228606791 Offshore Pipeline Engineering Materials Welding Module Construction Practice

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE

CONSTRUCTION PRACTICE 5.1 CONVENTIONAL CONVENTIONAL LAY .......................................... ................................................................. ............................................. ............................... ......... 300 5.1.1 Welding Techniques – Carbon Steel Pipelines ............................................. ....................................................... .......... 305 5.1.1.1 Shielded Metal Arc Welding (SMAW) ............................................. .............................................................. ................. 305 5.1.1.2 Mechanised Gas Metal Arc Welding ......................................... .............................................................. ......................... .... 312 5.1.1.3 Semi-automatic Flux Cored Arc Welding .......................................... .......................................................... ................ 318 5.1.1.4 One Shot Welding Processes ............................................ .................................................................. .................................. ............ 319 5.1.2 Welding Techniques – Corrosion Resistant Alloy Pipelines .................................. .................................. 327 5.1.2.1 Gas Tungsten Arc Welding. ............................................. ................................................................... ................................... ............. 327 5.1.2.2 Mechanised Gas Metal Arc Welding. ........................................... ................................................................. ...................... 330 5.1.2.3 Semi-automatic Flux Cored Arc Welding. ............................................. ......................................................... ............ 330 5.1.2.4 One Shot Welding Processes. ........................................... ................................................................. .................................. ............ 330 5.1.2.5 Precautions When Welding CRA Materials. ............................................. ...................................................... ......... 331 5.1.3 Inspection Techniques............................ Techniques................................................... ............................................. .......................................... .................... 334 5.1.3.1 Defect types............................................. types................................................................... ............................................ ...................................... ................ 335 5.1.3.2 Visual Inspection.......................................... Inspection................................................................. .............................................. ................................ ......... 341 5.1.9.4 Radiography ............................................. .................................................................... .............................................. ................................... ............ 342 5.1.9.5 Ultrasonic Examination ............................................ .................................................................. ......................................... ................... 346 5.1.9.6 Magnetic Particle Inspection............................................... Inspection...................................................................... ............................... ........ 348 5.2 J LAY....................................... LAY.............................................................. .............................................. .............................................. ......................................... .................. 348 5.2.1 Welding Processes ............................................ ................................................................... .............................................. ............................... ........ 349 5.3 COILED COILED LINE PIPE ................................................ ........................................................................ ................................................ ............................... ....... 350 5.4 LANDFALLS.............. LANDFALLS.................................... ............................................ ............................................. ............................................. ................................. ........... 351 5.5 TIE-IN OPERATIONS ........................................... .................................................................. ............................................. ................................. ........... 352 5.6 REFERENCES REFERENCES .............................................. ..................................................................... .............................................. .......................................... ................... 353

299

Main Menu


5.1 CONVENTIONAL LAY

A conventional laybarge for large diameter offshore pipe is designed to lay pipe using the S-lay technique, Figure 1. This technique, which is currently applicable to water depths up to about 600 metres, is so called because the pipe follows an ‘S’ shaped curve as it moves from the laybarge to the sea-bed. The pipes are welded to each other in the horizontal position on the barge and the pipeline then passes over an inclined ramp or ‘stinger’ which gradually lowers the pipe into the water. This region of the S curve is known as the ‘overbend’ and as the pipe leaves the overbend region it is inclined almost vertically as it descends to the sea bed. Close to the sea-bed it once again returns to the horizontal position so that it eventually rests on the sea-bed. This region is known as the ‘sag bend’ region.

Figure 1. Laybarge Laybarge methods : a) J lay, lay, b) S lay, lay, c) Reel lay (1). In order to prevent the t he pipe from buckling in the regions r egions of maximum bending, the bend radius is controlled by keeping the pipe under te nsion, so that the pipe actually follows a ‘lazy S’ shape. The tension is applied to the pipe by tensioners on the barge which are usually arrays of rubber wheels or belts which surround the pipe and apply an axial force to the pipe through the friction generated between the tensioner tensioner and the pipe external coating. The force is maintained at the barge end of the pipeline either by an array of

300

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE winches on the barge which pull on anchors which are placed on the sea bed ahead of the barge by tug boats, or by thrusters on the barge which dynamically position it with respect to the pipeline, or by a combination of the two. The force on the pipeline is i s reacted at the sea-bed end of the pipeline by the dead weight of the pipeline and friction between it and the sea-bed. Obviously the larger the force applied by the tensioners to the pipeline, the more gradual will be the bending radii in the S portion of the laying curve. Also as the pipe weight increases it is necessary to apply a greater force to the pipe to maintain the desired bend radii and so prevent buckling, particularly in the sag bend portion of the curve. This may entail adjusting the capacity of the tensioning stations on the laybarge. For example the tensioners on the EMC ‘Castoro Sei’ laybarge have been increased from their original 180tons capacity to 330tons, to allow the barge to lay large diameter pipes in deeper water. (2) The cradle rollers which carry the pipe on the laybarge stinger may be instrumented so that a continuous check of the stresses during pipelaying may be kept. If any anomalies occur this could indicate a problem with the pipeline which requires investigation. Also a buckle detector may be drawn through the pipeline as pipelaying proceeds to ensure that a buckle has not occurred in the sag bend region of the pipeline. This buckle detector, which is similar to a lightweight gauging gauging pig, is positioned beyond beyond the lower part of the S curve away from f rom the laybarge, so that it is always beyond the sag bend. Should a permanent deformation greater than 95% of nominal pipe diameter (i.e the gauge plate diameter) occur in the pipeline the gauge plate will stick and the laying contractor is alerted. As individual pipe lengths are welded onto the growing pipeline, the barge is winched forward and the new section of pipeline passes over the stinger towards the sea-bed. In the case of anchor positioned barges tugs are used to continuously reposition the anchors ahead of the barge so tthat hat the barge can keep moving forward. It may sometimes be necessary to lay two pipelines at once, one being a smaller diameter line to carry car ry condensate or glycol, glycol, for example. In this case the smaller pipeline can be ‘piggy backed’ onto the main line, or laid separately using a mini firing line and stinger alongside the main firing line. In the early days of the use of fusion bonded epoxy (FBE) coated pipe offshore, problems were experience with movement of the concrete weight coating on the pipe in the tensioners. The concrete weight coating was applied over the smooth FBE coating and, since the pipe ends were not weight coated (in order to allow access for welding), it was possible for the weight coating to slide along the pipe, thus closing up the gap at the pipe ends. The problem was overcome by applying applying a sand treatment to the FBE before it hardened in order to increase its surface roughness. The concrete weight coating would then ‘key’ into the roughened surface and there was sufficient friction between the two coatings to prevent slippage in the tensioners. The tension at which the pipe is laid may have to be adjusted to suit the laying corridor and sea-bed profile. For example, for the Troll oil pipeline the section of pipeline in the fjord approach had to be laid very accurately in up to 540m of water. The accuracy was needed because of the undulating sea bed profile, tight bend radii, numerous changes of

301

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE direction and the need to position the pipe on pre-installed gravel berms. This resulted in the specification of a +/-3m laying corridor, which was achieved using a low tension set up (35-95 tonnes). This requirement conflicted with the normal high tension approach to deepwater pipelay and was accommodated by means of a specially extended stinger. Outside the fjord, the pipeline had to cross the Norwegian Trench and a conventional high tension set up (183 tonnes) was used to ensure satisfactory pipeline conditions during operation (3). A pipeline submerged weight/water depth layability l ayability curve can be formulated to to define the laying limits of a laybarge laybarge with a specific configuration, Figure 2. The ability of S lay barges to lay pipe in deep waters can be improved by increasing the power of the tensioners, by altering the angle of the stinger (to a steeper angle), by using higher yield pipe steels, and by using new design methods such as limit state design. desi gn. The maximum depth achievable varies with pipe diameter as shown in Figure 3. Even for small diameter pipe the maximum depth limit for S laying is unlikely to exceed 1000 metres, and beyond this there is a need to go to J lay (see section 5.2).

Figure 2. Static layability curve for S lay barge

(2)

.

Current so-called 3rd generation laybarges have been in operation since the 1970’s, although refinements in their layout and operation have taken place during this time. In order to increase their laying capability for large diameter pipes in deep water

302

Main Menu


Figure 3. Water depth versus pipeline diameter for various projects (2). The deck of a conventional laybarge consists of a number of covered, fixed workstations laid out along the barge. As each operation is completed the pipeline moves through the workstations a pipe length at a time. A typical barge layout might be as in Figure 4, and the number of workstations might be 3-5 for welding, 1 for inspection and repair, and 23 for coating.

Figure 4. Typical barge layout for S Lay (1).

303

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE In addition to the main firing line welding stations there may be additional double joint stations separate from the main firing line. This allows two standard length pipe sections (normally around 12m length) to be rotated under a submerged arc welding head. In this way the pipe is double jointed offline with a high productivity weld process, and the pipeline production rate when welding these sections in the main firing line is effectively doubled compared to the use of single pipe lengths. The pipe material used for carbon steel offshore pipelines is similar to that used for onshore pipelines. However, company requirements for offshore pipes are usually more stringent than for onshore pipes, because of the more severe laying and operating conditions, and the increased cost of failure of the pipeline. Tolerances on wall thickness, diameter, and ovality in particular are controlled to give good fit-up for welding and to prevent buckling during laying. Early offshore pipelines in the North Sea were welded using the same welding processes as had been used traditionally for onshore pipelines. Thus, manual ‘stovepipe’ welding using cellulose coated electrodes was common. In addition because of the high cost of laybarge hire there was pressure to limit the time spent at each workstation on the barge. Although the multipass girth welds could be completed at a number of workstations, the inspection of the completed welds was carried out at one station and this operation was often the rate controlling process. Radiography was the accepted girth weld inspection technique but it was common to view the resulting radiographs with the films still wet after developing, in order to save time. This viewing method meant that it was possible to miss crack-like defects in the weld. The combination of cellulosic electrodes, the pressure to maximise the welding rate, the stress seen by the newly completed weld as it passed over the stinger, and the less than ideal inspection methods led to a number of problems. Firstly, cellulosic electrodes produce a weld which is high in hydrogen and this reduces the toughness of the weld. For onshore pipelines this dissolved hydrogen normally diffuses out of the completed weld before any significant stressing takes place and the toughness properties recover. However, a laybarge weld sees a high stress as it passes over the stinger within minutes of it being completed. Secondly, fast weld completion rates led to large weld bead sizes with consequent coarse grain size and lack of inter-run grain refinement, meaning that weld toughness was once again poor. Thirdly the viewing of wet radiographs meant that crack like defects could be missed. The combination of crack like defects and low weld toughness meant that it was not uncommon for welds to break as the pipe passed over the stinger resulting in the need to recover the pipe from the sea-bed. Also, even if the pipe was successfully laid, later radiographic interpretation quality audits, carried out on dry radiographs, could reveal weld defects which were outside the fabrication code defect acceptance levels. It was then necessary to carry out expensive fitness-for purpose analyses to justify continued operation of the pipelines. As a result of these problems, there was a move towards the use of low hydrogen and mechanised gas metal arc welding systems for the welding of pipe on laybarges. The welding techniques adopted for carbon steel pipelines on conventional laybarges are described in the following sections.

304

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE 5.1.1 Welding Techniques – Carbon Steel Pipelines

5.1.1.1 Shielded Metal Arc Welding (SMAW) For girth welding of fixed pipe to pipe joints on a laybarge the direction of welding can be:(1) vertical down from 12 o’clock through to 6 o’clock (the fastest welding method, commonly known as ‘stovepipe’ welding). (2) vertical up from 6 o’clock through to 12 o’clock (the slowest welding method, commonly known as ‘conventional’ welding). (3) vertical up for the root pass, vertical down for the fill and cap passes (a compromise for overall welding speed, known as ‘composite’ or in welder’s slang ‘dolly mix’ welding). These techniques are shown in Figure 5.

