DEEPWATER RISERS Steel catenary, flexible risers battle for technical supremacy

Jeremy BackmanEditor - Europe

• Cross-section of a typical Coflexip non-bonded flexible riser pipe.

• Wellstream current and future manufacturing capabilities, collapse water depth (carcass/barrier only, no hoop spiral support).

• C-shaped hoop spirals used in the NKT/Furukawa riser.

Flexible risers are being usurped by metallic systems in some of the newer deepwater developments. Shell chose steel catenary risers for its three latest TLPs in the Gulf of Mexico, and Petrobras has followed suit for gas export from the Marlim semi. In Norway, titanium is making its mark in the riser for the Heidrun TLP.

The downside of metallic systems is the extra expense perceived for many applications. For this reason, flexible risers will remain popular for all water depths. However, there are serious design challenges to surmount as development heads deeper.

In a paper delivered at Deeptec '96, Aberdeen, Svein Are Lotveit of SeaFlex put collapse capacity and weight limitations at the top of his list. He also drew attention to related concerns for deepwater flexible pipe, namely:

• General demand for large bore risers
• High pressure, high temperature and aggressive fluids
• Complicated replacement/repair of deep water risers, which needs to be avoided.

SeaFlex, a Norwegian engineering consultancy, is also part of the technical support team evaluating Coflexip and Wellstream pipes at Shell's facilities in The Hague. The project, known as the Deep Water JIP, is led by four major oil companies.

Coflexip and Wellstream are two of the dominant producers of non-bonded flexible risers (bonded pipes are currently not in the picture for deepwater risers). Another emerging heavyweight is NKT (Furukawa). Lotveit went on to outline this trio's current R & D work.

The market leader for deepwater, large diameter and high pressure applications is Coflexip, which favors a five-layered construction for its pipe. The innermost section is a carcass, providing collapse resistance. This is normally cold formed from stainless steel (316L), although Coflexip is working on alternate, T-shaped carcasses made from aluminum or ferritic stainless steel.

A key thrust of its development has been the middle layer hoop spirals. Main aim has been to increase the pressure rating and to improve service life. Traditionally, the hoop spiral has a Z-shaped cross-section (Zeta). However, a new hoop spiral wire geometry has been developed, called T-wire, which has been used in several test pipes but has so far not been installed offshore.

Advantages of the T-wire geometry are said to be:

• Larger and stronger hoop spirals can be made
• Less susceptible to initiation of fatigue cracks than the Zeta. In accelerated tests, it has been shown to improve service life of Coflexip risers.

The T-wire is currently being tested by the Deep Water JIP. To date, thicknesses up to 14mm have been examined, and wires up to 16 or even 18mm may also shortly be tested. This will increase substantially the pressure rating of large-bore Coflexip risers. A further increase in capacity is achieved through use of a secondary hoop spiral outside the T-wire.

T-wire's main impact will be in extreme pressures and dynamic risers. However, long length production of Coflexip risers with T-shaped wires will likely involve substantial investments, meaning that the cost of these risers will be relatively high.

Tension armors in the company's risers are normally made of rectangular steel wires. To increase the depth range of flexible pipes, Coflexip has qualified tension armors of reduced weight composite material. One composite riser has been installed offshore Brazil and is still in operation.

However, the cost of composite-reinforced systems is higher than a steel-reinforced pipe. They are only justifiable, therefore, at great depths where traditional pipes are too heavy. Maximum depth for a steel-reinforced flexible riser in a free-hanging configuration is typically 1,000 meters. For greater water depths, riser weight is critical and there composite may become attractive again, according to Lotveit.

During 1994-95, a number of failures occurred in risers in high temperature service (above 80C). Both Coflexip and Wellstream responded by re-evaluating their designs, making significant changes. This work is still some way from completion.

Coflexip had considered 20 different designs before selecting a prototype for testing. A 20 meter test pipe and a second sample with the old end fittings were both manufactured and tested by cycling the temperature from ambient to 130C. The new end fitting included a steel sleeve inserted under the main fluid barrier in order to control accurately the diameter and circularity of the barrier close to the main pressure seal.

While the failure rate of the old design in tests confirmed misgivings, the new sample did not perform perfectly either. Coflexip then designed and built five new pipes with different end fittings, incorporating different methods of PVDF layer termination. Since June 1995 when the first set of tests were completed, the second series of pipes have proven these end fittings to be stable.

One customer was sufficiently convinced to accept the first dynamic jumper with a new end fitting design, for short-term service (two to five years) in the Far East. This has been in service since last October. Following further tests, Coflexip is now predicting 20 years of service life.

Wellstream

Wellstream's long-term R&D program for non-bonded flexible pipe focuses on extending the product's capability through higher design pressures with large diameters; higher design temperatures (up to 180C); deeper water; and extremely sour production fluids, up to 5,000ppm H2S. It aims to achieve these targets by 2000, but this will involve improved technology and manufacturing processes.

Current Wellstream can cold form a 1.8mm thick stainless steel strip to a 10mm thick carcass. Tests performed by the Deep Water JIP have shown the collapse resistance of the 10-in. inside diameter structure to be 120 barg, when supported by Wellstream's Flexlok hoop spiral. Allowable water depth with this design is 1,040 meters.

To increase collapse resistance for larger diameter pipes, a new carcass machine is being developed capable of producing up to 16-in. ID: this should be in place by late 1997. Wellstream is also evaluating other profiles and materials such as aluminum alloys which reduce weight and cost while providing greater collapse resistance and sufficient corrosion resistance.

