Acoustic Emission Monitoring for HIC-Affected Vessels

The following question was posed by a member of the Inspectioneering Community:

I need advice concerning the following problem: we are operating a gas scrubber at 600 PSIG. The metal thickness is 32 mm (1.28"). The gas stream contains 5,000 ppm of hydrogen sulfide. The material is SA 212 grade B. The H₂S level has gone up only in the last two years. On ultrasonic scanning of the vessel, HIC is suspected.

The vessel was stress relieved after fabrication. In the last ten years the vessel had an internal coating installed, coal tar epoxy due to internal corrosion. The operating temperature is around 150 degrees Fahrenheit. The latest WFMPI does not reveal any internal surfacing of the cracks. UT scanning was done from the outside and internal scanning is not possible due to surface roughness. We would like to operate the vessel until we arrange a replacement. Would you advocate acoustic emission (AE) continuous monitoring of the vessel for the time being? AE monitoring during an immediate hydrotest and subsequent continuous monitoring is what I have in mind. Is AE monitoring done for HIC-affected vessels?

First, I'd like to mention that we've published many articles related to acoustic emission testing, damage mechanisms such as HIC, and other related subjects. I'd like to share my own thoughts on this topic.

AE monitoring has been done for HIC affected vessels, with limited success. The limitation, which greatly affects the confidence level of results, is that traditional AE testing stresses, i.e. Kaiser affect overpressure and felicity effect types of re-stressing, do not generate the types of stress necessary to generate elastic strain waves at the HIC laminar crack tips. Some laboratory work has been done in an MPC joint industry project where an AE testing company was able to detect this type of damage with no stress increase. Keep in mind though that the vessel was in a laboratory (very quiet) and that the H₂S solution was very harsh. Not like what you would see in the field. If the damage is stress oriented, traditional over pressure stress testing is more likely to detect it and this may be the main damage you are interested in, anyway.

With the information (grade of material, H₂S content, temperature, service, etc. I assume water is present) you supplied it is quite possible that hydrogen may be gathering at laminations or non-metallic inclusions in the base material. As long as any potential HIC remains normal to hoop stresses, i.e. laminar only, does not usually detract from the load bearing capabilities of the material, as long as it is a safe distance from the welds and HAZs, see the API/ASME 579 FFS Standard for rules. It is typically when HIC becomes stress-oriented that it raises a concern that could lead to leaks or failure.

You did not mention the age of the vessel other than the coating being 10 years old. This is usually pushing the generally accepted age maximum of coatings (10 to 15 years in non-harsh environments). How did you perform WFMPI on the interior of the vessel? I assume the coating was removed in selective locations and then repaired. Was there a primer used? and what type? How has the coating held up?

You may want to consider areas on the outside of the vessel opposite areas of coating loss as well as the areas UT has already indicated with automated ultrasonic (AUT) scanning in the 0 degree compression wave mode. This should be able to detect and map out the HIC (blister activity) damage. You can then monitor it for growth rates at whatever frequency you desire. Some have reported limited success using combined shear wave angle beam AUT to detect any stress oriented cracking emanating from edges of blisters. Some like to perform the 0º  scanning first, to see if there are any blistered areas, then do the shear wave scanning for crack detection. Some do both at the same time, although this can add unnecessary time to the scanning and data analysis. Your 150ºF temperature will allow this type of scanning. It does require a high level of expertise to characterize the cracking that can be associated with blistering. Be careful though, blistering may mask or hide the presence of cracks underneath them.

Now we have advanced methods such as TOFD (time of flight diffraction) and phased array UT where progress has been made in detection, characterization and sizing of this type of damage.  Qualification demonstration testing for the application and or careful check of UT references is highly recommended.  Automated Phased Array UT is quite popular now and can be quite effective at crack detection and characterization. I recommend the same attention to qualification demonstration testing for the AUT operator, hardware and software.

Many are using Risk Based Inspection (RBI). An important aspect of RBI is the effectiveness of the planned inspection strategy in finding the damage. That has a big impact on the uncertainty debit or credit charged in the probability of failure calculation. Make sure that you are realistic and consistent about the inspection strategy you create which includes; type of NDE (will it find the type/s of H₂S damage you are looking for and how confidently?), expertise of operator and data interpreter, procedure, severity of inspection conditions (e.g. surface temperature, surface condition, ease of access, etc.) scope of inspection, percentage of susceptible area being inspected, and that you factor their impact on probability of detecting the damage. Often a combination of inspection methods will be used.

Ultimately, for various complex reasons, it is usually very difficult to conclusively tie AE results to fracture mechanics, i.e. how the crack needs to be (size, orientation, material ductility, etc.) in order for AE to detect it. In this particular case, with all the other methods used appropriately, to cover an appropriate percentage of welds and parent plate, it is up to the owner operator to determine what is appropriate in this regard. Finally, with all of this completed, understanding the amount of uncertainty and risk for current and future safe operation of the vessel, your AE action plan might be viable as an on-stream monitoring method. It is important that the AE operator, hardware, stress fields and software be validated as appropriate for the job/scenario.

Any inspection strategy should be viewed in terms of leak and fracture probability, potential consequences in the event of leak or fracture, critical flaw sizes, maximum tolerable flaws sizes, a realistic understanding of damage mechanisms and their dynamics for that specific case, operating practices, time to failure based on anticipated stress fields, codes and standards, industry best practices, and internal mechanical integrity program requirements.

Interested in reading more about inspection, damage mechanisms, and many other asset integrity management topics? Click the button below to download a free copy of John Reynold's updated and revised 101 Essential Elements in a Pressure Equipment Integrity Management Program guide.