12

12

12

6

6

6

VERTICAL DOWN (STOVEPIPE)

VERTICAL UP (CONVENTIONAL)

COMBINATION (DOLLY MIX)

Figure 5. Pipe welding modes.

The shielded metal arc welding process is a fusion welding process and, therefore, comprises: (a) a heat source (in this case the electric arc). (b) protection for the arc and the molten weld pool (provided by the gas shield around the arc and the slag layer on the weld pool, both of which come from decomposition of the electrode flux covering). (c) filler metal to fill the weld preparation (this comes from the electrode core wire) (Figure 6).

305

Main Menu


Figure 6. Schematic of shielded metal arc welding (SMAW) process. The importance of these various components is described below : The electric arc This is a low voltage, high current discharge (15-30 volts, 50-300 amps in SMAW) which occurs between the end of the electrode and the work piece, and generates temperatures in the region of 6,000 oC. The heat from the arc melts the tip of the electrode and also melts the abutting edges of the weld preparation and forms a molten pool on the work piece. Molten droplets from the electrode tip are propelled across the arc into the weld pool by gravity, surface tension and complicated electromagnetic forces. Most SMAW welding of pipelines is carried out using direct current power supplies, with the electrode connected to the positive side of the power supply and the welding return lead (incorrectly referred to as ‘the earth return’) connected to the negative side of the power supply. Occasionally, electrode negative or alternating current power supplies may also be used for special applications. Protection for the arc and weld pool The flux coating on the electrode decomposes during welding and fulfils three main functions: (1) It provides a gas shield around the arc and weld pool - this helps to maintain a stable arc, and also prevents contamination of the molten weld pool by oxygen and nitrogen in the surrounding air. (2) It melts to form a slag covering over the weld pool which 'fluxes out' impurities from the weld pool and also helps to contain the molten weld pool by surface tension effects, so facilitating control by the welder, particularly when welding vertically or overhead.

306

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE (3) It can be used to add alloying additions to the weld pool, since the electrode core wire itself is normally a simple carbon manganese steel. Electrode flux formulations vary according to the type and application of the welding electrode, and the welding characteristics of individual makes of electrode also vary according to the manufacturer’s own preferred formulation. Electrodes therefore have to be carefully evaluated to ensure that they are easy to use and appeal to the welder, as well as produce welds with the correct mechanical properties. Electrode types Offshore SMAW pipeline welding makes use of electrodes with two main types of coating: cellulosic, and basic. Cellulosic Electrodes These electrodes are coated with cellulose in the form of wood pulp which decomposes in the heat of the arc to give a gas shield of hydrogen, carbon monoxide and water vapour. The arc is harsh and deeply penetrating and the slag is light and fast freezing, allowing welding to be carried out in the rapid vertical down direction. This makes these electrodes suitable for stovepipe welding of pipeline girth welds. A disadvantage with these electrodes is the large amount of hydrogen present in the gas shield which is generated around the arc, since this can be dissolved in the weld pool and then cause cracking in the weld metal or HAZ (see Section 3.1.13) . For this reason careful control of the welding procedure, especially preheat is required. These electrodes were used for the early offshore pipelines but have largely been superseded by alternative techniques for the reasons previously described. Basic (Low Hydrogen) Electrodes These electrodes are designed for use when hydrogen cracking is a problem, the coating being based on calcium carbonate which produces a gas shield of carbon monoxide and dioxide and very little hydrogen. They are also used where good weld metal toughness is a pre-requisite. Liberation of the gas shield around the arc is slow, which means porosity can occur at stop/starts in the welding process and the welder must maintain a short arc while welding. Traditionally low hydrogen electrodes have been used mainly in the slow vertical up mode and have not been economic for offshore laybarge pipeline construction. Recently low hydrogen vertical down (LHVD) electrode have been used for making pipeline girth welds on offshore laybarges. These electrodes find application particularly for small diameter flowlines, where the mechanised welding systems are not suitable. The LHVD electrodes have been available for some years, but have been slow to be accepted, since they require a slightly different welding technique compared to the traditional cellulosic electrode. They can be used for root bead welding of multipass pipe girth welds, but they are not as quick as cellulosic electrodes for this part of the joint and require special welder training. As a result, mixed cellulosic/LHVD procedures are sometimes used. These combine the speed and relative ease of use of the

307

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE cellulosic electrode for the first two weld passes (the root and hot pass) with the good toughness and freedom from weld metal hydrogen cracking of the low hydrogen electrode when used for the fill and cap passes. 5.1.1.1.1 Welding techniques

The traditional method of joining steel pipe using SMAW welding is, to make a full penetration weld between pipe ends which have a 30o pipe end bevel, and which have been set up so there is a small gap between them. The pipe ends also have a root face (‘nose’, or ‘landing edge’) of about 1.5mm to prevent the root bead weld pool from melting away the adjoining pipe material too quickly. Several weld runs may be required in order to completely fill the weld preparation (Fig 7). Considerable skill is required during deposition of the first or root weld run in order to bridge the gap between the pipe ends with molten metal, at the same time ensuring consistent penetration of the pipe bore. Less skill is required to fill the remainder of the weld preparation since the problem of over-penetration or 'burn-through' is much reduced once the root run has been completed.

1. Root run, 2. Hot pass, 3. Hot fill, 4-5 Fillers, 6. Cap Figure 7. Weld deposition sequence for SMAW girth weld.

As mentioned previously, the chosen welding direction may change depending on the application. For full penetration pipe to pipe girth welds, the technique normally used is vertical down welding using cellulosic electrodes - 'stovepipe' welding. The weld preparation is as shown in Figure 8 (a) . The vertical down technique is considerably quicker than vertical up welding, for example a weld pass in a 950 mm pipe will take two welders (one welding on each side of the pipe) about 13 minutes when welding vertically down, whereas 32 minutes should be allowed when welding vertically up.

308

Main Menu


60+- 10 0

!

60 +- 10 0

+ 1.5 - 10 mm

!

+ 1.5 - 10 mm

+1 1.5 - 0 mm

+

0.8

2.4 - 0

a) Weld preparation for vertical down weld

mm

b) Weld preparation for vertical up weld

Figure 8. Typical weld preparations.

However, despite its slowness, the vertical up technique with cellulosic electrodes has to be used in some circumstances, for example for the welding of small diameter pipe, say less than 200mm diameter, when its superior control is necessary. An alternative procedure, which is used for small diameter pipe to pipe joints, is to use conventional welding for the root run only, (see Figure 8b for weld preparation) and then to complete the weld using vertical down welding - this is known as composite or 'dolly mix' welding. This procedure combines the ease of use of the vert ical up root bead with the speed of the vertical down fill and cap passes.

5.1.1.1.2 Welding Procedure

Laybarge welding for offshore pipeline construction is normally carried out to international or national codes such as, API1104, BS4515, or DnV96, although many companies also have their own additional requirements. Before any welding is carried out, a welding procedure is submitted by the contractor for approval. After the initial procedure proposal has been approved, with any revisions thought necessary, the contractor is normally required to carry out a procedure qualification test weld which is examined non-destructively and destructively to the requirements of the relevant specification. This is carried out to demonstrate that the proposed procedure can be applied without difficulty and is capable of meeting the specification requirements. For onshore pipeline projects there has been a trend in recent years to accept pre-qualified welding procedures, provided these have been witnessed by an independent third party. However, because the welding procedure is critical to the success of the whole laybarge operation it is normal practice to qualify the procedures for each offshore contract using laybarge personnel and equipment and the pipe which is to be used for the project. In this way there is the greatest chance that there will be no unexpected problems when laybarge operations start. The items specified on a welding procedure sheet are covered below:

309

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE Welding electrodes The principal requirement for welding electrodes are that they shall deposit weld metal which has adequate strength and toughness, and is compatible with the properties of the pipe material. Secondly, but by no means of secondary importance, is the ability of the electrodes to be used without difficulty by the average welder, ie. they must have good welding characteristics. For full penetration welding the types of electrode used are designed to give a nominally 'overmatching' weld metal. This means that the weld metal should have higher tensile strength than the minimum specified tensile strength of the pipe material. This is required by both American (API 1104) and British Standards (BS4515) pipeline welding standards, and is demonstrated by a transweld tensile test. For pipe grades up to X65 this also means that the yield strength of the weld metal usually overmatches that of the pipe, although only the tensile strength is measured in the transweld tensile test. This overmatching of tensile properties means that the potential effect of defects in the weld region is minimised, since any plastic strain will be seen by the pipe rather than the weld metal. For higher yield pipes it is more difficult to guarantee yield strength overmatching and there is currently much debate about the need to specify a weld metal yield strength overmatching requirement as well as a tensile strength overmatching requirement in pipeline specifications. An exception to the general rule of overmatching is the use of a lower yield strength, nominally undermatching, electrode when depositing the root run in multi-pass girth welds. This serves the purpose of allowing some of the stresses generated during welding of the root bead (for example due to weld metal thermal contraction and pipe handling operations) to be accommodated by the weld metal. The reason for this is to reduce the stress build up in the root bead heat affected zone (HAZ), in order to avoid hydrogen induced cracking. The lower yield electrodes also normally produce a smoother root bead profile than higher yield electrodes. Once the root bead is in place, subsequent weld runs (with the higher yield strength electrode) help to temper the critical heat affected zone region and produce a finished weld which overmatches the pipe in tensile properties. Welding parameters The welding procedure must include the weld preparation dimensions, the direction of welding (vertical down, vertical up, or composite), and also the electrode sizes, currents and voltages and welding speeds which are to be used. These points ensure that the weld deposit is being made in a reasonable manner, both to ensure that the strength will be correct, and also to ensure that the welder is able to control the weld pool adequately, and so keep the incidence of defects to a low level. Preheat temperatures and, weld interpass times To avoid excessive HAZ hardening and to minimise the amount of hydrogen which is likely to be trapped in the HAZ after welding, a number of requirements are specified in the welding procedure. The first of these requirements is that the weld shall be preheated before any weld is deposited. Preheating serves the dual purpose of drying

310

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE the components by driving off moisture and slowing the cooling rate after welding by reducing the thermal sink effect of the cold pipe material either side of the joint. The preheat level is normally determined from the pipe chemistry, or by carrying out weldability trials. These trials consist of welding together, under controlled conditions, pipes with chemistry towards the maximum from any particular order. This test weld is then sectioned and the pipe HAZ is tested for the hardness and the sections examined for cracking. Based on these results, a recommended preheat is decided. For current steels, the vast majority of pipes in temperate climatic conditions can be welded with a preheat of 50o C. However there may be situations which demand higher preheats than this, for example X80 grade pipe, or heavy wall (e.g. greater than 25mm) pipe, or welding in adverse weather conditions. Fittings, such as tees and flanges etc, always need a higher preheat than is normal for pipe to pipe joints. This is for two reasons, one of which is that the steel used in these fittings is frequently of a more hardenable type than the pipe. The other, that the thickness of the fittings is somewhat greater than the pipe so it acts as an additional heat sink, which means that the heat will be dissipated much more rapidly. The preheat normally used for welding fittings is of the order of 150 oC. The other major part of a welding procedure which controls cracking tendency is the time limitation between the start of one run and the start of the next run. Stipulation of a maximum interpass time ensures that when the first weld run is deposited a second weld run (the ‘hot pass’) is placed on top of it as quickly as is reasonably possible so that the heat may be maintained in the joint. This second run also gives the additional benefit of reinforcing the first run or root run so that it can withstand any movement better. There is also a designated maximum time lapse between the second and third runs for the same reason. On a laybarge, these interpass times are usually adhered to without difficulty, since the pipe is moved to the next station on the barge as soon as possible. Proving the procedure Once the submitted welding procedure has been examined, the next stage is to test that procedure to see if it meets the specification. To this end, when a contract has been awarded, the contractor is required to carry out at least one simulated laybarge weld, preferably on full lengths of pipe, using the designated procedure. The weld is visually inspected throughout its execution, and it is inspected normally using radiography and the magnetic particle technique on completion, to check that the weld meets defect acceptance limits in the specification. Increasingly automatic ultrasonic inspection systems may be used for girth weld examination, especially in combination with mechanised welding processes (see below). After non-destructive examination the joint is then cut up into mechanical test specimens and mechanical testing is carried out according to the relevant specification. Mechanical testing usually involves transweld tensile tests to prove that the weld metal tensile strength overmatches the minimum specified tensile strength of the pipe, Charpy V notch toughness tests to prove adequate notch toughness, and macro and hardness examination to prove that the weld and HAZ are defect free and not susceptible to