The Flexlok hoop spiral currently comes in three sizes, 4.8, 6.4 and 8mm. However, a prototype 10mm spiral has been produced. Machinery for 10 and 12mm profiles will be in place by the end of next year.

Last year the Flexlok profile was refined using finite element analysis optimization. The design was proven to be successful in a recent dynamic test for a 6-in. riser with 420 barg internal pressure. This riser included for the first time (from Wellstream) a secondary hoop spiral, the Flexpress, for increased pressure capacity.

A thicker Flexpress layer will be applied on the second Deep Water Riser JIP Wellstream pipe, planned to be tested early in 1997. The upgraded Flexlok machine will also be capable of applying a thicker and wider Flexpress layer to increase the pressure capacity of the larger diameter pipes.

Wellstream is also in the process of developing a carbon fiber thermoplastic composite strip to replace the steel tensile armor layer for deepwater sour service and onshore arctic applications. The composite armor is lighter in weight, stronger and more resistant to corrosion than steel. This extends the use of the flexible pipe to dynamic risers for deeper water applications.

Static qualification will likely be achieved by early 1997, with dynamic qualification later in the year and dynamic field demonstration in 1998. The company has also worked closely with Emerson & Cumming to develop a syntactic foam thermal insulation suitable for water depths down to 1,000 meters. This material has been used by Wellstream on Norsk Hydro's Troll project (340 meters water depth). E&C is qualifying similar materials for depths down to 2,000 meters.

NKT

NKT is currently establishing a large manufacturing line for non-bonded flexible pipes in Kalundborg, Denmark. From next year risers will be produced of the Furukawa design that were qualified for dynamic service by Shell in the mid-1980s. Static flowlines will be produced this year.

Production capacity will typically be 70 km of flexible pipe per year, with a maximum of 10-in. ID and 5,000 psi design pressure. Manufacturing is based on turntables with large capacities, allowing up to 10 km of 10-in. pipe to be produced in one length.

Although the NKT/Furukawa riser has been considered promising, the two companies have until recently been reluctant to put sufficient resources into entering the flexible riser market. They have chosen to do so now when demand for flexible pipes is buoyant.

But as Lotveit points out, they also face hard competition. Coflexip has increased its production capacity in France and opened a new factory in Australia. Wellstream is established in the market, and metallic pipes in flexible riser configuration are being viewed as feasible for deep water applications.

Most of NKT's R&D relates to getting production up and running. But there is also a continual dialogue with the oil companies to ensure that NKT's products are in line with industry needs.

One of its first steps needs to be to gain third party verification of the design analysis tools, in line with API Spec 17J. NKT has also been liaising with Shell to develop a composite armored flexible flowline. A prototype may be constructed this year. The aim is reduced weight and use for sour service.

Steel catenaries

Speaking on behalf of steel catenary risers at the same IIR-organized conference, Dr Hugh Howells of 2H Offshore Engineering, Woking, UK, outlined why steel had won the day on Shell's new US TLP developments. Compared with vertically tensioned risers, he pointed out, the steel catenary offers reduced deck space, elimination of the riser base and tensioner and a fixed valve stack: this adds up to a simple arrangement at reduced cost.

Scope for flexibles was also limited by Shell's pressure requirements, he added. The Auger platform's export lines are already close to the maximum diameter/internal pressure combination of flexible risers. Lines larger than 12-in. diameter are beyond current flexible riser capability, which would limit use of flexibles for the Mars Field export lines.

However, steel catenaries wouldn't appear to travel easily to the deep waters west of Shetland, due to the harshness of those operating conditions compared with the Gulf or Brazil. Factors weighing against steel risers off Scotland are larger waves and currents, which would increase extreme loading.

These currents might necessitate use of vortex-induced vibration suppression strakes over longer lengths. Also, the greater average wave height would increase the risk of fatigue damage. Assuming that vessels remain the in-vogue production method west of the Shetlands, Howells prognosticated that drift offsets of up to 25% water depth might be needed in this region for a catenary moored vessel, in order to accommodate intact and damaged mooring systems. This compares with under 10% offset maximum for TLPs in the Gulf, which would mean substantially greater compliancy requirements for the steel riser.

One way of handling these large vessel offsets and dynamic motions, he suggested, might be to use buoyancy arrangements for the steel catenary similar to those applied to flexibles. Loads applied to the vessel would be lower than with flexible risers, as these are 20-40% heavier than steel equivalents in production mode.

Installation of steel catenaries and attached flowlines could be performed using conventional laying methods such as S-lay, J-lay and reel lay. To cut costs, risers and flowlines could be installed in one operation, eliminating the need for a subsea tie-in between riser and flowline.

Vessel tie-in procedures would be the most difficult aspect of steel catenary installation off the Shetlands, said Howells, so greater care would be needed to control curvature. But the extra controls required would not impact cost significantly. In fact, weight of flexible risers may be 40-80% heavier than steel lines during installation, assuming lines are filled.

Howells concluded that steel catenaries could prove a money-saver for TLPs in this environment - perhaps up to 100,000 per riser - through lesser requirements for the riser base and flowline tie-ins.

REFERENCE:"Recent Developments in Flexible Risers for Deep Water; " "Analyzing the Practicalities of Moving to Steel Catenary Risers in the Atlantic Frontier;" Deeptec '96, organized by IIR in Aberdeen.

Copyright 1996 Offshore. All Rights Reserved.