Technical Basis for Revision of API 574 Piping Structural Minimal Thickness Values

?

In March 2020, Marsh JLT Specialty published a study of the 100 largest industry failures (each failure must be > $175 million in loss) from 1972 to 2019 [1]. Over that 47-year stretch, the study shows an average loss of $432 million per failure, or $43.2 billion in total loss. However, this average increased dramatically when reviewing the most recent industry failures from 2018 to 2019. Those eight failures had an average loss of $562 million per failure, or $4.5 billion in total loss. Further, in the full time period, 39% of the 100 largest losses occurred in refineries, 26% in petrochemical facilities, and 24% in upstream facilities. When reviewing the 2018 to 2019 time period, refineries accounted for 50% of the asset losses, with petrochemical facilities at 25% and gas processing at 25%. Meanwhile, upstream facilities accounted for 0% of the major failures noted from 2018 to 2019 (see?Table 1). Refining represents the largest portion of the industry failures, and when investigated further the biggest culprit is associated with piping, particularly thinning of piping (the other common causes were start-up/shutdown incidents, natural disasters, or not disclosed).

Table 1. Distribution of 100 Largest Losses in Hydrocarbon History

To manage piping systems for safe and reliable operation, a robust mechanical integrity program will include a focus on piping inspection and provide guidance on acceptable limits for evaluating the piping thickness readings being obtained. Establishing a required minimum thickness (i.e., structural minimum thickness) is a key element for this evaluation. Unlike vessels, the piping structural minimum thickness (Tstruct) is not solely governed by hoop stress (?; where P=internal pressure, r=outside radius, and t=pipe thickness) due to pressure only, but can be governed by longitudinal stress (?; where F=axial force, A=pipe cross-section area, M=bending moment, and S=pipe section modulus) resulting from pressure and bending moment due to weight and/or thermal loads (see plot of stress orientation in?Figure 1). To determine if the required thickness is governed by longitudinal stress, the system would need to be assessed for sustained loads (weight plus pressure), and assessed for thermal loads, if appropriate.

Figure 1. Tensile moment.

Existing API RP 574 T struct?Values

It is impractical to assess every piping system using stress analysis; thus, the American Petroleum Institute (API) published Tstruct?values for carbon and low-alloy steel pipe in the 3rd edition of API RP 574,?Inspection Practices for Piping System Components, in 2009 [2]. Table 7 in the current 4th edition of API RP 574 lists minimum structural thickness and minimum alert thickness for pipe sizes from NPS ? to NPS 24 (see?Table 2) [2]. Users may manage their piping systems to a minimum alternate thickness with values greater than either the minimum structural thickness or the pressure design thickness, whichever governs the Tstruct?value. Note, API RP 574 is API’s companion document to API 570,?Piping Inspection Code, and provides the API 570 adherent best practices guidance for managing an effective piping reliability and integrity program.

Table 2. API RP 574 minimum thicknesses.

In recent years, users and members of API RP 574 have tried to determine the basis for the Table 7 Tstruct?values. Such efforts were unsuccessful because the knowledge of the basis for Table 7 values was lost with API RP 574 membership changeover. As a result, The Equity Engineering Group was contracted by API to develop a referenceable basis for the structural Tstruct?values published in API RP 574. The work also included developing new Tstruct?values for low alloy carbon steel and stainless steel. This article will outline the five design considerations incorporated in the work and provide general insight into the results. Final Tstruct?results are not provided in this article as the API RP 574 Task Group is still in the process of balloting and approving the developed results. Once the API RP 574 ballot has passed, a second article will present the Tstruct?results and detail how they should be utilized in a piping mechanical integrity program.

Design Basis – Five Design Considerations

The process for developing Tstruct?values for piping systems can be complicated, as performing a piping flexibility analysis for the countless number of piping layout configurations is an extensive and impractical effort. For the Tstruct?assessments, the analysis was simplified to a 5-span simply supported beam with a concentrated load (P) on the middle span and a distributed load (w) over the 5-spans are assumed (see?Figure 2). The following five design considerations, as shown in?Table 3, were evaluated in this study and are discussed in further detail below.?

Figure 2. 5-span simply supported beam.Table 3. Five design considerations.

Tensile Moment

For a simplified 5-span beam piping system, the distributed load and any potential concentrated loads must be determined. For these calculations, it’s assumed that no axial load is present, but a concentrated load is assumed for most cases that reduces the longitudinal stress to SL?= Pr/2t + M/z. Note that M is the maximum bending moment in the 5-span simplified beam due to both the distributed and concentrated loads. The longitudinal stress is limited to the B31.3 hot allowable stress.

Compressive Stress

Compressive stresses occur at the midspan on the top of the pipe. For these calculations, the buckling rules of ASME Section VIII Division 2 (ASME VIII-2) Para. 4.4.12.2.i “Cylinder – Axial Compressive, Compressive Bending, and Shear” are utilized (Note that external pressure is not considered in the study). For the 5-span simplification, there is no axial force, and thus, column buckling is not possible. Therefore, the column buckling check in ASME VIII-2 was ignored for this application, and only the local buckling criterion was checked (see?Figure 3).

Figure 3. Only local buckling criteria are checked.