311

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE cracking. In the case of mechanised welding processes, which can be prone to lack of fusion defects (see later), there may also be a requirement to carry out crack tip opening displacement (CTOD) tests on the qualification weld, particularly if the procedure is to be used for an offshore pipeline. The value of the CTOD test is that it gives information which can be used to carry out fitness-for-purpose assessments of defective welds which may avoid the need for costly and difficult repairs. Once the welding procedure has been qualified the competence of the welders to use the qualified procedure must be tested. The welder's test consists of making a weld to the qualified procedure. Some specifications then allow the welder to be assessed using only non-destructive methods, such as radiography and magnetic particle inspection. In this way the welder can be qualified on his first production weld. Others call for mechanical tests, similar to those used to qualify the original procedure. The weld is acceptable if it meets the mechanical property levels and/or contains defects which are within the acceptable defect levels given in the specification. These acceptable defect levels are workmanship levels, in other words they represent a quality level which should be achieved by a reasonably competent welder. It is well known that greater amounts of defect are acceptable on a fitness-for-purpose basis, but these limits are only used where it would be difficult to carry out a repair (eg tie-in welds, defects discovered in service, or welds in subsea pipelines). 5.1.1.2 Mechanised Gas Metal Arc Welding Gas metal arc welding (GMAW) is available in two forms: semi-automatic and mechanised. In the GMAW process the filler wire is in the form of a continuous reel of bare wire so that welding can proceed virtually uninterrupted. The weld pool is shielded by a stream of gas (usually argon or carbon dioxide, or a mixture of the two) which is supplied through a nozzle in the welding gun (Figure 9). In semi-automatic form the filler wire is fed through a flexible tube to a hand held gun, but the welding arc length is controlled automatically.

Figure 9. Schematic of gas metal arc welding (GMAW) process.

312

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE In the 1960's the introduction of large diameter X60 pipes made from normalised (high carbon equivalent) steels for onshore pipelines led to widespread HAZ hydrogen cracking problems when using cellulosic SMAW electrodes. Semi automatic GMAW welding, which is a low hydrogen welding process, was used in an effort to overcome this, Figure 10. However, the process was relatively new and problems were experienced with the equipment and with lack of fusion defects. With the introduction of controlled rolled steels with better weldability, cellulosic electrodes once again became popular.

Figure 10. Semi-automatic GMAW in use on a land pipeline in the 1960’s As previously mentioned the use of cellulosic electrodes for offshore pipelines is now uncommon, especially in the North Sea. Mechanised welding systems, Figure 11, are more attractive for laybarge use because : • • •

• • •

they are capable of high production rates. they produce a low hydrogen weld deposit. the workstation is fixed (the pipe moves) so that equipment damage is less of a problem than onshore. the workstation is usually enclosed so that gas shielded processes can be used. less welder skill (hand to eye co-ordination)is required. welds are more reproducible and lend themselves to mechanised ultrasonic inspection.

In manual welding, the welder can accommodate variations in joint fit-up by altering his current, travel speed, and electrode angle. This is particularly important for the root bead, since the joint is set up with a root gap which can vary slightly around the joint circumference. When using mechanised welding this root bead control is more difficult and two main solutions have been used.

313

Main Menu


Figure 11. Mechanised pipeline welding (courtesy CRC Evans). These solutions have been adopted in various proprietary welding systems which fall into two groups, those in which all welding is carried out from outside the pipe and those in which the root bead is deposited from inside the pipe. The first group, which includes the Serimer ‘Saturnax’ equipment, the Saipem ‘Passo’ and the Allseas ‘Phoenix’ system normally uses a small root gap at the base of the weld preparation. A removable copper backing strip, which is built into an internal alignment clamp, is used to support the root bead as it is being deposited, Figure 12.. All the weld passes are made from outside the pipe using some form of motorised welding carriage travelling on a steel band (or sometimes a horseshoe clamp arrangement with rack and pinion drive) which is clamped around the pipe. The welding head, and sometimes the spool of filler wire, is carried on the carriage, or ‘bug’ and the welder constantly monitors and controls the positioning of the welding torch in the weld bevel by means of handwheels on the bug, or motorised slides controlled by a remote pendant.

314

Main Menu


Figure 12. Internal welding clamp with copper backing shoes (courtesy Saipem). The second group of systems consists of the long established CRC-Evans system, the Autoweld system, and a newly developed system by Serimer. These systems use a closed root gap, with a small internal weld bevel which is filled by a series of internal welding heads built into an internal alignment clamp, Figure 13, 14, 15. The external weld runs are deposited by bugs similar to those used in the previously described systems.

Figure 13. Internal alignment and welding clamp (courtesy Autoweld Systems).

315

Main Menu


Figure 14. Internal root bead welding on 1200mm diameter pipe (courtesy CRC-Evans).

Figure 15. Schematic of CRC Evans mechanised GMAW weld preparation.

The relative merits of the two types of systems are the simpler internal clamp in the case of the copper backing shoe systems, and the freedom from the danger of copper contamination of the root bead in the case of the combined internal/external welding systems. Both systems use a reduced pipe bevel angle of 5-7 o compared to the 30 o bevel used for manual SMAW welding. This is because the deposition rate of the GMAW welding systems is limited by the maximum wire diameter and current that can be used. This in turn is controlled by the need to maintain a small, fast freezing weld pool, since with these processes there is no slag covering to help support the weld pool against the effects of gravity as welding proceeds around the pipe. High production rate has, therefore, to be achieved in other ways. Hence the narrow joint preparation needing less weld metal, and the use of several welding heads (for example 6 welding torches for the

316

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE internal root bead on a 900mm or 1050 mm diameter pipe). In addition, with the internal root welding systems the external hot pass can be started while the internal root bead is still being completed. Figure 16 compares the weld volume for mechanised GMAW and manual SMAW welds.

Figure 16. Macrosections from mechanised GMAW and manual SMAW welds (courtesy CRC-Evans).

The end preparation for mechanised GMAW is machined on the barge using a purpose made machine and this takes typically 7-8 minutes. Setting up time for the joint is faster than for SMAW, and typical welding times for the root bead and hot pass are half those for SMAW. The filler wire used in GMAW welding is again essentially a plain low carbon steel but with controlled additions of deoxidising elements such as manganese and silicon. The wires are also often microalloyed with titanium to improve toughness and sometimes also alloyed with nickel or molybdenum to improve strength. The shielding gas used for the root pass is usually an argon-carbon dioxide mixture, which gives a good external weld profile. For all other passes either pure carbon dioxide or argon-carbon dioxide mixtures are used. Although the mechanised GMAW welding system produces a weld deposit low in hydrogen so that HAZ cracking should not be a problem, it can produce welds which are susceptible to other types of defect. First, because the process uses low currents and fast welding speeds, it is a low heat input process. This means that the weld cools rapidly and can exhibit high hardness which may lead to stress corrosion cracking problems if the pipeline sees sour service. This high hardness is particularly associated with the capping pass HAZ, which does not receive the benefit of the tempering effect of a subsequent weld pass. In extreme cases in the past when welding high carbon equivalent pipeline steels, it has been necessary to post heat the finished weld or to deposit the cap using SMAW, when the increased heat input helped to reduce the problem. Modern low carbon equivalent steels up to X65 strength level do not usually give HAZ hardness problems. However, with higher strength steels, such as X70 and X80, care may be needed to avoid problems with high cap weld metal and HAZ hardness when using mechanised systems. Secondly, the low heat input, and steep sided weld bevel, of the mechanised GMAW process means that it is prone to lack of fusion defects, Figure 17. This is especially true

317

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE if the welding head is not maintained in a centralised position in the weld bevel, when lack of sidewall fusion can occur. Constant operator supervision, and adjustment, of the welding heads is, therefore, required. As described above, this is usually carried out either directly by means of a thumb wheel on the welding carriage or by servo motors controlled by a separate fly lead. Another problem in the past was that of centre line weld metal cracking. This could occur because of the high depth to width ratio of the weld bead (especially the hot pass), which could lead to segregation of weak impurity films to the weld centre line. These films act as a plane of weakness and could crack during weld bead contraction. The remedy was to control both the weld preparation and the welding variables so that satisfactory weld geometry is obtained. The problem has largely disappeared with these changes and the use of modern clean steels. For small diameter offshore pipelines the mechanised GMAW systems cannot be used and either manual low hydrogen SMAW, semi-automatic GMAW, or semi-automatic flux cored arc welding (FCAW) are alternatives. 5.1.1.3 Semi-automatic Flux Cored Arc Welding As described earlier, semi-automatic gas metal arc welding was first introduced for onshore pipelines but fell out of favour because of problems with equipment and with lack of fusion defects associated with the process. More recently flux-cored arc welding (FCAW) wires have been produced which can be used in place of the traditional solid wires. These wires operate with a different arc characteristic compared to solid wires and are less prone to lack of fusion defects. These flux cored wires can be used with mechanised welding systems but they introduce the problem of slag removal between weld passes, which can be difficult in the narrow weld bevel. Although they have been used for fill and cap welding with such systems (normally using manual root bead welding and conventional weld bevels), the flux cored wires may have more application when used with semi-automatic welding equipment. Flux cored wires for semi-automatic application are available in two main types, gas shielded (GSFCAW), Figure 19, and self shielded (SSFCAW), Figure 20. The gas shielded types are more versatile and tend to produce better weld mechanical properties, but they suffer from a sensitivity to disruption of the gas shield by the wind. The self shielded wires, on the other hand are more tolerant to laybarge welding conditions but the welding procedure needs to be carefully controlled in order to optimise the mechanical properties, particularly toughness. The most common self shielded wires are those produced under the Lincoln ‘Innershield’ brand name. These can be used in the vertical down mode and because the filler wire is on a spool there are fewer stop/starts than with manual welding. Therefore, productivity advantages are claimed with the use of such processes. However, although SSFCAW filler wires which can be used for root bead welding are available, they require a higher level of skill and training than the fill and cap wires. It is, therefore, currently normal practice to weld the root and hot pass with SMAW before filling and capping with SSFCAW.

318

Main Menu


Figure 19. Gas shielded flux cored arc welding (courtesy AWS).

Figure 20. Self shielded flux cored arc welding (courtesy AWS).

Semi-automatic flux cored wire welding has found limited application offshore for the welding of small diameter pipelines, but its major use to date has been for carrying out repairs to mechanised GMAW welds. It has also been used occasionally for depositing the capping pass on mechanised GMAW welds, when there has been a problem with poor weld bead profile with the mechanised GMAW system.

