Stress Corrosion Cracking in Steam Turbine: Two Case Studies

Abstract

Stress corrosion cracking in steam turbines had been an old problem though some modern steam turbines have almost eliminated this problem by several methods. The methods include design modification to reduce the stress levels below the threshold stress level for stress corrosion cracking, inducing compressive stress by different means and using pure steam [1, 2]. Some of the older steam turbine discs are prone to stress corrosion cracking. Two cases where such machines experienced stress corrosion cracking in their discs are discussed here. The row 6 disc of an integral steam turbine rotor developed cracks in the root sections. Some of the cracks were mechanically opened for the evaluation. Evaluation of the fracture surfaces with a scanning electron microscope showed evidence of intergranular mode of cracking. Optical microscopy of a cracked root confirmed intergranular mode of cracking. In addition, it showed branching of cracks. Based on these findings, it was concluded that stress corrosion cracking was the reason for the cracks. In addition, finite element analysis was used to calculate the stress distribution in the blade root of the disc. The location of the maximum equivalent stress coincided perfectly with that of the actual crack location in the disc root section. Unfortunately, redesign of the root geometry to minimize the local stress concentration is very difficult due to the size limitation of the blade roots. Small amount of chlorine was identified on the fracture surface and the chlorine could have come from the steam used. The customer was advised to analyze their steam quality and to improve the quality of the steam if needed. The cracked portion was removed from the disc and weld-build up to machine new root sections with the same type of roots. Root section of the row 6 disc of another steam turbine developed failure. This disc had radial entry type blades. Portion of the disc root and some blades were liberated from the disc due to the cracking. The fracture surface had heavy oxide layer on it. Evaluation of the fracture surface with a scanning electron microscope revealed intergranular mode of failure. Energy dispersive spectroscopy analysis of the fracture surface found oxides on the fracture surface. Optical microscopy showed secondary cracking and branched cracking. All these evidences confirmed that the failure occurred due to stress corrosion cracking. In addition, it was suspected that forging was not heat treated properly due to measured lower toughness and different microstructure. The lower toughness was believed to be a result of improper heat treatment rather than that of embrittlement. Methods to mitigate the risk of stress corrosion cracking were proposed.