The Structural Repair of a North Sea Platform Using Underwater Wet Welding Techniques

Abstract

ABSTRACT Underwater wet welding repair techniques using ferritic electrodes have, until recently, been limited to low strength steels with carbon equivalents below 0.40% because the fast quench that occurs after welding produces hard heat affected zones susceptible to hydrogen cracking. This paper describes the first documented structural wet weld repair of a North Sea offshore platform constructed with steel that had a carbon equivalent greater than 0.40%. It describes the events that lead to the failure and documents the repair using underwater wet welding techniques. BACKGROUND The underwater inspection of an Amoco U.K. North Sea Platform in June 1987 revealed that a diagonal brace connecting Leg B3, at a point 26 feet above water to Leg E3, at a point 36 feet below water (Figure 1) had a severed weld which had connected the brace to the E3 leg node (Figure 2). The failure was the result of ship impact on the brace. Initial structural assessment concluded that the damage did not affect the integrity of the platform and no emergency measures were required. However, in order to prevent further degradation of the damaged member from wave action, the brace was removed in July 1987. A failure analysis supported by a metallurgical examination of the fracture surface was undertaken by Amoco Corporation Research at Naperville, Illinois. The findings indicated that the failure was the result of the impact loads which occurred when the ship struck the brace. The impact resulted in a circumferential weld metal crack between brace 3Bl and the jacket leg stub JL3E which was located underwater. Multiple fractures initiated at lack of penetration weld defects on the ID of the brace weldment with several fracture initiation sites in the first pass of the weld. Figure 3 shows a cross section of the weld failure which contained a lack of penetration defect, forming a crack like defect in the ID of the weld. The cracking initially propagated in a brittle manner and changed to a ductile shear fracture as it approached the weld cap. There was no evidence of fatigue cracking on the fracture surfaces examined. The impact loads from the ship did not initiate cracking in the brace-to-stub weldment located +26 feet above water. Figures 4 and 5 show cross sections of the +26 foot brace-to-stub weldment. Note that the excessive root pass reinforcement in Figure 4 and the suck back in Figure 5 are defects with smooth contours which would be less likely to propagate as a result of impact. A comparison of the above water and the underwater weldments was also made to determine if there were any significant differences in the chemical composition or mechanical properties between the -36 ft and the +26 ft welds. The testing revealed that both welds had similar weld metal compositions suggesting that they had been welded with the same type of electrodes. The Charpy impact toughness of both weld metals were found to be satisfactory for this service.