Welding of titanium alloy risers (May 2000)

Graham Hutt, Stolt Offshore Limited

Originally published in Welding and Metal Fabrication, 2000, Vol. 68, No. 5,
May, pp 9, 10 and 12 by DMG World Media
http://www.dmgworldmedia.com/

Graham Hutt is Group Welding Manager for Stolt Offshore, based in Aberdeen, Scotland. He is responsible for providing welding and materials engineering support for the Company's offshore construction activities related to pipeline and riser installation and underwater welding. Prior to joining Stolt Offshore he spent five years at TWI in Cambridge, engaged in arc welding research.

E-mail: ghutt@stoltcomexseaway.com

This work was conducted jointly by Stolt Offshore Limited and GKSS Forschungszentrum with support from the European Commission THERMIE-JOULE Programme (Contract OG/175/95).

With offshore recovery of oil and gas deposits moving into deeper waters and the pressure, temperature and corrosion conditions experienced becoming harsher, interest in the use of titanium alloys has grown. Graham Huttprovides the results of one research project, the joining processes best suited and describes the equipment developed for laybarge operation.

Floating production systems have opened the way for low cost recovery of offshore oil and gas reserves, especially for deepwater locations. Riser pipes are required for transportation of the product from the sub-sea wells to the production vessel and these are subjected to very severe operational and environmental loadings. The configuration of the riser is selected to accommodate the motion of the vessel and take several forms varying from a simple, free hanging catenary of pipe to more complex wave shapes developed by adding ballast and buoyancy in selected areas. A typical sub-sea development may require several such risers ranging in size from 6 to 14 inch diameter with individual lengths ranging between 500m and 1500m. Traditionally, these are manufactured from flexible pipe which is highly suitable for a large number of situations. However, as field developments move into ever-deeper waters and involve the handling of high pressure, high temperature and highly corrosive products, there is a concern that currently available flexible products will become a limiting factor.

Interest has, therefore, turned to titanium alloys and the possibility of exploiting the combined properties of high strength, low weight, low elastic modulus and exceptional fatigue and corrosion resistance.

A project to develop and demonstrate the necessary technologies for the design, fabrication and installation of such systems has recently been completed. Major aspects of this work were the selection of a suitable alloy and the development of high productivity welding methods suitable for offshore fabrication and installation from a pipe lay ship.

Selecting the material

The available titanium alloys were assessed against the performance criteria with the objective of selecting candidate material for the main test programme.

Within the family of Alpha-Beta alloys, Ti-6Al-4V is the most readily commercially available. It has been used in a variety of industries for many years and possesses all of the required properties for the riser application. At the outset of the project, only limited information was available on the behaviour of this alloy in seawater environments. It was therefore necessary to carry out an extensive validation exercise using the two nominated welding processes i.e. Orbital Gas Tungsten Arc Welding (GTAW) and Radial Friction Welding (RFW). Butt welds were produced using a Grade 29-alloy pipe and these were submitted for detailed mechanical testing including S-N fatigue tests in seawater at elevated temperature.

Welding processes

Titanium alloys have been exploited in a wide variety of industries, including offshore applications, but there are few examples of on-site fabrication on the scale required for fabrication of deepwater risers. Each riser may incorporate several hundred welded joints and, therefore, the selected welding processes must be able to deliver acceptable productivity but without compromising the high quality needed for this application.

GTAW cold wire

Gas Tungsten Arc Welding (GTAW) is traditionally used for titanium and produces high quality welds. Joint completion rates are, however, relatively slow and to optimise productivity it is necessary to employ multi-station welding techniques and narrow gap joint configurations. With the exception of welding sub-assemblies (double jointing), there is no opportunity for rotating the pipes and consequently it was necessary to develop mechanised, orbital GTAW procedures for the pipe girth welds. Extensive work was performed to validate orbital procedures in terms of mechanical, corrosion and fatigue performance and to establish this welding process as the benchmark for production quality. Fig.1.

GTAW hot wire

The hot-wire variant of the GTAW process is an attractive proposition for titanium because it retains all the recognised qualities of cold wire GTAW however, with much enhanced deposition rates.

The North Sea's first titanium drilling riser for Conoco's Heidrun field has provided an important reference for the design and fabrication practices to be applied to the smaller diameter, dynamic, production risers.

The drilling riser system comprises 24m long sections of seamless pipe, 24-inch nominal diameter, with wall thickness ranging from 22mm to 33mm. An extensive development program was undertaken to evaluate and optimise joint design, welding parameters, and shielding methods using automatic hot wire welding. The joints were made from one side only, and a narrow-groove joint design and hot-wire filler metal additions were used. This practice resulted in substantially fewer weld passes and increased deposition rates.

