650°C for five minutes

Weld codePipe codePositionMaximum HAZ hardness
(HV5)
Change in max hardness
(HV5)
As-weldedAfter PWHT
W1/W1P A (12Cr5Ni2Mo) Cap HAZ 330 330 0
Root HAZ 345 345 0
W2/W2P B (12Cr6Ni2Mo) Cap HAZ 303 315 +12
Root HAZ 332 319 -13
W3/W3P C1 (12Cr6Ni2.5MoTi) Cap HAZ 315 304 -11
Root HAZ 327 313 -14
W4/W4P D (13Cr5Ni1Mo) Cap HAZ 345 330 -15
Root HAZ 351 327 -24
W5/W5P B (12Cr6Ni2Mo) Cap HAZ 341 315 -26
Root HAZ 347 312 -35
W6/W6P E (11Cr1.5Ni0.5Cu) Root HAZ 306 254 -54
Root WM 332 319 -13
W7/W7P C2 (12Cr6.5Ni2.5MoTi) Root HAZ 355 315 -40
Root WM 308 301 -7

3.2 Toughness testing

Table 6 and Fig.3 present the results of the fracture toughness and impact tests respectively. Neither showed a substantial reduction of properties after PWHT, although it is noted that the lowest impact values at -50°C for both weld metal centreline and fusionline notch positions were after PWHT. Impact toughness for the fusionline was 65-82J over the range -50 to 0°C, whilst CTOD at maximum load was 0.15-0.29mm at 20°C. The corresponding figures for the weld metal centreline were 25-70J and 0.15-0.23mm.

Table 6 Fracture mechanics test results for superduplex stainless steel weld metal (all from weld W7, tested at -20°C)

SamplesConditionNotch positionMeasured CTOD*
( δm), mm
W7-01 to 03 As-welded WMCL 0.21, 0.15, 0.22,
(0.19)
W7-04 to 06 PWHT WMCL 0.12, 0.20, 0.23
(0.18)
W7-07 to 09 As-welded FLMT 0.15, 0.20, 0.19
(0.18)
W7-10 to 12 PWHT FLMT 0.22, 0.29, 0.21
(0.24)
* Presented as individual values with average in parenthesis
WMCL = weld metal centreline
FLMT = fusionline mid-thickness
PWHT = 650°C for 5 minutes
Fig.3. Effect of PWHT at nominally 650°C for 5 minutes on impact toughness of HAZ and superduplex weld metal in weld W7/W7P (pipe C2, 12Cr6Ni2.5MoTi)
Fig.3. Effect of PWHT at nominally 650°C for 5 minutes on impact toughness of HAZ and superduplex weld metal in weld W7/W7P (pipe C2, 12Cr6Ni2.5MoTi)

3.3 Corrosion testing

Table 7 lists the results of the SCC tests. None of the specimens with the root machined flush showed any evidence of cracking in the environment A (calculated pH=3.3, 110°C) and no tests on such specimens were performed in environment B. When specimens were tested with the root surface intact, most of the specimens showed intergranular cracking in the HAZ, at a variety of locations in the HAZ ranging from immediately adjacent to the fusionline (e.g. W3, steel C1, 12Cr6Ni2.5MoTi, superduplex wire) to about 0.5mm from the fusionline (W1, Steel A, 12Cr5Ni2Mo). No such cracking was found in any weld after PWHT. There was also some variation in crack depth between specimens. In particular, W2 (pipe B, 12Cr6Ni2Mo, welded with superduplex wire) showed very shallow cracking (25-30µm). Weld W5 (also pipe B, welded with SMSS wire) showed no cracking on the section examined. The other specimens cracked through most of the thickness. A second environment was examined (environment B), with pH raised to a calculated value of 4.5 by an addition of 500mg/l NaHCO 3 . Again similar trends were observed, i.e. as-welded specimens tended to crack and PWHT specimens did not. Crack location was similar to that in the environment A. Weld W2 (pipe B, 12Cr6Ni2Mo welded with superduplex wire) and W4 (pipe D, 13Cr5Ni1Mo, welded with matching composition wire) showed no cracks and only shallow cracks were found in W5 (pipe B, 12Cr6Ni2Mo welded with approximately matching composition wire).

