Hydrogen-assisted cracking in welds, which is also referred to as ‘hydrogen cracking’ or ‘delayed cracking,’ often requires time to occur. The reason for this is that time is required for the hydrogen to diffuse to areas with crack susceptible microstructures. Prior to inspection for hydrogen cracking, general good practice indicates that a sufficient delay time should be allowed to elapse — to allow any cracks that are going to form to do so and for the cracks to grow to a detectable size. What is a ‘sufficient’ delay time? Why does a delay time tend to be required for some applications (e.g., installation of a hot tap branch connection) and not for others (e.g., construction of an offshore pipeline from a lay barge)? This paper will address these and other related questions and present the results of recent experimental work on this subject.
When determining appropriate delay times prior to inspection, it is important to consider not only the time-dependent nature of hydrogen cracking, but also the expected susceptibility of the weld to cracking. From a time-dependent nature standpoint, longer delay times decrease the chance that cracking can occur after inspection has been completed. From a probability standpoint, if measures can be taken to assure that the probability of cracking is extremely low, then determining an appropriate delay time becomes a moot point. In other words, if the weld is never going to crack, it does not matter when you inspect it. The probability of cracking can be minimized by using more conservative welding procedures (i.e., by designing out the risk of hydrogen cracking during procedure qualification). For example, if hydrogen levels are closely controlled by using low-hydrogen electrodes or a low-hydrogen welding process, or if the hydrogen in a weld made using cellulosic-coated electrodes is allowed to diffuse away after welding by careful application of preheating and slow cooling, or the use of post-weld preheat maintenance (i.e., post-heating), the probability of cracking is significantly reduced, and immediate inspection may be justified. This alternative approach to time delay prior to inspection for hydrogen cracking, which can allow for immediate inspection, will be presented.