5.1.1.4 One Shot Welding Processes

The previous welding processes are fusion welding processes which use the electric arc as a heat source and use a multi pass welding procedure. This is because the arc has limited penetrating power and an open V butt weld joint design has to be used. The consequence of this is that several layers of weld metal are needed to fill the weld preparation. The size of weld bead which can be controlled by the welder or welding machine depends on the rate of heat input of the welding system (basically the power input (volts x amps) divided by the welding speed) and the ability of any slag system associated with the process to support the molten weld pool until it solidifies. The SMAW process has a relatively high heat input but has a slag covering to support the weld pool. Since the SMAW welding electrode is quite large (6-10mm diameter if the flux layer is included) a large bevel opening is required for electrode access. The welder then adjusts his welding speed and electrode angle to ensure that there is sufficient heat in the weld pool to get good fusion, but not enough to produce a weld pool which will is too big to control. Mechanised GMAW on the other hand has to use a lower heat input than SMAW because there is no slag covering on the weld pool to help hold it in place against the effects of gravity until it solidifies. Fortunately the GMAW filler wire is of smaller diameter (about 1mm) compared to a SMAW electrode so that a narrower weld bevel can be used, which reduces the weld volume required and helps to compensate for the lower heat input of the process. The narrower weld bevel also results in a weld bead which is less wide, but deeper than for a SMAW weld which creates greater surface

319

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE tension forces between the weld pool and the pipe. This again helps to hold the molten weld pool in place until it solidifies. However, despite the reduced weld preparation volume, a number of successive weld passes are still required to complete the joint. A number of new pipe welding processes attempt to speed up the welding operation by completing the joint in one weld pass. These processes are loosely described as ‘one shot’ welding processes, although strictly speaking they should be divided into those that use a high energy density beam to progressively weld the joint in one pass (such as laser and electron beam welding) and those that are truly ‘one shot’ (such as explosive and flash butt welding)(5,6). In laser (7) and electron beam welding(8) a very narrow beam of energy is used, with a parallel sided weld preparation and little or no gap. Usually no filler wire is added, the weld pool being formed by melting of the parent material at the joint line. The processes normally work in the ‘keyhole’ mode, that is the electron or laser beam melts right through the material and emerges from the other side, Figure 21. A column of molten material is produced around the beam, which is held in place by surface tension, and as the beam moves along the joint line this column solidifies behind it. (The action can be visualised by imagining a hot wire travelling through a block of butter). Other variat ions on the processes include the use of small quantities of filler wire to help control porosity.

Figure 21. Schematic of keyhole welding (courtesy AWS). The advantage of electron beam and laser welding are that weld completion times are short, no filler wire is needed, the joint preparation is simple (plain pipe ends), and the process is mechanised. The main disadvantage, for conventional S lay use, is that there is difficulty in controlling the weld pool in all positions around the joint circumference when welding with the pipe axis horizontal. Therefore, the processes have attracted

320

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE most interest for J lay operations, where the pipe axis is horizontal and the forces acting on the weld pool are uniform around the circumference. Alternative one shot welding processes are flash butt welding, radial friction welding, MIAB, SAG forge, homopolar pulse and explosive welding. Flash butt welding has been used extensively for onshore welding of pipelines in the CIS countries, Figure 22, but attempts to interest offshore pipeline operators in the process have so far been unsuccessful. The process relies on the use of a welding alignment clamp through which a low voltage, high current supply from an external power source is applied across the abutting pipe ends. In the case of pipes up to 300mm diameter an external clamp is used, and for pipe greater than 300mm diameter an internal clamp is used.

Figure 22. Flash butt welding of 1400mm diameter pipe in the former Soviet Union (5). The current discharge from the power supply causes arcing at any small gaps between the pipe ends and this arcing causes localised heating of the pipe ends, Figure 23. When the arcing extends over the whole of the pipe circumference the ends of the pipes are brought together by hydraulic rams which are built into the welding clamp. The pipe material in the heated zone is brought together under the applied pressure, without melting, and a solid phase weld is formed. An upset of displaced material is formed on the internal and external diameters of the pipe. The internal upset or ‘flash’ is removed by hot shearing, in the case of large diameter pipes, using special rams which are built into the clamp. The external upset is removed by a separate orbital milling operation.

321

Main Menu


Figure 23. Schematic of flash butt welding (NB. shown for butt welding not pipe).

The advantage of the process is that it is rapid, with a welding time of approximately 3 minutes for a 1420mm diameter pipe and that it does not require skilled welders. However, a major concern for offshore operators is the coarse grained microstructure which is produced in the weld zone as a result of the large thermal input during welding. This microstructure is characterised by low toughness, which would not be acceptable for offshore applications. Therefore, when McDermott attempted to develop flash butt welding for offshore pipeline applications they experimented with an induction normalising treatment which resulted in a significant improvements in toughness. This treatment was applied by an induction coil placed immediately after the welding station, followed by water quenching. The time taken for this treatment was approximately the same as that required for the welding operation. McDermott carried out a series of trials with the equipment in conjunction with Statoil with a view to its use on the Zeepipe project(9). One concern was the occurrence of a softened zone adjacent to the weld where the controlled rolled steel had been heated during welding and normalising. Wide plate tests were carried out on the welds which demonstrated that these soft zones were

322

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE not of practical significance to the overall performance of the weld since they were localised and supported by stronger material on either side. Another concern is the narrow joint line which makes inspection more difficult. The most serious defect in flash butt welds is likely to be a lack of bond or ‘cold shut’ at the weld interface. Since this defect is tightly closed, and is normal to the surface of the pipe, it is not easily detected by radiography. McDermott reported good success in correlating ultrasonic and radiographic test results with weld parameter strip chart recordings. They proposed that such in-process monitoring could be used as an effective quality control tool provided that optimum welding parameters had been previously established. Despite the extensive development programme carried out by McDermott and Statoil flash butt welding was not used on the Zeepipe project since competitive bids for the work were received from barges offering conventional mechanised GMAW systems. Subsequently it is believed that flash butt development work for offshore use has ceased. Radial friction welding is another solid phase welding technique, the heat for welding in this case being created by friction. Bevelled pipe ends are set up with no root gap and a steel collar with an internal profile to fit the bevelled joint is rotated between the pipe ends. As it is rotated, the collar is also radially compressed by means of a die, so that frictional heat is created between the collar and the pipe ends, Figure 24. An internal clamp is necessary to hold the pipe ends together and to prevent collapse of the pipe bore.

Figure 24. Schematic of radial friction welding (10).

323

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE Welding occurs when metal is displaced from the abutting surfaces and clean material is brought together under pressure thus creating a solid phase weld. Rotation of the collar is stopped to consolidate the weld. Radial friction welding is attractive for small diameter pipe welding, since a range of materials, such as carbon steel, stainless steels and even internally clad material can be welded rapidly. Once again no skill is required. However, there are a number of concerns, including assessment of weld properties and non-destructive testing. Ultrasonic testing, in combination with in-process monitoring, has been proposed for quality control, but there are currently no international standards for defect acceptance. Although laboratory versions of the equipment have been under evaluation since at least 1977 (10) it is only in the last few years that serious efforts have been made to develop the equipment for the offshore market. MIAB (magnetically impelled arc butt) welding was originally developed for the mass production industries such as the automotive sector. Thin wall tube (which need not be circular) or pipe is butted together with a small gap between the ends and a magnetic coil is placed around the joint. A direct current is applied across the pipe ends and the arc created then rotates at high speed around the joint under the influence of the magnetic field, Figure 25. The arc heats the pipe ends which are then forged together to make a solid phase weld. The Japanese have developed portable equipment to weld small diameter gas pipelines in urban areas. The UK Welding Institute (TWI) have also built a MIAB machine for Nova of Canada which is designed to weld onshore pipe in the diameter range 80 to 325mm, and to operate at sub-zero temperatures. The equipment has been under evaluation since 1991, but does not appear to have been used commercially to any great extent.

Figure 25. Schematic of magnetic impelled arc butt (MIAB) welding. MIAB welds are similar, in principle, to those produced by flash butt welding and, therefore, similar quality control and inspection problems exist. The Japanese workers used a combination of in-process monitoring, visual inspection, and 2% destructive testing I to check weld quality. The major limitation of MIAB welding for offshore pipeline applications is that it is only capable of joining pipe with a wall thickness of up to 7mm wall maximum, because of uneven heating when it is applied to heavier wall

324

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE components. Therefore, at its current stage of development it would only be suitable for small diameter service pipelines. SAG (Shielded Active Gas) forge welding was developed in the 1980’s by a Norwegian company, AMR Engineering (11). The pipe welding version of the process uses high frequency resistance heating to heat the two abutting pipe ends . The heating is carried out in an atmosphere of hydrogen (the ‘active’ gas) which reduces the surface oxides and allows metal to metal contact. A forge pressure is then applied to consolidate . It is claimed that a wide variety of materials can be joined, including clad pipe. The equipment developed was limited to pipe of 300mm diameter, although much larger diameters could, in theory, be welded. An extensive evaluation of the process was carried out by DNV, on behalf of Statoil, for the welding of duplex stainless steel pipe. Acceptable tensile properties and Charpy impact toughness values of 81-129J at -20 oC were quoted, together with low hardness (250HV). However, despite these promising results it appears that SAG forge welding has not been taken beyond the development stage. Homopolar pulse welding is being developed in the USA and it is a resistance forge welding process which uses a high amperage direct current discharge from a homopolar generator (12). A homopolar generator converts the stored rotational energy of its spinning rotor to electrical energy by means of electromagnetic induction, Figure 26. Once the energy is discharged across the abutting pipe ends a forge cycle consolidates the weld.

Figure 26. Schematic of homopolar pulse welding (10). Pipe welds can be made in under five seconds without the need for filler metal, although the generator charge up time is typically 2-3 minutes. Pipe diameters up to 300mm, and grades up to X65 have been welded and adequate mechanical properties can be achieved although some fine tuning of the joint geometry and process parameters may be required. It is claimed that the process is cleaner than flash butt welding, has a less

325

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE extensive heat affected zone, and a better profile. Once again non-destructive testing techniques and defect acceptance levels remain to be optimised for the process. Homopolar pulse welding development work is being sponsored by a number of operators, particularly with a view to its use for J lay operations. Explosive welding of pipelines has been evaluated for at least 25 years but the first widespread application is still awaited. Explosive welding is another solid phase welding technique, the explosive being used to bring together the two surfaces to be welded in such a way that oxides and impurities are jetted from the surface and the two clean surfaces then bond under the action of the applied force. Two main techniques have been developed, one using an external sleeve into which the two adjoining pipe ends are welded by placing the explosive charge inside the pipe (13), and the other using a bell and spigot joint in which balanced internal and external charges are detonated simultaneously(14), Figure 27.

Figure 27. Section through explosively welded bell and spigot pipe joint A field trial of the latter system was carried out by Premier Lilley Construction on behalf of Trans Canada pipelines in 1984. A 6km length of 1050mm diameter pipe was welded , with an induction heat treatment after welding in order to improve mechanical properties. It is claimed that toughness compares favourably with that produced by conventional welding processes and weld hardness is said to be less than 250HV. Postweld inspection was carried out by a modified RTD Rotoscan system. A 10 minute cycle time for welding is claimed which would make the process competitive with conventional welding systems. One of the main disadvantages of this system is the noise, a problem which is said to be reduced with the Volvo-Nobel sleeve welded system. However, although prequalification trials were carried out under the auspices of the DNV, the system does not appear to have been used in anger. Application of explosive welding to pipe welding on a laybarge does not appear to have been considered, presumably of concern about the novel joint design, as well as the other obvious concerns about noise and safety. The only serious proposal for the use of explosive welding offshore appears to be that of an underwater repair technique but even this proposal seems to have been abandoned.

326

Main Menu


5.1.2 Welding Techniques – Corrosion Resistant Alloy Pipelines

The Corrosion Resistant Alloy materials which are most commonly used for offshore pipelines include : •

Weldable martensitic stainless steels.

•

Duplex and Superduplex stainless steels.

•

CRA internally clad carbon steel.