Table 7 Results of stress corrosion cracking tests.

Weld codeParent pipeRoot consumableMax root HAZ HV10PWHTMachined surface: Env. A (pH = 3.3, 110°C)Root intact: Env. A (pH = 3.3, 110°C)Root intact: Env. B (pH = 4.5, 120°C)
W1 A (12Cr5Ni2Mo) 25Cr 345 No No cracks Cracks Cracks
W1P A (12Cr5Ni2Mo) 25Cr 345 Yes No cracks No cracks No cracks
W2 B (12Cr6Ni2Mo) 25Cr 332 No No cracks Small crack No cracks
W2P B (12Cr6Ni2Mo) 25Cr 319 Yes No cracks No cracks No cracks
W3 C1 (12Cr6Ni2.5MoTi) 25Cr 327 No No cracks Cracks Cracks
W3P C1 (12Cr6Ni2.5MoTi) 25Cr 313 Yes No cracks No cracks No cracks
W4 D (13Cr5Ni1Mo) 12Cr6.5Ni1.5Mo 351 No No cracks Cracks No cracks
W4P D (13Cr5Ni1Mo) 12Cr6.5Ni1.5Mo 327 Yes No cracks No cracks No cracks
W5 B (12Cr6Ni2Mo) 12Cr6.5Ni2.5Mo 347 No No cracks No cracks Shallow cracks
W5P B (12Cr6Ni2Mo) 12Cr6.5Ni2.5Mo 312 Yes No cracks No cracks No cracks

Extensive, very shallow surface penetrations (about 5-10µm deep), with an intergranular morphology ( Fig.4), were found in the HAZ and parent steel of welds W1 and W1P (pipe A, 12Cr5Ni2Mo, welded with superduplex wire), the first of which was as-welded and the second had been given PWHT. Similar shallow intergranular corrosion was observed in the HAZ and parent steel of W2, W5 and W5P (all in pipe B, 12Cr6Ni2Mo).

Fig.4. Shallow intergranular features on the surface of specimen from weld W1P (pipe A, 12Cr5Ni2Mo) after test in environment B
Fig.4. Shallow intergranular features on the surface of specimen from weld W1P (pipe A, 12Cr5Ni2Mo) after test in environment B

4.0 Discussion

4.1 Current Understanding of the Mechanism of IGSCC of Supermartensitic Stainless Steel

Cracking was at least partly intergranular with respect to prior austenite grain boundaries in most cases, e.g. Fig.5, but some cracks had areas with an apparently transgranular morphology. Figure 6 shows a crack running through an area with retained delta ferrite in the HAZ of weld W3, where the morphology appears to be more transgranular although it is noted that the prior austenite grain structure is not clearly defined. All other authors who have reported this cracking phenomenon have indicated an intergranular morphology. The intergranular crack appearance suggests that the sensitisation mechanism is a consequence of the formation of Cr-carbides and adjacent Cr-depleted zones, as in austenitic and ferritic stainless steels. However, there are a number of differences between the sensitisation of the austenitic stainless steels and the ferritic stainless steels, which are essentially single-phase throughout the welding thermal cycle, and the supermartensitic grades, which undergo several phase changes during welding. For SMSS, IGSCC has been reported to only occur in weld roots, [4] where multiple thermal cycles are experienced. This is not the case for austenitic and ferritic grades, where only one thermal cycle is required. However, this observation does support a Cr-carbide precipitation sensitisation mechanism for SMSS, as little or no carbide formation would be expected to occur during one weld thermal cycle in material that has been transformed to austenite during welding. This is due to the very low M s temperatures (around 200°C for the highest alloy grades [14] ).

