Light weight composite materials fortify Mobil platform blast walls

A firefighter wearing the new Draeger oxygen breathing unit.

The new Gent Orator CSR voice alarm system.

Probably the first repair application for CFRP composites were aluminum ships owned by Britain's Royal Navy. Techniques were developed by the Admiralty Research Establishment (now the Defense Research Agency) which were subsequently adapted by the Royal Australian Navy to repair steel ships.

Ten years later, the idea was dusted off when Mobil was examining ways of blast strengthening the Beryl B platform. Mobil knew of studies performed by Devonport Managemnt (DML) at the Devonport Royal Dockyard that had demonstrated the viability of the technique.

An engineering study by WS Atkins concluded that CFRP could be applied on Beryl B, while a cost analysis revealed that, despite the high materials unit cost, the technique would in fact be the most economical way of achieving the desired strengthening level. Reasons for this were:

• The low weight of the materials and process equipment removed the need for mechanical handling equipment.
• The compact material packages could be taken into areas where access was tight without dismantling existing equipment or temporarily removing parts of the existing structure
• The high mechanical properties of the materials meant they could be applied in small amounts to strengthen only critical parts of the structure, allowing a structurally efficient solution to be developed.
• There was no need to shut down platform operations as the installation process did not require hot work.

The design and validation program was split into five phases: development of a robust application process; characterization of materials properties; structural design and analysis; manufacture and testing of prototypes; design review.

Implementation would be based on DML's Resin Infusion under Flexible Tooling process that had been developed to allow high quality composites to be formed in-situ and bonded to the existing structure. Key features of this process are:

• The dry fibers are pre-formed in a workshop and the necessary process materials attached to the pre-form before packaging. All that is required at site is to unwrap the pre-form, attach it to the existing structure and connect the resin supply. As the resin flows into the dry fiber pre-form, it develops both the composite material and the adhesive bond between the composite material and the existing structure
• Dry fiber pre-forms are made to near net shape, i.e. close to final dimensions. Therefore it is possible to strengthen or repair easily three dimensional structures such as tee joints.
• The dry fiber pre-form is very flexible, allowing it to conform well to uneven surfaces. The thin, even gluelines that result are thought to contribute to the high measured strength of the joints, of the order of 28MPa lap shear strength.
• The process provides high fiber volume fraction composites (above 55%) that have very good strength and stiffness. As hot working would not be allowed offshore, suitable substitutes for any electrically driven equipment had to be found. Also, it would not be possible to heat the working areas. Resins had to be identified, therefore, that could be used at temperatures as low as 5!C, and the detailed set-up of the process had to be modified to ensure a good quality composite material and adhesive bond would be formed at this temperature.

Results of the preliminary engineering study and process trials indicated that the best solution would be based on three types of carbon fiber with Young's modu* between 230 and 550 Gpa, with laminate thicknesses of 12-30mm. The high modulus fibers were used to provide a composite material with a Young's modulus very close to that of steel, while the standard modulus fibers were used to provide strength in secondary directions at reasonable cost. Test laminates were made and a wide range of tests performed, including:

• Full characterization of short-term mechanical properties for each laminate.
• Investigation of the effects of temperature and strain rate on laminate properties.
• Investigation of the effects of surface preparation on bond strength.
• Investigation of the effects of strain rate and temperature on the adhesive bond strength.

To validate the design, half-scale models of the strengthened blast wall were built and blast tested. In all, 14 blast tests were carried out. Good correlation between blast test results and the predictions of the FEA model were seen and the mode of failure was as expected. Following a final design review, Mobil decided to implement the CFRP solution on Beryl B.

The final design required 13 columns to be strengthened to achieve an almost tripled increase in blast capacity on two walls of size 40 x 8 metres. In some cases, the space between the flange faces of the columns to be strengthened and the adjacent structure was only 25mm, while in most of the areas, access was very congested. The benefits of using the light weight, compact and flexible materials quickly became apparent.

The system was installed by a small team in three, two-week visits (four visits had been allowed for in the schedule). Considering that this was a development project, it was considered a major achievement that every stage of design, validation and implementation was completed within schedule and to budget.

Copyright 1997 Oil & Gas Journal. All Rights Reserved.