H2 Bakeout Interview Rev 1

Hydrogen bakeout, also known as hydrogen outgassing, is inconsistently applied in different companies. It is important to understand the risk of hydrogen-induced cracking in weld repairs. Hydrogen can diffuse into metals and weaken bonds, causing porosity and cracking. Thermal charging, where hydrogen diffuses into steel at high temperatures, and corrosion reaction charging, where hydrogen is produced during corrosion, are the main sources of hydrogen charging. Knowledge of the process stream and contaminants can guide the decision to perform bakeout. Cyanide and arsenic increase the probability of hydrogen charging. Visual inspection and other techniques can determine the need for bakeout. Preheat or post-weld heat treat makes bakeout easier to perform. Weldability tests can also be used. Austenitic stainless steels do not require bakeout. The need for hydrogen bakeout, which is also known as hydrogen outgassing, can be inconsistently applied between companies and even within the same company. This is often dependent on a level of conservatism and a degree of understanding by the materials and corrosion resource being consulted. Hydrogen bakeout questions arise during routine maintenance repairs, attempting to get a lot of attention during turnaround periods, particularly when an unplanned weld repair needs to be performed without negatively impacting the turnaround schedule. Turnaround teams can be faced with assessing the risks of impacting critical paths to set up and perform a hydrogen bakeout prior to weld repair versus the potential for a more difficult and time-consuming repair if the repair welds crack due to the presence of hydrogen diffused into the steel. The requirement for bakeout, generally for ferritic steels, which have been in service, stems from the fact that hydrogen in metals can cause porosity and cracking in the weld deposit. Elemental hydrogen can diffuse into the metals. These hydrogen atoms accumulate at high-stress areas such as grain boundaries and in voids and other inclusions. The presence of hydrogen atoms weakens the bonds at these locations, facilitating crack initiation and propagation. Elemental hydrogen can combine to form molecular hydrogen or hydrogen gas at lower, such as ambient temperatures, and this molecular hydrogen becomes trapped in the steel. The key to deciding whether or not to bakeout before performing a weld repair is recognizing the two most common sources for hydrogen charging into the steel that's been in service. Those sources can be broken down into thermal charging and corrosion reaction charging. Thermal charging is probably the easiest source to understand. Many process streams contain molecular hydrogen, and this molecular hydrogen is constantly in equilibrium with elemental hydrogen. At higher temperatures, generally considered to be greater than 400 degrees Fahrenheit, this equilibrium favors the formation of more elemental hydrogen. This makes the elemental hydrogen more available to diffuse into the steel. The solubility of elemental hydrogen feels also higher at elevated temperatures. Although many process units have hydrogen-containing streams with operating temperatures greater than 400 F, the most common units experiencing thermal charging are hydro-processing units such as hydrotreating and reforming. If hydrogen service equipment such as a hydrotreating reactor is shut down in a planned, meaning a slower manner, where, for example, a hot purge is performed in a hydrogen-free environment, there is generally sufficient time at temperature for most of the diffused hydrogen to be removed from the equipment, even in a heavy-walled reactor. However, if the equipment experiences an emergency shutdown or is otherwise rapidly cooled through the 400 F to 600 F temperature range, then we have to assume that sufficient hydrogen is still charged into the steel, and this makes weld repairs vulnerable to hydrogen-induced cracking unless a hydrogen bake-out is performed. One could argue that thin-walled equipment is not susceptible to cracking in this instance, and indeed it is less susceptible, but practically speaking, repairs in thinner-walled equipment is certainly not immune to cracking without hydrogen bake-out. Hydrogen charging from corrosion reactions is probably the more common problem, although often the least understood. There does not need to be hydrogen present in the process stream for hydrogen charging to occur in a corrosive environment. The corrosion process alone is a source of hydrogen for the steel, even in the absence of molecular hydrogen, H2. For example, in sour water, which is generally considered to be water with a hydrogen sulfide content greater than 50 ppm, carbon and low-alloy steels produce iron sulfide as a corrosion product. Atomic hydrogen is produced at the surface of the steel in these cases, and this atomic hydrogen generally combines to produce molecular hydrogen, but some of the atomic hydrogen diffuses into the steel. In streams without H2S, corrosion reactions involving, for example, water molecules can be broken down by electrochemical reactions at the steel surface, and the corrosion processes can release hydrogen ions, which can then enter the steel. So knowledge of the process stream, and in particular its