Fig.5. Intergranular cracking in the HAZ of weld W1 (pipe A, 12Cr5Ni2Mo) tested in environment A
Fig.5. Intergranular cracking in the HAZ of weld W1 (pipe A, 12Cr5Ni2Mo) tested in environment A

The phenomenon seems to occur at specific HAZ locations, suggesting a critical combination of thermal cycles is required, i.e. to put carbon back into solution and then to form chromium carbides and associated Cr-depleted zones without subsequent 'healing'. Some welds showed cracking at two specific locations in the HAZ, one of which was very close to the fusionline. This suggests that there may be more than one critical location for cracking. This may be rationalised by considering the carbon (and perhaps nitrogen) that may be present in solution at the various locations. In most of the transformed HAZ, carbon and nitrogen in solution after one thermal cycle will depend on the levels present in the steel and the extent of carbide/nitride dissolution during the first thermal cycle. Complete dissolution of Cr and Mo carbides occurs above about 720-800°C. [14] In addition, adjacent to the fusionline for duplex and superduplex weld metals there will be a fairly narrow band where diffusion of carbon and nitrogen into the HAZ may occur from superduplex weld metal, which has higher levels of both elements (especially nitrogen) compared to the parent steel. Experience with duplex stainless steels indicates that this zone may be about 50-100µm wide. [14] This zone may subsequently sensitise at a higher rate than the remainder of the HAZ due to the N and C enrichment.

It was noted that steels C and D, with a Ti addition, cracked particularly close to the fusion boundary. Titanium would be expected to form carbides and nitrides preferentially and tend to lower the C and N content in solution and hence act as a stabilising element as the same way that it does in austenitic and ferritic stainless steels. At very high temperatures, above about 1300°C, stabilised austenitic and ferritic stainless steels show dissolution of the stable Ti-carbides and may subsequently sensitise in such regions if reheated to temperatures around 500-600°C, which promote Cr-carbide formation, leading to so-called 'knifeline' corrosion. [15] Therefore, the location of cracking close to the fusionline in SMSS grades that contain Ti is consistent with the temperature range over which such dissolution of Ti-carbides might be expected. These facts all suggest a sensitisation mechanism that is related to the formation of Cr-depleted zones associated with Cr-carbides. It is noted also that the region close to the fusion boundary also typically contains a small fraction of retained delta ferrite within 100-200µm of the fusion boundary, which could have contributed to cracking susceptibility, perhaps via precipitation of Cr-carbides on the ferrite-martensite boundaries. However, no strong correlation between the location of delta ferrite and the IGSCC crack path was found.

This observation is supported by very fine scale chemical analysis in a transmission electron microscope, which has confirmed the presence of Cr-carbides and Cr-depleted zones in lean grades. [2,3,18] However, no such evidence has been found for the high grades, so it is impossible at present to be conclusive for these grades. Nevertheless, it seems unlikely that IGSCC of these two classes of SMSS would be a result of widely differing mechanisms. In the absence of evidence that is inconsistent with such a mechanism, it is postulated that sensitisation of high alloy SMSS grades is also a consequence of Cr-carbide precipitation, whilst recognising that no positive proof has been obtained to date.

A mechanism of localised near-surface sensitisation has also been observed, associated with the formation of Cr-oxide on the weld surface. [5] The formation of the oxide is associated with prior austenite grain boundary diffusion of chromium, which leads to the development of Cr-depleted regions adjacent to the near-surface prior austenite grain boundaries. Shallow intergranular corrosion associated with this sensitised layer has been observed in high alloy SMSS grades.

Bend specimens tested in hot acidic chloride media only cracked with the as-welded root and, hence, with surface oxide and a stress concentrator. This has been observed by other authors [4] although in highly acidic solutions, smooth specimens have been found to crack. [9] This indicates that the weld surface oxide or stress concentration or both encourage crack initiation but are not essential. The effect of the oxide is presumably related to the Cr-depletion adjacent to grain boundaries immediately beneath the surface. Not all weld specimens cracked in the higher pH environment (B), suggesting a fairly strong effect of pH on IGSCC, similar to the situation for austenitic stainless steels.