Since these materials have generally poorer weldability than plain carbon steels, and since it is important that the parent pipe corrosion properties are maintained in the girth weld, more care is required when welding them compared to carbon steels. This means that production rates are slower, and this, combined with the higher cost of welding consumables and parent pipe, means that these pipelines are significantly more expensive to lay than carbon steel pipelines. However, it is important to consider the whole life cost of the pipeline. It may be possible to use a carbon steel pipeline to carry a product which might normally give rise to unacceptable levels of corrosion (e.g wet sweet gas) if the product is dried and/or if inhibitor is injected into the pipeline. However, the cost of such treatments (chemical and plant costs associated with addition and recovery) over the lifetime of the pipeline frequently exceed the initial cost of a CRA pipeline. Also the effect of plant breakdown, or change in future operating conditions, must be considered. The welding techniques which are used for CRA materials are slightly different to those for carbon steel pipelines as described below : 5.1.2.1 Gas Tungsten Arc Welding. The gas tungsten arc welding (GTAW) process consists of a welding torch containing a non-consumable tungsten electrode surrounded by a gas shield nozzle, Figure 28. The electric arc is initiated between the tungsten electrode and the work piece causing heating to occur in both. If DC electrode negative current (workpiece positive) is used then most of the heating occurs in the work piece, creating the weld pool. Since tungsten is a high melting point metal (3410 oC), and since high power torches are water cooled, the tungsten electrode does not melt. Any filler wire which is needed in the weld pool has to be added separately, either by hand in the manual version of the process, or by a wire feed unit. The weld pool and the added filler wire are protected from oxidation by an inert gas shroud which is passed through the gas nozzle surrounding the electrode. The inert gas in Europe is usually argon, and in the USA it is more common to use helium, although mixtures of the two gases are sometimes used. The GTAW process has the advantage that there is no welding slag and the inert gas gives a high quality weld metal. Also, since filler wire is added separately to the main arc the level of heat input to the weld and the amount of filler wire can be controlled independently, unlike the GMAW process. This means that the manual version of the GTAW process can cope with variations in joint fit-up (root gap, root face etc.) and still produce a smooth root bead profile. The disadvantage of the process is that it in manual

327

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE form it is highly skilled, since it is a two handed process (Figure 29), and, because of the relatively inefficient heating method it is very slow compared to other arc welding processes.

Figure 28. Schematic of GTAW welding equipment.

Figure 29. Manual GTAW showing separate addition of filler wire. However, because with CRA pipelines it is critical to obtain a high quality corrosion resistant root bead with the minimum of crevices, the GTAW process is often used. In order to further protect the root bead internal surface from oxidation (which would decrease its corrosion resistance) it is normal with most CRA materials to purge the inside of the pipe during welding with an inert gas in order to remove the air. The purge gas used is normally argon, although in some situations Formier gas, or argon-nitrogen mixtures may be used. The purge gas is restricted to the immediate area of the root bead

328

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE by either inflatable dams, or in the case of most pipelines, by an internal alignment clamp with a special purge chamber located between expandable polymer seals, Figure 30.

Figure 30. Internal alignment clamp containing purge dams (courtesy CRC Evans).

Mechanised GTAW welding is often used in preference to manual GTAW welding since the welding parameters and filler wire placement can be controlled more accurately and less manual skill is required to operate the equipment. For mechanised GTAW the welding torch is positioned on a carriage, together with a miniature wire spool and wire feeder, Figure 31. The equipment can include facilities to oscillate the welding head, automatically control the arc length, and s ometimes pulse the welding current for better weld pool control.

Figure 31. Mechanised GTAW pipe-welding head (courtesy Arc Machines).

329

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE Because the GTAW process is inherently slow, it is usually used for the first few weld passes for large diameter pipe, and then a faster process such as GMAW is used for filling and capping. Another variant on GTAW, plasma GTAW welding is also sometimes used. In this process the tungsten welding electrode is surrounded by a copper nozzle which acts to constrict the arc and produce a more intense source of heat. This process has not seen much use for laybarge firing line girth welding, but has been used for seam welding and double jointing of CRA pipe. 5.1.2.2 Mechanised Gas Metal Arc Welding. The same mechanised GMAW systems used on carbon steel pipelines can also be used with most CRA pipelines. The CRA filler wires do not usually have as good metal transfer characteristics as carbon steel filler wires and so in some cases pulsed welding currents are used to improve this situation. The other disadvantage of mechanised GMAW systems is that they are not as suitable for the critical root bead as GTAW welding. For carbon steel pipelines the GMAW root bead is deposited either from inside the pipe or from outside onto a clamp with copper backing shoes. However, for CRA pipework these options are not used, since pipe diameters are usually too small for the internal root welding option, and clients do not like to use copper shoes since any copper contamination might cause a corrosion cell. Some pipelines have been constructed using mechanised GMAW with ceramic coated copper shoes and some have been constructed using pulsed GMAW root beads without any backing. However, the option to use a GTAW root (and hot pass), together with GMAW fill and cap is also used. This welding process combination is used since any defects associated with GMAW root beads are more significant in CRA alloys as they may act as a site for corrosion. 5.1.2.3 Semi-automatic Flux Cored Arc Welding. Although FCAW consumables for welding some of the CRA materials exist (for example duplex stainless steel), they are not particularly suitable for laybarge girth welding. Therefore, to date FCAW has not been used to a great extent. 5.1.2.4 One Shot Welding Processes. Since some one-shot welding processes are solid phase welding processes (i.e. there is no melting of the components), they lend themselves to the welding of CRA materials where oxidation of the surfaces must be avoided. Also one-shot processes are rapid and can offer significant cost advantages compared to the slower arc welding processes. The two one shot processes which have seen the most development so far for welding of CRA materials are SAG forge and friction welding. SAG forge welding was investigated for the welding of duplex and clad pipes by Statoil, but to date it has not been used in anger offshore. Radial friction welding has also seen development by Stolt Comex Seaway who have developed the system to a stage where ir can be installed on a laybarge (15). The system is suitable for pipe of 150 to 300mm diameter, with a typical

330

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE overall cycle time of 4.5 minutes, which includes line-up, welding, internal machining and non-destructive testing. The actual welding time accounts for 25 to 35 seconds of this total time. 5.1.2.5 Precautions When Welding CRA Materials. CRA materials depend for their corrosion performance on the correct microstructure and composition in the weld region, and they must also meet the usual strength, toughness and hardness limits. There are, therefore, are a number of precautions which must be observed and a number of factors which must be taken into consideration when selecting welding consumables. These are discussed below for the individual materials. 5.1.2.5.1 Martensitic Stainless Steels As discussed in Module 3B, traditional martensitic stainless steels are not readily weldable, without the need to use extended post-weld heat treatments in order to temper the hardened heat affected zone (HAZ). Weldable martensitic and supermartensitic stainless steels are now available which have low carbon contents in order to give satisfactory hardness and toughness in the HAZ after welding. However, if matching filler wires are used to weld these steels then the weld metal can suffer from poor toughness. The alternative is to use an overalloyed filler wire, such as duplex or superduplex stainless steel. These weld deposits will have good toughness and a corrosion resistance at least equal to that of the parent material. However, a weld made with a duplex filler wire may not be as strong the parent pipe and even using a superduplex filler wire does not guarantee overmatching of the higher grades of parent material (the supermartensitics). Also duplex stainless steel loses tensile strength more rapidly than martensitic stainless steel at elevated temperatures, so this may limit the application of the pipeline at high temperatures. Finally the composition of the duplex stainless steel weld metal will vary according to the dilution with the parent material and this may lead to varying properties. Consequently matching filler metals with good toughness are under development (16). Radial friction welding of martensitic stainless steels is an alternative to fusion welding. Tests have been carried out to demonstrate that welds with adequate corrosion and mechanical properties can be produced when radial friction welding consumable rings made from matching martensitic stainless steel are used (17). 5.1.2.5.2 Duplex and Superduplex Stainless Steels In order to ensure that the pipe and weld material have adequate corrosion properties the following precautions are recommended for welding of duplex and superduplex stainless steels (see also Reference 18): •

There should be no contact with carbon steel during fabrication and installation. Local pick up of carbon steel on the surface of the material can lead to a galvanic cell and the initiation of pitting. This means that tools and clamps should be faced with stainless steel or non-metallic material, and grinding discs should not have previously been used on carbon steel.

331

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE •

The consumables used should match the base material, with a slight increase in the amount of nickel in order to maintain the correct phase balance in the weld metal (a 50/50 austenite/ferrite ratio as in the parent material is desirable but 40/60 is usually acceptable). For the welding of 22%Cr duplex stainless steel it is sometimes the practice to weld the root pass with the higher alloy 25%Cr superduplex filler wire, before filling the remainder of the joint with a 22%Cr filler wire. In this way the corrosion properties of the root bead are improved and therefore safeguarded somewhat against variations in the welding procedure.

•

The bore of the pipe should be purged with inert gas during welding to reduce oxidation of the root bead. It is not realistic to expect all the oxygen in the pipe bore to be removed during purging and it is normal practice to measure the oxygen level in the bore immediately before welding using a purge meter. A practical maximum level of oxygen to aim for is 0.5%, as any less is associated with a low partial pressure of nitrogen in the purge and this can lead to a loss of nitrogen in the root bead. Since the pitting resistance of the parent metal and the weld depends on nitrogen, then the corrosion properties of the root bead can suffer if purging is carried out too thoroughly. An alternative approach is to use a shielding gas (and possibly purge gas) with a deliberate addition of nitrogen of 3-5%. The purge gas should also be maintained until at least 5-8mm of weld is deposited so that the heat from subsequent weld passes does not oxidise the root bead.

•

There should be no preheating before welding, and the maximum interpass temperature should be controlled to 100-150 oC, depending on pipe wall thickness and grade of duplex. This is required in order to prevent very slow cooling rates in the weld and HAZ which would lead to the precipitation of unwanted intermetallic compounds in the microstructure. These intermetallic compounds reduce the corrosion properties and toughness of the joint. Precautions for superduplex pipe are usually more severe than those for duplex pipe, since the higher alloyed steel is more susceptible to precipitation than the lower alloyed steel.

•

The heat input from the welding process should be controlled to a specified maximum in order to control the cooling rate of the weld. Again, the maximum heat inputs for superduplex steel are usually lower than for duplex stainless steel.

•

A ‘cold pass’ technique should be used. This means that the heat input for the second (cold) pass should be lower than that for the root pass. Since the weld bead size is related to the heat input (large heat input equals slow welding speed, equals large weld bead size) this means that the root bead should always be thicker than the second pass. This ensures that the heat from the second pass does not penetrate through the full thickness of the root bead and cause oxidation and precipitation effects.

•

Root pass repairs are usually not allowed on duplex and superduplex pipelines because of the difficulty in re-establishing the purge inside the pipe. They should only be allowed if it can be demonstrated that they can be carried out without prejudicing the integrity of the corrosion performance of the joint. Repairs to the body of the weld should be carried out with a sufficient ligament of weld metal

332

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE beneath the repair that the internal root bead surface is not oxidised by the repair weld heat input. Most duplex stainless steel pipelines have been welded using the manual or mechanised GTAW process for the root bead and then either filled with the same process (in the case of thin wall/small diameter pipe) or filled with the GMAW process (in the case of thicker wall/large diameter pipe. Procedures have also been developed for mechanised GMAW root bead welding of duplex stainless steel using pulsed welding currents and an internal gas purge, or ceramic coated copper backing shoes. 5.1.2.5.3 Clad Pipe Cladding materials can be austenitic stainless steels, such as grade 316L, duplex stainless steel, or one of the high nickel alloys, such as Incalloy 825 or Inconel 625, depending on the service conditions. The CRA clad layer is usually about 3mm thick and the backing steel is normally an API grade carbon steel of whatever thickness is required to withstand the maximum operating pressure. There are two main ways to weld clad pipe: •

Use a high alloy CRA consumable throughout the joint. The consumable should be designed to match, or overmatch, the clad layer and cope with dilution of the carbon steel host pipe (dilution of a carbon steel into a CRA material can reduce its corrosion resistance and result in high hardness unless the CRA material is chosen carefully). This is an expensive option and it can be argued that the corrosion properties of the CRA filler material are not needed in the filling runs of the joint. However, the advantage of this approach is its technical simplicity, there being only one consumable type used throughout the joint.