contaminants, can help guide the decision on whether to perform hydrogen bake-out. For streams containing cyanides or arsenic, the best practice is to always perform a bake-out. This is because both cyanide and arsenic poison the reaction of elemental hydrogen combining to form molecular hydrogen. As a result, the probability of hydrogen charging when cyanide or arsenic are present is far greater than without these poisons. Many companies have adopted the exception to this rule for killed carbon steel piping such as ASTM A106-B that has less than half-inch wall thickness. The thinking behind this exception is that center-wall silicon-killed carbon steel pipe has minimal inclusions to trap hydrogen, and therefore has a very low probability of hydrogen charging, and therefore it's not as susceptible to cracking. In the presence of cyanide and arsenic, some companies will still bake out the center wall as a piping as a precaution. Conversely to the A106-B exemption, some grades of ferritic steels, particularly older grades and or those with less chemistry control, for example, the old A212, A70, firebox quality steels, A283, sometimes A285, these are considered more vulnerable to hydrogen embrittlement because of the higher likelihood of inclusions and voids that can trap hydrogen since they're just generally not very clean. They can be not very clean steels. The presence of arsenic is not uncommon in refining streams as it's present in many crude oils. Most of the unit process or environmental engineers should have knowledge regarding the presence of arsenic based on crude assays and related feedstock data. Arsenic's toxic and catalytic poisoning characteristics tend to make it a known element in both product and in waste streams. For cyanide, some refiners perform sampling, but one should generally expect to find cyanide in thorough water streams associated with light ends processing of cracking units such as FCCs and gas plants and coker units. Risk repairs on these units can be particularly prone to cracking when outgassing is not performed. The choice to forego bakeout on thin-walled A106B piping is a company-specific decision and is sometimes best informed by doing an individual risk assessment for that particular repair. When cyanide and arsenic are not present, a visual inspection of the components to be repaired can be very useful in providing guidance on the need for hydrogen bakeout. When there is significant corrosion observed in ferritic-based materials, wells, and heat-affected zones, it is best to perform hydrogen bakeout because there is a high probability of hydrogen diffusion into the material. Visual and perhaps other techniques such as UT thickness inspections reveal no measurable corrosion, including no localized weld or heat-affected zone corrosion. And for welds, a useful indicator is whether the original fabrication weld pool flow lines are still present, then bakeout can usually be omitted without danger of delayed hydrogen cracking. For cases where some loss is observed, one can generally consider the risk of cracking to be between low and medium. These are cases when some judgment is needed and perhaps can be aided by a risk assessment. If visual or ultrasonic inspections reveal the presence of hydrogen blisters or provide indications that could be microfissures, these findings would, of course, suggest that significant levels of hydrogen charging are present. Care should be used if hydrogen blisters are present because exposure to outgassing temperatures could also drive cracking due to increased pressure and the blisters at temperature. In these cases, measures such as drilling blisters generally from the outside can be employed to alleviate this concern. If the repair will require a preheat or a post-weld heat treat, the choice to bakeout is generally easier because there will already be heat-treating equipment on the job and electric heating pads will already need to be installed. In these cases, it's only the bakeout time that needs to be considered, but most organizations will agree to perform bakeout in these instances. If bakeout is not considered practical, but there is evidence of significant corrosion, a weldability test is sometimes performed. Unfortunately, this test is not usually applied during a turnaround since it requires waiting 24 to 48 hours after welding to assess the results. Basically this test involves welding a bead on the affected material using the same welding process and consumables, so basically using the same WPS, welding procedure specification, as planned for the actual repair. After the test bead is completed, a 24 to 48 hour waiting period begins, after which a PT or preferably an MT inspection is performed. The idea is that if delayed hydrogen cracking does not occur after the waiting period, then bakeout is not needed. We should note that bakeout is not needed for austenitic stainless steels in refining services. The face-centered cubic, or FCC, structure of the austenitics promotes hydrogen diffusion out of the material readily. The ferritic materials that are austenitic stainless steel clad or overlaid are not considered vulnerable to hydrogen cracking from corrosion reaction hydrogen charging, but of course thermal charging of the ferritic base materials in hydrogen service can occur.