An Analysis of a Completed Temper Bead Weld Repair Performed on an Amine Tower

Abstract

An amine tower was inspected and was shown to have wall loss over the majority of the circumference. A repair plan was developed which included welding onto the tower while the tower remained in operation. The two step repair plan first required welding two Inconel 625 rings to the SA516-70N steel tower and then welding a stainless steel sleeve to the Inconel 625 rings encapsulating the corroded area. Since the repair welds could be exposed to an amine environment if the steel tower corroded through, API Recommended Practice 945 (API RP 945) was used to aid in the qualification of the welding procedure. API RP 945 recommends a post-weld heat treatment (PWHT) to reduce hardness and relieve stress, but since the planned repair was to be made in-service, PWHT was not preferred. To address the hardness aspect a temper bead technique was used to successfully qualify a welding procedure, without PWHT, in accordance with 2006 NBIC and 2004 ASME Boiler and Pressure Vessel Code Section IX with hardness values below 200 Brinell. The temper bead welding procedure used Inconel 182 SMAW electrodes and required strict welding heat input control and weld bead placement. The heat input could be monitored by controlling the welding parameters or by using the run-out ratio diagram. The temper bead passes needed to be deposited in such a manner that the weld toe of the temper bead was no more than 3/32 in. (2.4 mm) away from the weld toe of the initial layer. To address the residual stress aspect the procedure qualification weld was thermo-mechanically modeled to predict the residual stress distribution on the inside surface of the amine tower. The repair procedure was performed on the operating vessel to emplace the straps and the sleeve. Shortly afterwards, the tower wall was breached and the internal environment reached the annulus inside the sleeve. Operation continued for several months until the replacement vessel was available. Once the vessel was removed from service, a section of the repaired area was examined for residual stresses and hardness in the carbon steel. The peak residual stresses were lower than predicted by the analysis from the qualification stage. However, the measured heat affected zone (HAZ) hardness was well above the desired level of 200 Brinnell. Analysis showed that the increased hardness level correlated with improper temper bead placement [i.e., temper bead to weld toe spacing greater than 3/32 in. (2.4 mm)] along with other indications of deviations from the qualified procedure.

Modeling of Tensile Residual Stresses Extending More Than 5 Weld Widths From Weld Joint

“…Some results on weld residual stresses have been obtained, such as those by Karlsson and Josefson [5], Leung and Pick [6], Hong et al [7], Dong [8], Zhang et al [9], Dean et al [10], Yang et al [11,12], Mohr et al. [13] where multi-pass welds effects were simulated directly using the commercial code ABAQUS, with user-developed welding heat source models.…”

Simulação numérica e experimental de soldagem em operação de um aço API 5L X80 com processo GMAW.

“…Um dos desafios da técnica passe de reaquecimento é o controle sobre a entrada de calor, Mohr et al [50], não conseguiram alcançar valor de dureza requerido em função do alto valor da entrada de calor, soldaram uma junto de filete e comparação com simulação numérica.…”