•

Use a high alloy consumable for the root pass, and then use a pure iron buffer layer for the second pass, before switching to a carbon steel consumable for the remaining filler and cap pass. Although potentially cheaper than the first option, this technique has the disadvantage that it is more complicated, and the integrity of the joint depends critically on the effectiveness of the buffer layer in preventing mixing of the CRA root weld metal and the carbon steel filler weld metal. It has been suggested that pulsed GTAW welding can be used for the buffer layer, since this is a low dilution process. However, even with this process there is a danger that dilution of the CRA material into the buffer layer may result in a hard and brittle zone in this region. Also any breach in the CRA root bead (for example due to an undetected root defect such as lack of penetration) would result in contact between the aggressive media and the non-CRA part of the weld.

As a result of concerns about the second, mixed welding, approach only a handful of clad pipelines have been welded in this way, most having been welded using the single, overalloyed consumable, route (19).

333

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE 5.1.3 Inspection Techniques

On a conventional laybarge one station is devoted to the inspection of the girth welds. Traditionally this inspection has been carried out by X ray film radiography and, as already mentioned, for the early pipelines, viewing was carried out on wet films. The reason for this was that each operation on a conventional laybarge is a critical path event, since the laybarge cannot move forward until the operations at each station are completed. The time taken to position the film around the weld, expose the film, remove it and develop it and then view the film for defect acceptance (‘sentencing’) was therefore critical. Initially it was found that the only way that radiography could keep pace with other operations, such as welding and field joint coating was to view the films as soon as they had been developed, whilst they were still wet. The viewing of wet films was not ideal and it was found that when the films were reexamined in the dry condition important defects (mainly tight linear defects such as cracks and lack of fusion) had been missed. As a result effort was put into introducing rapid developing systems, together with rapid drying. This meant that dry films could be produced within the 8-10 minutes cycle time required to match the other laybarge operations. As a result the standard of radiography was much improved. However, even with good practice radiography suffers from some well-known shortcomings. In particular, it is only capable of detecting tight defects if they are in line with the beam. If the X ray head is slightly off line with the centre of the weld, or if the defect is inclined from the vertical, then it is possible to miss the defect. Also radiography does not give an accurate measure of defect depth and for this reason the common welding standards use defect length and type as the basis on which to sentence the welds. For example BS4515 limits surface breaking linear defects (e.g. lack of penetration) to 25mm length in any 300mm length of weld, and buried linear defects (e.g. lack of fusion) to 50mm length in any 300mm length of weld. This is on the basis that surface breaking defects are twice as significant from the point of view of fracture of the weld as buried defects, so the allowable length is halved. The above limits are known as workmanship limits, that is they are defect le vels which a competent welder or welding system ought to be able to meet. Since the welds to which they have been traditionally applied are multipass welds it is assumed that defect depths are limited to one weld pass which in most welds is only a small proportion of the pipe wall thickness. Usually when the weld defects exceed the workmanship limits the weld is repaired. However, sometimes this may be difficult or costly to carry out a repair if a critical weld is affected (for example a tie-in weld). In this case it is possible to carry out an engineering critical assessment (ECA) of the defective weld using fracture mechanics principles. The important parameters, which must be known for these calculations to be carried out, are the stress on the weld, the material fracture toughness and the defect dimensions, including depth. When radiography is used the defect depth is not measured, but it can usually be assumed that defects are limited to one weld run in depth, which is about 3mm and this depth value is normally used for the purposes of ECA calculations.

334

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE Manual ultrasonic inspection, where the operator moves an ultrasonic probe around the joint by hand, is able to give information about defect length and depth although it is not usually possible to identify the type of defect. Also manual ultrasonic inspection is slow and prone to variation in the level of accuracy due to human error or operator fatigue and has no permanent record. For this reason manual ultrasonic inspection is not widely used, but in the last ten years mechanised ultrasonic systems have been developed which use multi-probe heads mounted on a track surrounding the pipe. The systems are best suited to the inspection of girth welds made with mechanised welding processes, since the weld geometry is more consistent than for manual welds, allowing easier interpretation of the results. With mechanised ultrasonic system the opportunity for operator error is reduced and a hard copy print out of the weld scan is produced. The systems which have been developed are able to inspect a typical girth weld within the required time interval for laybarge applications. A final fact in their favour is the absence of health and safety concerns compared to radiography, since on a laybarge care has to be taken to screen operators from the ionising radiation associated with radiography. A recent development has been real time radiography, where the photographic film is replaced by a low light level video camera, so that the image can be viewed in real time, or near real time. The ability to manipulate the image electronically is also an option, so that image enhancement methods can be used and the results can be stored on CD ROM. There is also the option, where X ray source alignment is critical, to move the source whilst viewing the effect on the screen. In this way planar defects which might normally be missed could be detected. However, there are a number of disadvantages to the systems. Firstly, the image is normally not as good as a radiographic film image, since the resolution is limited by the pixel size on the screen (similar to the limitation of current digital cameras compared to photographic film). Secondly, the systems are not truly real time, near real time being a better description. As a result the systems have not seen extensive use offshore, although there have been some specialised applications where they have been used. For example, many duplex stainless steel pipewelding standards call for ‘no root defects’. Since many standards also forbid the repair of root defects it is cost effective to examine each girth weld after completion of the root run to check for defects. In this way if defects are present it is possible to cut out the weld immediately without the wasted effort of completing the whole weld only for it to be cut out. Real time radiography has been applied to the examination of the root run in the partially completed weld. The types of defects found in pipeline girth welds, and the details of the techniques used to detect them are described in the next sections. 5.1.3.1 Defect types Defects which occur in pipeline girth welds can be: •

•

those which are due to human error such as poor workmanship, or failure to follow qualified procedures. those which are due to poor materials, such as defective pipe or welding consumables.

335

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE •

•

those which are due to external influences, such as poor weather conditions or excessive handling stresses during pipelaying. those which are due to a progressive cause such as corrosion or fatigue in service.

Common defects, which are mainly from the first two categories, are shown in Figure 32a-t. Not shown in these diagrams are arc strikes, which are local hard spots on the external surface of the pipe caused by accidental striking of the welding electrode on the pipe surface rather than in the weld bevel. Arc strikes may also be caused by arcing between the pipe and the welding return connection, or by the use of electri cal prods for magnetic particle examination. The use of prodes has largely been superceded by electromagnetic yokes and this problem is then avoided. Arc strikes are normally treated by grinding them to remove any hardened layer, followed by magnetic particle examination to ensure freedom from cracking. Mismatch (Figure 32a) is caused by ovality in the pipe or failure to use a clamp, or adequately support the components before welding. Clamps will ‘round out’ some pipes, particularly those of high diameter/wall thickness ratio and low grade, but not if the pipes are of heavy wall thickness or high strength. Excessive mismatch can be associated with other defects, such as lack of penetration, as shown in Figure 32b. External concavity, or lack of fill, Figure 32c, is caused by an insufficient number of weld passes, and can occur at the 3 and 9 o’clock positions especially with manual welding if a stripper pass (an additional filler pass over a small part of the pipe circumference to counteract concavity) is not used. Excess root penetration, Figure 32d, is caused in manual welding by incorrect electrode manipulation and travel speed (the welder controls the level of penetration of the root bead partly by altering the angle of the welding electrode to close or open the ‘keyhole’). Cap undercut is a region at the side of the cap where the pipe material has been melted and washed away, leaving a groove, Figure 32e. This is caused by too high a welding current and incorrect electrode manipulation. Root undercut is similar to cap undercut but the groove is on the internal surface of the pipe, Figure 32f. Root concavity, or suck back, is most common at the 6 o’clock (overhead) position and is caused by too high a current or the root gap being too large, Figure 28g. Usually this defect is relatively unimportant, provided the weld thickness is not reduced too much, as it has a smooth profile. A burn through, or window, is a localised severe form of root concavity, caused when the root bead weld pool fails to fill the keyhole formed during welding, leaving a small hole in the root bead. This hole is then covered, but not filled by, the hot pass, leaving a ‘window’ shaped depression in the root bead, Figure 32h. Lack of penetration is distinct from concavity in that the unmelted edges of the machined root face are still visible, Figure 32i. This defect can be caused by too low a

336

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE welding current, wrong electrode angle, or too tight a root gap. This defect is normally limited in depth to the depth of the root face, about 1.5mm. Slag inclusions are ‘body of the weld’ defects caused by too low a welding current or incorrect cleaning of the weld bead surface between weld passes, Figure 32j. If these inclusions are linear and on each side of the root bead, they are known as waggon tracks, because of their appearance on the radiograph, Figure 32k. Waggon tracks are a feature of manual welding with cellulosic electrodes. They are caused by failure to grind the external surface of the root bead (which is normally convex) sufficiently between weld passes so that the hot pass is unable to melt out any remaining slag at the edges of the bead. Lack of sidewall fusion is caused by insufficient melting of the faces of the weld preparation, Figure 32l. This can be due to too low a welding current, too large a weld pool, or wrong electrode angle. Mechanised GMAW welding systems are particularly prone to this type of defect because of the steep angle of the weld bevel and the low heat input welding system. The fact that the lack of sidewall fusion is tight and inclined slightly to the radiographic beam means that it may not show up well on the radiograph. However, lack of sidewall fusion is normally limited in depth to one weld run, and provided this is the case ECA analyses often demonstrate that long lengths can remain in the weld without posing a threat to the integrity of the pipeline. Lack of interrun fusion occurs between weld passes, and so is normally at right angles to the radiographic beam, making detection more diffic ult, Figure 32m. Scattered porosity, Figure 32n, can be caused by too low or too high a moisture level in the welding electrode flux covering (cellulosic electrodes need some moisture in the coating to provide the shielding gas), or the welder holding too long an arc. If the porosity is surface breaking it can also be caused by too much weaving of the electrode or overheating of the weld pool. In the case of GMAW disruption or contamination of the shielding gas can also result in porosity. Cluster porosity, Figure 32o, can be caused by isolated contamination of the weld faces, local loss of shielding gas, or failure to use the correct technique at the weld stop/start positions. Elongated root porosity, or hollow bead, is caused by a failure of the weld to outgas as it solidifies, Figure 28p. This is caused by too small a root gap, too fast a welding speed, and sometimes by excessive amounts of some alloying elements in the parent material. Transverse hydrogen cracks, Figure 32q normally occur in the weld metal of high strength steels and may lie on alternate 45o planes (hence the term ‘chevron’cracks). They can be difficult to detect by radiography since they are usually very tight and are aligned parallel with the natural ripples on the weld bead surface. Longitudinal weld metal cracks, Figure 32r, can be due to hydrogen but are usually solidification cracks caused by a high depth to width ratio weld bead shape, and/or stress on the weld as it solidifies.

337

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE Root heat affected zone cracks, Figure 32s, are usually due to hydrogen, although they can also be caused by excessive stress on the pipe during laying. Although they usually initiate and propagate for some distance in the heat affected zone, they can also propagate into the weld metal further up the weld. Hydrogen cracks are some of the most critical to the integrity of the girth weld because their extent is unpredictable and they can be missed by radiography because of their orientation and tightness. The precautions adopted in the construction welding procedure to minimise hydrogen cracks have already been mentioned (preheat, interpass temperature and time control etc.). However, another cause of hydrogen cracks is the use of poor repair welding procedures. It used to be the practice on large diameter pipes to repair any root defects by single run back welds with cellulosic electrodes. Since these were of low heat input, and it was difficult to apply preheat when carrying them out, hydrogen cracks were often introduced. It is now common to specify that such welds should consist of at least two weld runs, in order that some tempering of the heat affected zone in the pipe takes place. The exception to this practice is for mechanised GMAW welds when a misfire occurs on one of the welding torches used on the internal root bead. If a misfire occurs then the missing part of the root bead is replaced with a single weld run, using low hydrogen vertical down electrodes or semi-automatic gas metal arc welding. In this case the repair weld run is tempered by the later external weld runs.