Static Deflection

Other than exceeding tensile stress or causing local buckling of the pipe, the pipe needs to maintain operational efficiency. It should be free-draining to avoid freezing (in the case of an undrained piping system). Limiting the beam deflection can ensure no detrimental change for either item. The API RP 574 subgroup decided to limit static deflections to 50% of the pipe diameter or 0.75 inches, whichever is less. A static deflection equation for a 5-span beam using the concentrated load (P) and the distributed load (w) was used to calculate the maximum deflection and compared to the acceptance criteria.

Natural Frequency

To minimize the concern for flow-induced vibration and avoid excessive flexibility of a piping system, a minimum natural frequency of 4 Hz was considered in the design criterion. The natural frequency equation with respect to span deflection was derived to be?. Substituting 4 Hz into this equation and solving for the deflection results in a deflection of 0.77 inches or less. The previously described deflection criterion of 0.75 inches is more limiting than the calculated flow-induced vibration minimum criterion of 4 Hz. Thus, no natural frequency criterion is required in addition to the static deflection criteria.

Local Support Stress

While global net section bending stresses have been considered thus far, there is a concern that stresses local to pipe supports could govern over the global stresses. To address this concern, several numerical methods (e.g., WRC 107/537, PD5500 Annex G, B31W, WRC 448, etc.) were investigated for evaluating stresses local to pipe shoes on the piping. All methods are versions of Biljaard’s Method and have limits on geometry that are mostly not applicable to the relatively small-diameter piping systems being examined. Thus, over 1,000 finite element analyses (FEAs) were performed to quantify stresses at shoe supports for larger pipe diameters (NPS 10-24), assuming a Class 150 pressure class. The FEA study was limited to carbon steel material. Elastic-Plastic and Bifurcation Buckling assessments were performed to look at plastic collapse and buckling failure modes, respectively. An example of an Elastic-Plastic FEA result is shown in?Figure 4.

Figure 4. Example of Elastic-Plastic FEA.

Assumption for Tstruct?Assessments

This structural Tstruct?study can have endless combinations, so assumptions had to be made to limit the assessment combinations to a practical number. Thus, assessments were only performed for A106-B (CS), A312-316 (SS), and A335-Paa (1.25Cr) materials in sizes NPS ? to NPS 24, and two assessment temperatures were chosen for each material. These temperatures affect the B31.3 allowable stress, the pipe elastic modulus, and the assumed operating pressure (based on pressure class rating). The study was determined to include a flanged gate valve, two weld neck flanges, and an additional 250-lb load at the midspan for pressure classes 150 to 600 (see?Figure 5). For pressure classes 900 to 2500, the flange gate valve was removed because it produced impractical and highly limiting results; it is also expected that such heavy valves would not be positioned at the midspan and that the valves would be adequately supported. The additional load of 250 lb is intended to incorporate unexpected loads the piping may see in service, such as a person standing on the pipe. For pipe sizes NPS 2 and smaller, the 250-lb load was removed as it produced impractical and highly limiting results. Similar to the unsupported valves, it is expected that small-bore piping would not be subjected to these loads in service.

Figure 5. Concentrated load assumption.

The subgroup decided to utilize the published ASME B31.1 span length values for a pipe filled with water as the support spacing for this assessment. Still, it included a maximum span length of 20 feet to align with the existing 20-foot limit in the current edition of API RP 574. Insulation weight also needed to be established to account for total weight on the system. To establish a set weight due to insulation, a table was created for typical pipe insulation thicknesses per pipe size. With the assumption of weld neck flanges and a flange gate valve, the weight of these components was estimated from a combination of the SIMFLEX-IV and CAESAR II flange weight databases. Typical flange and valve weights were assumed; however, actual flange and valve weights can vary based on the manufacturer. A summary of the assumptions for these assessments is provided in?Table 4.

Table 4. Assumptions.

Conclusion

A step-by-step assessment procedure was developed to assess Tstruct. Assumptions were required to narrow the assessment cases to a manageable number. These were documented in the study. Over 1,000 FEA cases were completed to assess stresses local to pipe shoe supports. In the end, a comprehensive and fully documented Tstruct?assessment was performed.

The preliminary results indicate that the current API RP 574 Tstruct?guidance may be unconservative for Class 300 and higher piping systems (see?Figure 6). This is mainly due to the addition of component weights and the assumed operating pressure set at the pressure class assessment temperature. The FEA results for the local shoe supports indicate they limit the structural Tstruct?for NPS 10 to NPS 24, but only the NPS 24 exceeds the API RP 574 existing Tstruct?values. API RP 574 is in the process of balloting the new structural Tstruct?tables, and an API Technical Report will be published documenting the detailed design basis for these tables.

Figure 6. Preliminary structural Tstruct?results for carbon steel at 400 ℉.

Article By : By Anthony Feller, Senior Engineer II at The Equity Engineering Group, Inc., and Kraig S. Shipley, P.E., Principal Engineer II at The Equity Engineering Group, Inc. This article appears in the July/August 2024 issue of Inspectioneering Journal.

Link : API RP 574, 5th Edition: Piping Structural Minimal Thickness Values (inspectioneering.com)

References

1. Marsh.com, 2018, “The 100 Largest Losses in the Hydrocarbon Industry,” MarshMcLennan, Marsh JLT Specialty, https://www.marsh.com/en/industries/energy-and-power/insights/100-largest-losses-in-the-hydrocarbon-industry.html.
2. API RP 574, 2024, “Inspection Practices for Piping System Components,” Fifth Edition, American Petroleum Institute.
3. ASME B31.3-2022, 2022, “Process Piping,” The American Society of Mechanical Engineers.

?