For the drilling riser, the relatively short section lengths allowed rotation of the pipe and welding in the AWS 1G position. In contrast to this, fully orbital techniques are required for the much greater length of production risers.

Orbital GTAW equipment with hot wire addition has been applied to titanium pipe welds with very promising results. The particular characteristics of the titanium weld pool are favourable in terms of supporting the higher deposition available for the hot wire addition. Demonstrations have shown that the number of weld passes can be halved when compared to cold wire GTAW with a 14mm wall pipe requiring typically four weld passes.

Radial friction welding

Radial Friction Welding (RFW) involves using a vee-shaped consumable ring that is rotated and radially compressed between two firmly fixed pipe ends, Fig.2. A solid phase weld is formed in typically 12 to 25 seconds. Full penetration of the joint is facilitated by a recessed support mandrel. Both the root side and the cap are machined in subsequent operations to produce uniform profiles.

RFW offers a number of advantages over fusion processes and in the context of newer materials these include three of particular importance.

• Suitability for materials which are difficult or expensive to weld by conventional means.
• Highly reproducible quality and properties.
• Consistently high production rates independent of material type and pipe size.

The welding process is targeted at pipe diameters of 12in and smaller and therefore an important segment of the market for specialised flowline and riser materials. Much of the development work has been concentrated on 6in diameter but more recently a production RFW machine suitable for 6 to 12in pipe has been commissioned. The new equipment, Fig.3, is shown during proving trials at the Stolt Offshore facility in Halifax, England. The intention is to install the equipment on a lay ship for flowline and riser installation.

The mechanical properties of titanium alloys are strongly dependent on their microstructure and, therefore, a precise understanding of the RFW weld microstructure was required. Detailed mechanical testing included measurement of tensile, fracture and fatigue properties.

Welds were produced using a matching composition filler ring, Fig.4. The tensile testing results confirmed that the weld metal over matched the parent material and HAZ.

Fracture toughness was assessed using Single Edge Notch Bend (SENB) specimens.

Specimens were tested in the 'as welded' and the stress relieved condition (2hrs at 720°C). For both conditions, the results for K _EE and CTOD δ _{5 max} as well as the R-curves indicate a reduced fracture toughness in the HAZ and weld metal when compared to base material.

In order to improve the fracture toughness, Post Weld Heat Treatment (PWHT) was investigated using temperatures near the super β transus. The experimental work was divided into two stages. In the first stage a series of different PWHT were devised and conducted in an electric heated furnace in an air atmosphere. The resulting microstructures were investigated by conventional light microscopy after removal of the surface oxide layer. In stage two, the specimens treated by the selected procedures were subjected to hardness and micro flat tensile testing. In addition, specimens were produced for fracture toughness tests (CTOD δ ₅ ).

The results showed that all heat treatments at 950°C improve the fracture toughness of the weld metal compared to the 'as welded' condition without any significant degradation of the base material with the most suitable condition being of 15 minutes duration with a cooling rate, 150°C/min.

Fatigue tests in synthetic, circulated seawater (ASTM D 1141) were conducted at room temperature as well as at 150°C. The tests were carried out on servo-hydraulic testing machines under load control with a test frequency of 0.2 Hz (sinusoidal wave form, R=0.4) representing wave loading. Flat specimens were used for the tests and the surface condition was as machined to represent a surface roughness comparable to machined pipe weld sections. The specimens were oriented parallel to the axial pipe direction with the RFW joints located in the centre of the gauge length.

The results of the fatigue tests are presented in Fig.5 which also includes the results of the study on GTAW welds. All RFW specimens exhibited a longer fatigue life than the selected Riser Design Curve. The number of cycles to failure of the RFW specimens were comparable or, even in the upper stress regime ( σ _max = 680 MPa), higher than the GTAW welds.

The test results indicate that the test temperature (20°C/150°C) has no significant influence on the fatigue strength of the joints. The fatigue life of the PWHT specimens (2h at 720°C) was not improved when compared to the 'as welded' condition.

Conclusions

This programme of work is now complete and the results indicate that the metallurgical and mechanical properties of GTAW and RFW joints in Grade 29 alloy fulfil the main requirement for a future use in the installation of titanium production risers.

There is considerable scope for the use of titanium for production risers for deepwater field developments that involve the handling of high pressure, high temperature and highly corrosive products.

The use of orbital, hot wire GTAW provides much improved productivity while retaining the high quality required for this critical application.

For future commercial application, it is considered that the use of lay-barge installation techniques combined with the RFW joining process will provide the best technical and economic solution.

Copyright 2000, DMG World Media UK Ltd