338

Main Menu


Figure

32a.

Mismatch

or

hi-lo

Figure 32b. Mismatch with lack of penetration.

Figure 32c. External concavity or lack of fill.

Figure 32d. Excess penetration

Figure 32e. Cap undercut

Figure 32f. Root undercut

339

Main Menu


Figure 32g. Root concavity.

Figure 32h. Burn through or ‘window’

Figure 32i. Lack of penetration

Figure 32j. Slag inclusions

Figure 32k. Elongated slag lines or ‘waggon tracks’

Figure 32l. Lack of sidewall fusion

Figure 32m. Lack of interrun fusion

340

Main Menu


Figure 32n. Scattered porosity

Figure 32o. Cluster porosity

Figure 32p. Root pass aligned porosity, or ‘hollow bead’

Figure 32q. Transvserse weld metal crack

Figure 32r. Longitudinal weld metal crack.

Figure 32s. Longitudinal root heat affected zone crack

5.1.3.2 Visual Inspection Visual inspection is an often overlooked part of the overall inspection procedure. If the external appearance of the weld is neat, of smooth regular profile, and free from arc strikes and undercut, then there is a high probability that the weld is of good quality. In the case of pipeline girth welds, an experienced Welding Inspector can also tell the

341

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE welding direction and the class of electrode used, from a visual examination of the weld cap profile. This helps to determine if there has been any deviation from the welding procedure. Visual inspection should also be applied during the welding operation as well as at the end, since factors such as the standard of interpass weld cleaning can influence the final weld quality. There are a number of aids to visual inspection, including low power magnifying lenses, and weld inspection gauges for the measurement of depth of undercut, cap height, misalignment etc. Also available are borescopes and internal cameras, which can be used to examine the root bead. Although these are not used routinely on offshore pipelines they may be useful in the case of a disputed repair on small diameter pipe where internal access is otherwise not possible. 5.1.9.4 Radiography Characteristic pipeline defects, such as slag, porosity, undercut, lack of root penetration, burn throughs and arc strikes are three dimensional and can be found fairly easily by radiography. Narrow, two dimensional defects, such as cracks, especially those which are not aligned parallel to the radiographic beam, can be missed by radiography especially if best practice is not used. Since three dimensional defects are less significant than two dimensional defects from the point of view of structural integrity, this failing would appear to be a major disadvantage of radiography. However, radiography continues to be the main NDE technique applied to pipeline girth welds, and the satisfactory performance of most pipelines suggests that it is effective. This is probably because: a) the improved weldability of modern materials means that cracking is now relatively rare. b) the deterrent effect of radiography is sufficient to encourage welders to follow the qualified procedure and this reduces the chance of two dimensional defects as well as three dimensional defects. c) the use of qualified radiographic procedures and experienced interpreters means that all of the significant defects can be usually be found. Two forms of radiographic source can be used, X ray and gamma ray. The former has better sensitivity and is now normally specified for most girth weld inspection. The latter requires less specialised equipment but has poorer sensitivity and can therefore miss more defects, especially cracks. For this reason it is only used for special applications, for example for single wall examination of welds in very small diameter pipework, where the small size of the gamma source allows it to be placed inside the pipe, and on very thick welds, where the greater penetrating power of gamma radiography compared to X radiography is an advantage. A variety of radiographic techniques may be used, depending on the pipe geometry and the equipment available. For large diameter pipes the so-called panoramic technique is

342

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE used, Figure 33, where the X ray source is placed centrally inside the pipe and the radiographic film is placed in contact with the outside surface of the pipe.

Figure 33. Panoramic Radiography

FILM

S

F

S = Source of radiation F = Focus-to-film or Source-to-film distance

For offshore pipelines the X ray source is built into a crawler, so that the source can be carried along the pipeline from one girth weld to the next, Figure 34. The crawler is designed to stop when it recognises a signal from a low s trength gamma radiation source which is placed on the outside of the pipe at the weld to be examined. The radiographic film is then wrapped around the outside of the girth weld and exposed to the radiographic beam from the crawler, before being developed. The crawler than moves along to the next girth weld and the process is repeated.

Figure 34. Internal pipeline crawler. The panoramic technique gives a rapid result (one exposure). However it may sometimes be necessary to use the opposite approach, by placing the source outside the pipe and the film in contact with the weld root on the inside wall of the pipe, Fig 35. This is the single wall single image technique (SWSI) and is used when the pipe diameter is too small to give the correct focal length for an internally placed X ray source, or access to the centre of the pipe bore for the X ray source is restricted. This technique is mainly applicable to welds between short pup pieces of pipe, so would not be appropriate for offshore pipeline construction on a laybarge.

343

Main Menu


Figure 35. Single Wall Single Image Radiography. FILM

F S = Source of radiation F = Focus-to-film or Source-to-film distance S

If no access at all to the bore of the pipe is possible, e.g. when radiographing a girth weld in an in-service pipeline, then a double wall technique has to be used. The simplest technique is the double wall single image technique (DWSI), whereby the source is placed remotely from the pipe surface on one side of the pipe, slightly offset from the weld, and the film is placed in contact with the external surface of the weld on the opposite side of the pipe, so that a single image of the weld adjacent to the film is produced, Fig 36.

FILM

FILM

X °

90

F X

SECTION ON X-X

S

S F

/5

S = SOURCE OF RADIATION F = FOCUS-TO-FILM OR SOURCE-TO-FILM DISTANCE

Figure 36. Double Wall Single Image Radiography

344

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE For small diameter pipe (less than 90mm) the DWSI technique is not applicable and the double wall double image (DWDI) technique is used. This is similar to DWSI but the film is flat and is placed tangential to the pipe, so that it is only in contact with the weld at one point rather than over its whole length. The source is again offset but is placed so that images of both the near and far weld sections is produced alongside each other on the one film, Fig 37. Single wall radiographic techniques give radiographic results which are easier to interpret than double wall techniques and are to be preferred where applicable. FILM

FILM (OFFSET)

X °

90

X

SECTION ON X-X F

S

S F

/5

Figure 37 .Double Wall Double Image Radiography The quality of the radiographic image depends on many factors, such as the source to film distance, the energy of the source and the exposure time, the grain size of the film, and the film developing conditions. The important criterion is the resulting sensitivity of the radiograph and this can be measured by placing a penetrameter, or image quality indicator (IQI) alongside the weld during the exposure of the film. The DIN IQI consists of pairs of wires of decreasing diameter trapped inside a plastic holder. The sensitivity is assessed from the finest IQI wire which can be recognised on the film and should be compared with the allowable values given in the specification, for the appropriate wall thickness and radiographic technique used. The density of the radiograph should also be measured using a densitometer and compared with the allowable range in the specification. Interpretation of radiographs requires care and experience. The film viewing conditions should be correct with regard to film illumination and background lighting, and the inspector should allow some time for his eyes to become adjusted to the level of illumination before reading the radiograph. A knowledge of the cause and distribution

345

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE of weld defects is necessary to interpret the two dimensional image of what is usually a three dimensional defect. Some features in particular can cause confusion. One of these is the root bead slag intrusion, a surface breaking feature of limited depth found at the edge of the root bead which occurs occasionally when using cellulosic electrodes, Fig 34. It is often difficult to distinguish this feature on a radiograph from a root bead hydrogen crack, or fine lack of root fusion (missed edge) and in some cases it may be necessary to ‘grind and investigate’ the defect to see if it disappears within acceptable metal removal limits.

Figure 38. Root bead slag intrusion

5.1.9.5 Ultrasonic Examination Manual ultrasonic examination can be used as an aid to interpret a radiographic indication, especially on thick wall pipe. The manual version is not suitable for use as the primary inspection technique for pipeline girth welds since it is slow and dependent to a large degree on operator interpretation. Also the degree of operator concentration involved is very great and this can give rise to inconsistent results, between operators, or even between examinations when using the same operator. It is not suited, therefore, to making a large number of routine inspections, such as would be required on a laybarge. The interpretation of root bead indications in SMAW girth welds is also difficult with manual ultrasonics since mismatch, root undercut, excess penetration, and root hydrogen cracks can all give rise to similar reflections. In order to accurately position the defect, assumptions must be made about the exact length of the beam path from the probe to the weld root and this is difficult unless the outside surface of the pipe is marked in some way before welding to indicate the distance from the weld bevel. Even then the root gap with a manual welded joint can vary around the circumference and

346

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE from weld to weld, so that accurate defect location with respect to the finished root bead profile is difficult. When mechanised GMAW welding processes are used, the weld fit up and the resultant weld geometry are much more consistent. Also one of the most common types of defect found in mechanised GMAW welds is lack of sidewall fusion, Fig 39. These defects are tight and aligned at a slight angle to the through thickness direction. As such they can be missed by radiography, but lend themselves to detection by ultrasonics.

Figure 39. Lack of sidewall fusion in mechanised GMAW weld.

Multiprobe mechanised ultrasonic systems have been developed which run on the same track as the mechanised GMAW welding carriages, Figure 40. This means that the location of the probes with respect to the weld bevel is known and the defects can be identified by their position in the weld. Such systems have been used extensively on large diameter mechanised GMAW welded pipelines, initially in conjunction with radiography, but more recently as the primary inspection technique. A major advantage of such techniques is that they can operate just 6 welds or so behind the front end of the pipeline and therefore there is a rapid feedback on weld quality compared to radiography, where a whole days production (possibly more than 100 welds) may welded before the radiographs are taken and interpreted.

Figure 40. Mechanised ultrasonic system for pipeline girth weld inspection.

347

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE 5.1.9.6

Magnetic Particle Inspection Magnetic particle inspection (MPI) is only suitable for the detection of surface breaking defects, and so cannot be used as the primary inspection method for girth welds. Where access is available it may be used, however, as a support to radiography for the inspection of root beads in order to confirm freedom from hydrogen cracking. The inspection of fillet welds is not possible with radiography, and it is difficult with ultrasonics because of the subjective nature of the technique. Therefore, MPI is used to inspect the fillet welds on attachments such as weldolets, CP connections, pipe supports, hot tap split tees etc. This inspection will only confirm freedom from surface breaking defects. The body of the weld is not usually inspected, since most of these fillet welds are multipass welds. Any buried defects in such welds would, therefore, be expected to be limited to one weld run in depth (about 3mm) and therefore not significant structurally.

5.2 J LAY

The deepwater pipeline projects carried out to date mostly include small diameter pipe (250-300mm diameter), with the exception of the Shell Mars 450mm diameter pipeline. Future projects will entail laying large diameter pipe, up to 900mm diameter, in water depths up to 2,000m. If there is a need to lay rigid large diameter pipe in water depths greater than 1000m or so then a conventional S lay poses the risk of a pipe buckle or fracture, since the weight of the pipe causes a bending moment on the pipe as it leaves the stinger. A certain amount can be done to extend the capability of the conventional S lay barge by increasing the power of the pipe tensioners and extending the length of the stinger so that the pipe enters the water at a steeper angle. However, for very deep water in order to reduce the stresses on the pipe during deepwater laying it is necessary to eliminate the overbend region by arranging for the pipe to be vertical or near vertical as it enters the water. This usually means that the pipe must be welded, inspected and coated in a much smaller number of stations than on a conventional barge. This method of pipelaying is known as J-lay, since the pipe follows a ‘J’ profile as it progresses from its vertical position on the barge to its horizontal position on the sea bed, see Figure 1. The possible water depths for laying, versus the pipe diameter, for S and J-lay barges are shown in Figure 41. The main limit in the case of J-lay is the ability of the pipe tensioners to support the submerged weight of the pipeline. These tensioners must be capable of gripping not just conventional hard anti-corrosion pipe coating, but also soft thermal insulation.