4.2 The beneficial effect of PWHT with respect to IGSCC

Postweld heat treatment clearly has a beneficial effect on the resistance of SMSS girth welds to IGSCC. However, testing here was for a fairly short duration and longer term data are required to confirm its applicability to long term service, particularly in the light of reservations expressed by one end user with a 30 day test duration for qualification of SMSS for sweet service. [2] In the present work, one PWHT cycle has been examined on girth welds, namely a nominal 650°C for five minutes although actual temperatures were ~620-660°C, with 620-640°C at the root. Assuming that the Cr-carbide precipitation theory of sensitisation of SMSS to IGSCC is correct, the most likely mechanism by which PWHT is effective in eliminating sensitivity to IGSCC is by allowing chromium back-diffusion into the chromium-depleted zones. The chromium-depleted zone width has been estimated to be up to 20nm in lean grade material but may be <5nm in steel with about 6%Ni and 2%Mo. [2,3,18] Hence, in order for PWHT to be effective, the time and temperature must be sufficient for chromium to diffuse over a distance of this magnitude. Use of a simple x= √t calculation, based on published matrix diffusion coefficients in the range 4.9x10 -14 to 1.5x10 -13 cm 2 s -1 for chromium in iron with 10-20%Cr, extrapolated from higher temperature data, [19] which presumably relates to an austenitic microstructure, indicates a diffusion distance of about 40-70nm for five minutes at 650°C. Higher diffusion rates would be expected in the martensite and ferrite phases. Hence, this very simple calculation supports the proposed Cr-diffusion explanation of the effect of PWHT on eliminating sensitisation to IGSCC.

4.3 Avoiding potential detrimental effects of PWHT

In order for a PWHT cycle to be successfully applied to a SMSS weld, in addition to eliminating sensitivity to IGSCC, it must also be such that it does not have any significantly detrimental effects on other weld properties.

One undesirable effect of PWHT on the HAZ would be associated with heating to a temperature such that an excessive amount of austenite re-forms, leading to formation of un-tempered martensite on subsequent cooling. Un-tempered martensite has high hardness and low toughness in conventional martensitic stainless steels, which have carbon contents in excess of 0.03%, although for the low carbon SMSS grades, these effects are not pronounced and may not be significant. Examination of the effect of PWHT in simulated HAZs showed that 650°C was typically the temperature giving most hardness reduction of the steels studied but also showed substantial variation in response between SMSS grades, with some giving more hardness reduction at 625°C. [14] This indicates the importance of choosing PWHT for the specific steel in question, although broadly similar behaviour is expected for all SMSS grades based on the data obtained here. The reformed, stable austenite content was generally found to increase on tempering at 600-650°C, indicating that Ac 1 was exceeded over this range, hence some virgin martensite formation is possible if the upper temperature during PWHT is above this range. With induction heating, a temperature gradient develops through the pipe wall thickness, with the outside being hotter than the inside. For wall thicknesses of 11-18mm, induction PWHT trials indicated that the root was typically 15-35°C cooler than the cap. The greatest risk of un-tempered martensite formation and associated hardening is therefore in the weld cap, whilst it is essential for eliminating sensitivity to IGSCC in the internal environment that the temperature at the root is controlled. This requires that both root and cap temperatures are held within an acceptable range during PWHT. The limiting upper temperature will vary from grade to grade but based on the current data, which only extends to a cap temperature of up to 660°C, it is recommended that temperatures in excess of 660°C should be avoided. Further work is required to explore the suitability of PWHT temperatures exceeding 660°C.

Another potential detrimental effect of PWHT is that it will tend to increase oxidation of the weld area. Oxidation during welding has been demonstrated to have a detrimental influence on the pitting resistance of SMSS HAZs in mildly sour media [16] and hence any further oxidation from PWHT might also be detrimental. However, published work has indicated that PWHT at 650°C may be beneficial for service under mildly sour conditions, [17] presumably by lowering hardness, but it does not give immunity to cracking in sour media. Further work is required to explore this issue, although use of an inert gas shield during PWHT would eliminate the concern.

