LIFE MANAGEMENT OF PSA VESSELS – Part 2: Inspection and Remaining Life
In the first part of our blog series on the Life Management of Pressure Swing Absorbers (PSA), we discussed the importance of accurately measuring weld peaking and how it influences stress and fatigue life. As we delve into the second part of this series, our focus shifts to the quality of Non-Destructive Testing (NDT) inspections, a critical element in ensuring the safety and longevity of PSA vessels.
Detectable Defect Sizes Dependent on Inspection Technique
PSA vessels are subject to cyclic loading, which makes them particularly vulnerable to fatigue-related issues, especially in areas with welding anomalies like peaking and misalignment. As the vessels approach or exceed their design life, the probability of crack initiation and growth increases. This is where NDT inspections play an important role.
High-quality NDT inspections are essential for identifying cracking at an early stage, particularly in critical areas such as longitudinal seam welds. The quality, repeatability and accuracy of these inspections has a direct impact on the outcome of fatigue life and consequently, the re-inspection intervals. It is important to emphasise that for PSA vessels, opportunities to conduct internal inspections are limited because the adsorbent needs to be removed resulting in lengthy downtimes and significant cost. This means that operators must rely on ultrasonic inspections from the OD (outer diameter) to detect cracks on the ID (inner diameter). Where, the seam weld on the ID is the most likely location for crack initiation due to the hydrogen environment and the high stresses that may be generated by bending in the longitudinal weld seam due to peaking.
Table 1 shows potential detectable flaw sizes based on the used inspection techniques. However, this table comes with a caveat; it is important that the NDT technicians carrying out the inspection can show they are able to detect the defect sizes that will be used in the fatigue assessment. Therefore, performance demonstration on mock-ups is important to gain confident in the detection capability. In addition, the ultrasonic instrument combined the transducer / wedge(s) to be applied to the PSA vessel during examination must be calibrated using industry standard reference blocks.
Table 1: Initial assumed defect sizes depending on the inspection technique applied
| Initial Crack Size | Notes | |
| Height, mm | Length, mm | |
| 0.5 | 3.2 | WFMT applied with good practices per Becht’s inspection expert judgement. |
| 0.5 | 10 | This reference flaw has been used by some operators to qualify technicians to a procedure with known EDM notches.
Considered a good practice advanced UT flaw detection limit. |
| 1.5 | 10 | BS7910 Annex T largest surface breaking defect that will be missed on the back surface during an inspection by focused phased array.
Considered an average practice advanced UT flaw detection limit. |
| 2 | 15 | BS7910 Annex T largest surface breaking defect that will be missed on an as-welded surface with local dressing using MPI inspection.
Also considered a poor practice advanced UT flaw detection limit. |
Case Study: The Importance of Flaw Detection in PSA Vessels
One of the significant insights from a fatigue crack growth analysis is the correlation between the detection capabilities and the calculated remaining life, as shown in Figure 1 below. It should be noted the fatigue crack growth shown in Figure 1 represents a PSA vessel that had weld peaking of 8 mm as well as weld misalignment of 1.5 mm, resulting in a very severe bending stress across the long seam weld; see Part 1 of this blog..
The bending stress generated from the weld peaking and the misalignment were based on laser scanning shown in Table 2 with the Rb ratio calculated from a finite element analysis.
Table 2 Stresses obtained from the finite element analysis based on laser scanning
| Method of Measuring Weld Geometry Anomalies | Rb Ratio |
Hoop Stress due to Internal Pressure, MPa |
Additional Bending Stress due to Weld Anomalies, MPa |
| Laser scanning | 1.23 | 94 | 116 |
The fatigue crack growth rate per cycle, da/dN, is calculated using the Paris Equation, as follows:
where C and m are material parameters, and ΔK is the stress intensity factor range over the cycle calculated based on the membrane and bending stress.
Figure 1: Crack growth as a function of time in service for different initial crack sizes
The fatigue crack growth rates employed for the calculation shown in Figure 1 are based on an internal literature search of published fatigue crack growth rates for different partial hydrogen pressures, microstructures (base, weld and HAZ) and different pressure frequencies.
In the Figure 1 example where the PSA vessel had a significant amount of peaking and weld misalignment, it was found to have a critical remaining life of just nine months based on the detection of a flaw with dimensions of 1.5 x 10 mm. However, if it can be proven the NDT technicians can detect smaller flaws, say for example 0.5 x 10 mm, then the time to reach the critical crack size is more than doubled. It is important to note that using small initial crack sizes of 0.5 mm in height should only be applied in case where the NDT technicians have demonstrated that they are able to detect defects that that size.
Conclusion
In conclusion, the quality of NDT crack inspections is a critical factor that can significantly influence the life management and integrity of PSA vessels. As these PSA vessels age, operators must ensure the most advanced and reliable inspection techniques are employed as well as understanding the fatigue life consumption of each vessel. Doing so will allow for a more comprehensive life management plan that will help avoid failure and unscheduled downtime.
- In the next part of this series, we will explore the strategies for developing a life management plan for a fleet of PSA vessels based on fatigue assessment and the development of effective inspection criteria.