348

Main Menu


Figure 41. Lay capability envelopes for S lay and J lay(20). There are few purpose built J-lay vessels, most projects involve a modular tilting tower which is positioned on the bow, midships (through a moon pool) or stern of an existing vessel for specific projects(20,21). Since the number of workstations is limited by the height of the tower, it is normal to weld 2-6 pipe lengths horizontally on the deck (or onshore) using a high speed welding system such as submerged arc welding. The pipe string is then transferred to the vertical position and lined up with the end of the pipe string already in the tower. The tower normally contains the welding, NDT and field joint coating stations as well as the tensioners. Provision must also be made for abandonment and recovery equipment. 5.2.1 Welding Processes

Since there may only be one welding station on a J-lay barge it is important to maximise the productivity of the welding process. For this reason great interest has been shown in the use of one shot welding processes for J-lay. Laser and electron beam welding processes are applicable, since the welding position means it is easier to control the weld pool than it would be with the pipe axis horizontal. In the past 20 years several attempts have been made to develop laser welding systems for both S and J lay pipelines. To date the work has not been successful for a number of reasons, including poor weld quality (principally problems with porosity), poor mechanical properties, and difficulties with the optical guidance systems. Some of these problems can be solved by the addition of filler wires and the state of the technology is now such that laser welding may be back in contention for J-lay work. The limit on wall thickness is around 15mm for S laying and up to 30mm for J laying. Electron beam welding has also been under development for pipeline welding for many years. Total built a prototype system for J-laying in the early 1980’s and produced more

349

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE than 3000 trial welds in 600mm diameter pipe in wall thickness up to 32mm and grades up to X100 (22). A limitation of the electron beam process is the need to have the electron gun and workpiece in a vacuum chamber for high quality work. This means there is a need for seals around the pipe, and there may be a delay until the required vacuum is obtained. In order to save pumping time the Total equipment used a double vacuum chamber arrangement, the gun working at high vacuum, but the pipe being in a lower vacuum. The equipment was installed on a land based test bed which could be rocked to simulate severe sea states, but there is no record of it being used in anger on a laybarge. It is understood that the main reasons for this were difficulties with mechanical property requirements and the problems of maintaining the vacuum. More recently TWI and Saipem have been developing a low vacuum electron beam pipewelding system for use offshore. Trials are currently underway, with the aim of using the equipment on a J-lay barge. So far it has been found that it is necessary to add filler wire to the joint in order to improve the toughness to acceptable levels for offshore use. Since both these one-pass welding systems are still under development, attempts have been made to speed up conventional mechanised welding systems so that high weld production rates can be achieved at one weld station, instead of the 4 or 5 weld stations on existing vessels. This has been achieved by the use of multiple welding heads and by computer control systems which allow more rapid start up of the welding arcs. Care must be taken to control the maximum interpass temperatures of the welds to avoid deterioration in mechanical properties(23). Another way to overcome the problem of the limited number of fabrication stations in the J lay tower is to double, triple, or quadruple joint the pipe lengths in the horizontal position on the barge and then move the pipe to the vertical position in the tower for joining onto the main pipeline(20,21).

5.3 COILED LINE PIPE

The idea of installing offshore pipelines by spooling them onto a reel at a shore base and then unspooling the pipe at the installation site is not new. During the second world war fuel supply pipes were laid across the North Sea following the Allied invasion of Normandy. The project was called PLUTO : Pipe Line Under The Ocean. The steel pipes were 75mm diameter and were welded together in lengths up to 1.2 km and then wound onto drums of 12m diameter. The drums were towed behind ships and unreeled when required(24). For small diameter pipe (<450mm) there is the possibility of either welding the pipe strings together at a shore base as in the case of Pluto, or producing the pipe in continuous form before coiling. The drums onto which the pipe is reeled are of various diameters and can either be permanently mounted on the laying vessel, as in the Stena Apache, or can be dismounted to allow reeling to take place at the shorebase without tying up the vessel.

350

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE The limitation on the length of spooled pipe is either the crane capacity for lifting the spools onto the vessel, or the size of spool that can be accommodated on the deck of the vessel. In the case of continuous coiled pipe the pipe is made from strip which is formed and welded using the ERW/HFI welding process. In the case of butt welded spool sections individual pipe lengths of 12m are welded together using either manual or mechanised welding, inspected, field joint coated and spooled. A large area of land is needed to lay out the various work stations in a similar way to the layout of the firing line on a laybarge, and this is usually at a remote shorebase with good docking facilities. One of the obvious concerns is the fact that to reel and unreel the pipe it is necessary to apply plastic strains of up to 5% to the pipe and this could cause ovality and buckling. Also any strain ageing after this deformation could cause a reduction in toughness and an increase in yield strength of the pipe. For this reason it has become common practice to qualify pipe materials and welding procedures for reeled pipe on a bending rig prior to the start of a contract (25). Another disadvantage of coiled pipe is that weight coating is not possible so the pipe may not have the required on bottom stability unless the pipe wall thickness is increased.. The main application of coiled pipelines is for in-field flowlines, for example from the well to a transmission pipeline or to a gathering point. Pipeline lengths from around 1km to up to 20km have been laid using the spooling technique. The advantages of coiled pipe are the rapid laying time (useful if there is a short weather window), flexibility in the type of laying vessel which may be used, and improved flow characteristics in the case of continuous pipe because of the absence of girth welds. The Apache has laid 250mm diameter reeled pipe in a depth of 705m for Petrobas. 5.4 LANDFALLS

With rigid pipelines especially, it is not normally possible to lay pipe from the shoreline to the other shoreline (in the case of a crossing) , or offshore platform (in the case of an export pipeline) in one operation. Usually the pipe is laid from the platform, or far shoreline, to the near shore, and the pipe has to be positioned on the sea bed about 1km from the shore to allow the laybarge space to turn around. It is then normal practice to lay back to the shore from this point by winching the pipe from the barge (shore pull). The two pipe ends are then tied in either on the barge, or by means of a hyperbaric weld on the seabed (see next section). The land fall section itself may need to be in a trench or cofferdam if crossing a beach, a tunnel to gain access to any cliffs, if present, and a shaft to reach the top of any cliffs. Alternatively if the landfall area is congested with dunes, or rail and roadways, directional drilling may be the best option. The pull-in is carried out by winching using a pulling head welded to the end of the pipeline. Damage to the pipeline coating during pulling is an obvious concern and the coating of this section of pipe must be chosen with this in mind. The friction between the coated pipe and the touchdown points must also be assessed carefully, since this will affect the winch loads.

351

Main Menu

PIPELINE ENGINEERING: MATERIALS & WELDING MODULE: OFFSHORE 5.5 TIE-IN OPERATIONS

An offshore pipeline may involve several tie-ins. These include, for example, the shore approach, the tie-in to the riser pipe at the platform and any mid length tie-ins associated with pipe laydowns due to separate construction phases/seasons. Traditionally these tie-ins have been carried out by means of a hyperbaric welding operation. This involves placing pipe handling frames on the sea bed, cutting the pipe ends (to allow either a straight tie in or the insertion of a pup piece) manipulating the pipe ends into line, lowering a hyperbaric chamber over the tie-in location, and then welding, inspection and field joint coating of the tie in. The alternative approach, and one which has been used increasingly because of the cost of hyperbaric operations is to lift the pipe ends and carry out the tie-in above water on, or at the side of, the barge. A typical sequence of events for this operation used by European Marine Contractors Ltd on the English Channel Interconnector pipeline (26) is: i. ii. iii. iv. v. vi. vii. viii. ix. x. xi. xii. xiii. xiv. xv.

attach buoyancy tanks and lift clamps during pipelay raise the first pipeline using the barge davits remove laydown head and prepare pipe end raise second pipeline remove head and cutback so that the pipes are side by side but overlapping cut the second pipe at the appropriate tide bevel second pipe and bring the ends together place hydraulic alignment clamp achieve line-up and commence welding remove clamp radiography (double wall single image X ray) apply corrosion protection heat shrink sleeves strap wood to pipe for impact protection lay completed pipes back on seabed remove buoyancy and lift clamps

The main dangers with this approach are the stresses which are placed on the pipeline during the lifting and lowering operation. The lowering, in particular, must allow for the extra length of the pipeline when it is on the seabed. This is usually done by moving the laybarge gently sideways as the pipe is lowered to the seabed.

352

Main Menu


5.6 REFERENCES

1. PARLANE A J A, STILL JR Pipelines for subsea oil and gas transmission Materials Science & Technology 1988 314-323. 2. S D FARLEY, C SYKES, P SIMMONS Large diameter, deep water pipelay th Zeepipe 2A project OMAE 1996 15 International Conference on Offshore Mechanics and Arctic Engineering. 3. ANON Saipem gets ready for ultra-deepwater pipelay Offshore October 1998 58 4. R HARRISON, G SHARLAND, G SHARMAN, G CURTI, G MASTRACCHIO, H DALE Troll oil pipeline : pipelaying at the limits of today’s technology ?????? 5. K PROSSER Alternative Welding Systems for Pipelines IIW XI-E Document 8/87. 6. J HAMMOND A Review of ‘One Shot’ Welding Processes Int. Conf. ‘Joining and Welding for the Oil and Gas Industry’ London October 1997. 7. D RUSSELL Development of Laser Technology for Welding in the Oil and Gas Industries Int. Conf. ‘Joining and Welding for the Oil and Gas Industry’ London October 1997. 8. C PUNSHON, A BELLONI Reduced Pressure EB Welding for Offshore Pipelines Int. Conf. ‘Joining and Welding for the Oil and Gas Industry’ London October 1997. 9. P KVAALE, B LIAN, A KNOKLEBYE, T SIMONSEN Flash butt welding of TMCP steels International Conference ‘The Metallurgy, Welding and Qualification of Microalloyed (HSLA) Steel Weldments’. 10. S B DUNKERTON, A JOHANSEN, S FRICH , Radial Friction Welding for Offshore Pipelines. International Conference ‘Welding & Performance of Pipelines’. London Nov 1986. 11. P H MOE SAG Forge Welding . International Conference ‘Welding & Performance of Pipelines’. London Nov 1986. 12. R W CARNES Advances in Homopolar Welding of API Linepipe for Deepwater Applications. Int. Conf. ‘Joining and Welding for the Oil and Gas Industry’ London October 1997. 13. A ANDERDAHL, B PERSSON & G LANDE An Explosion Welding Method for Joining Pipelines International Conference ‘Welding & Performance of Pipelines’. London Nov 1986. 14. D G BATHAM, A LOYER & D V DORLING. High Impact Welding for Intermediate Diameter Pipeline Construction. International Conference ‘Welding & Performance of Pipelines’. London Nov 1986. 15. G HUTT Radial Friction Welding for Offshore Pipelines Duplex World Conference 21-23 October 1997. 16. H HEUSER, E PERTENEDER, J TOSCH Welding of 13%Chromium Supermartensitic Stainless Steels Pipeline 98 NEC Birmingham 19-21May 1998. 17. M SWIDZINSKI, G HUTT, C FOWLER. Mechanical and Corrosion Assessment of Advanced 13%Cr Steel Welds produced by Radial Friction Welding. Supermartensitic Stainless Steels 99, Belgium May 1999. 18. N SAUTE, M SAUVAGE Fabrication and Installation of Duplex stainless steel Sealine Offshore Nov. 1997. 19. P BUTLER, J G EMMERSON Welding the Maui A-B Pipeline Pipeline Digest May 1994, 7-8.

353

228606791 Offshore Pipeline Engineering Materials Welding Module Construction Practice

Recommend Documents