Detrimental microstructural effects at the edge of the PWHT zone, where intermediate temperatures will be experienced, are not anticipated, provided that the whole of the weld HAZ is heat treated, i.e. that the intermediate temperatures are experienced by parent steel. This assumes that the parent steel will have been tempered such that the carbon content in solution is very low. Detrimental microstructural effects in the HAZ and weld metal are of greater concern. These may include precipitation of (i) further carbides, e.g. on prior austenite boundaries or within or on the interface of any delta ferrite retained in the HAZ and (ii) intermetallic phase, secondary austenite or alpha prime phase in the delta ferrite in weld metal deposited with a duplex or superduplex consumable. These precipitation reactions may act to lower corrosion resistance and toughness in the weld metal or HAZ very close to the fusion line, although the present study showed that the toughness effects are not significant for a high grade SMSS HAZ and superduplex weld metal subject to PWHT at 650°C for 5 minutes. To avoid loss of toughness, it is recommended that the PWHT duration should not be substantially longer than 5 minutes whilst recognising that longer PWHT may still give acceptable results for many applications. Substantially shorter PWHT periods are not recommended due to the absence of data. Sensitisation is not expected provided that the whole of the HAZ sees the intended PWHT temperature. No loss of corrosion resistance associated with precipitation on delta ferrite in SMSS HAZs has been noted to date, although one reference cites it as an issue for conventional 13%Cr 4%Ni steels, but does not indicate the precise temperature range of concern, although it does state that tempering at around 600°C gives good corrosion resistance, [15] and hence problems are only likely to occur below this. Based on the results of the present work, a suitable lower temperature limit of 620°C is suggested for the HAZ.

Although some precipitation occurred in superduplex weld metal during PWHT, this was apparently restricted to the formation of secondary austenite. Secondary austenite tends to reduce corrosion resistance but this should not be a problem when welding SMSS. This implies that, although PWHT of superduplex weld metal is not normally considered advisable, in this case 5 minutes PWHT at 650°C does not seem to be detrimental. If longer PWHT times or higher temperatures were used, some loss of toughness in superduplex weld metal might occur, although this was not studied here.

5. Conclusions

  1. The sensitisation of lean grade SMSS HAZs has been linked to the formation of Cr-carbides on prior-austenite grain boundaries and adjacent Cr-depleted zones but this link has not been established for the high alloy grades. Formation of Cr-depleted zones on prior-austenite boundaries immediately underneath the welding oxide has been observed in high alloy grades. Hence some uncertainty remains over the mechanism of IGSCC of high grade supermartensitic stainless steel and the effect of PWHT. Nevertheless, there is a substantial body of information supporting a consistent beneficial effect of brief PWHT for a broad range of supermartensitic grades.
  2. It is recommended that PWHT should be applied to welds in supermartensitic stainless steel where there is a risk of intergranular SCC in service, i.e. in hot acidic environments. A PWHT temperature of 620-650°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 the current work extended up to 660°C. Heating and cooling should be fairly rapid. The most appropriate PWHT duration has not been established but there is fairly common agreement that 5 minutes is an appropriate duration.
  3. Whilst the beneficial effect of PWHT with respect to IGSCC has been demonstrated for 30 day exposure tests, longer term data are required to confirm the applicability of the effect to long term service.
  4. Due to the limited information available, the use of welded supermartensitic stainless steel in the PWHT condition will require qualification on a case by case basis. The qualification programme should consider the effects of PWHT on toughness and sour service performance, in addition to IGSCC. 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.
  5. No substantial change in toughness of superduplex weld metal was observed for PWHT at 650°C for 5 minutes, although secondary austenite was formed. A small reduction of root HAZ hardness was generally associated with PWHT.
  6. Postweld heat treatment may also have detrimental effects if not adequately controlled, e.g. (i) thickening of weld area oxides and associated loss of general/pitting corrosion resistance, (ii) formation of virgin martensite in the HAZ and increased hardness leading to reduced toughness and resistance to sour environments, (iii) loss of toughness in superduplex stainless steel weld metal, (iv) tempering of HAZ at temperatures that could induce sensitisation to intergranular SCC if the heat treated area is not wide enough.
  7. The precise response to PWHT is specific to each individual grade of supermartensitic steel, although the data indicate that all steels examined here were fairly similar and the beneficial effect of 5 minutes at 620-650°C, with respect to IGSCC, is applicable to 'lean' grades, with <1%Mo and 'high' grades with >2%Mo both with and without Ti addition.

Acknowledgements

The following companies, who sponsored the work, are thanked for their permission to publish the results: Advantica, BP, Chevron, ENI, ESAB, Industeel, JFE Steel, J Ray McDermott, Lincoln Electric, Nickel Institute, Serimer Dasa, Shell, Sumitomo Metal Industries, Tenaris and Total.

7. References

  1. Rogne T, Lange H I, Svenning M, Aldstedt S, Solberg J K, Ladanova E, Olsen S, Howard R and Leturno R E: 'Intergranular corrosion/cracking of weldable 13%Cr steel at elevated temperature', Proc conf Corrosion 2002, USA, NACE International, paper 02428.
  2. van Gestel W: 'Girth weld failures in 13Cr sweet wet gas flow lines', Proc conf Corrosion 2004, USA, NACE International, paper 04141.
  3. Ladanova E and Solberg J K: 'Transmission electron microscopy investigation of precipitation reactions in coarse-grained heat affected zone in two 13%Cr supermartensitic stainless steels', Proc Conf, Supermartensitic Stainless Steels 2002, Belgium, KCI publishing, pp205-214.
  4. Coudreuse L, Ligier V, Lojewski Ch, Toussaint P, 'Environmental induced cracking (SSC and SCC) in supermartensitic steels (SMSS), Proc Conf, Supermartensitic Stainless Steels 2002, Belgium, KCI Publishing, pp163-172.
  5. 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.
  6. Ladanova E, Solberg J K and Rogne T: 'Investigation of the precipitation mechanism in HAZ of supermartensitic stainless steels. Two-pass weld simulation'. Proc conf Eurocorr '04, Nice, France.
  7. Hashizume S and Ono T: 'Laboratory results of cracking behaviour of weldable 13%Cr steels girth weld joints in hot acid environments', Proc conf Eurocorr '04 , Nice, France.
  8. Hoerner B, Bayle B, Ligier V and Delafosse D: 'Stress corrosion cracking of welded joints of supermartensitic stainless steel in H2S free environment', Proc conf Eurocorr '04, Nice, France.
  9. Miyata Y, Kimura M, Nakamichi H, Sato K, Itakura N and Masamura K: 'Mechanism of intergranular stress corrosion cracking in HAZ for supermartensitic stainless steel', Proc conf Eurocorr '04, Nice, France.
  10. Enerhaug J, Olsen S, Rorvik G, Mikkelsen A, Rogne T Rostadsand P A, Haabrekke T: 'Robustness of supermartensitic stainless steel girth welds - experiences from the Kristin Field Development Project', Proc conf Corrosion 2005, USA, NACE International, paper 05097.
  11. Hashizume S and Ono T: 'Effect of PWHT temperature on SCC behaviour of low C-13%Cr steels weld joints in hot acidic production environments', Proc conf Corrosion 2005, USA, NACE International, paper 05094.
  12. Karlsson L, Rigdal S, Dhooge A, Deleu E, Goldschmitz M and Broek K van den: 'Mechanical properties and ageing response of supermartensitic (stainless steel) weld metals', Proc conf Stainless Steel World 2001, The Netherlands, KCI publishing, pp47-59.
  13. Dhooge A, Deleu E and Holy A: 'Weld simulation testing of different supermartensitic stainless steel grades', Proc Conf Supermartensitic Stainless Steels 2002, Belgium, KCI Publishing, pp225-232.
  14. Woollin P and Carrouge D: 'Heat affected zone microstructure in supermartensitic stainless steels', Proc Conf Supermartensitic Stainless Steels 2002, Belgium, KCI Publishing, pp199-204.
  15. Folkhard E, 'Welding metallurgy of stainless steels', Springer Verlag, 1988.
  16. Olsen S and Enerhaug J: 'Common pitfalls during SSC (sulphide stress cracking) and pitting testing of supermartensitic stainless steels for use in pipelines', Proc conf Corrosion 2002, USA, NACE International, paper 02038.
  17. Enerhaug J, Kvaale P E, Bjordal M, Drugli J M, Rogne T: 'Qualification of welded super 13%Cr martensitic stainless steels for the Asgaard Field, Proc conf Corrosion 99, USA, NACE International, Paper 587.
  18. Gregori A, Woollin P, van Gestel W: 'Analysis of in-service failures in girth-welded supermartensitic pipes', Proc conf Stainless Steel World 2003, Netherlands, KCI Publishing, p195.
  19. Smithells Metals Reference Book, 6th Edition, Brandes, 1983.

For more information please email:


contactus@twi.co.uk