Piping vibration is a leading cause of loss-of-containment and downtime in the oil and gas, petrochemical, and fertilizer sectors. In addition to environmental and personnel safety, piping vibration often impacts profitability by constraining flow rates.
The majority of process piping design in refining and petrochemical facilities is governed by thermal code compliance rules of the ASME B31 codes. Most design codes provide qualitative guidance on the importance of designing against vibration fatigue without specific methods. Following code design rules without best practices can lead the designer to focus on flexibility at the expense of vibration susceptibility. In cleaner lower temperature services, such as LNG and pipeline facilities, piping vibration can be the most persistent damage mechanism. Most challenging is that once detected, the current body of API/ASME and international standards is inconsistent in the evaluation approaches of in-service piping vibration. Once identified, assessments usually require subject matter expertise involvement because traditional assessment methods are usually conservative.1, 2, 3, 4
Figure 1 shows a collection of commonly used vibration assessment screening criteria. All the curves have been converted into RMS velocity units for comparison. Current evaluation methods for piping vibration have demonstrated limitations:
Technical basis of these approaches is not widely available and often misused.
Not having options for higher levels of assessment if screening criteria are not satisfied. This leads to many unnecessary remediation costs due to overestimated risks.
Vibration screening criteria based on maximum peak-to-peak displacement units are very sensitive to the amount of data collected in cases of random vibration. 1, 2, 3
Vibration testing of machinery is a well-developed predictive maintenance strategy with four levels of ANST and ISO certifications (ISO 18436-2). While high-quality training is offered by specialty consulting firms (E2G Included), surprisingly, most machinery vibration certification programs do not provide adequate technical coverage of this topic. For example, too often machinery analysts do not collect sufficient data to measure severity and characterize piping vibration signals, particularly in cases of random vibration with dominant frequencies in the 2 – 10 Hz range. This is just one example of why formalized training and procedural guidance is needed.
The Future API 579 – Part 15
The API 579-1/ASME FFS-1 (API 579) Standard is a 14-part international standard for the fitness-for-service evaluation of pressure-containing equipment subject to various in-service damage mechanisms.
The newest part, Part 15 – Piping Vibration, has been under development for over 10 years and is aimed at unifying existing approaches with systematic procedures while including guidance to ensure vibration data quality. Currently, the method is in its final draft form and will be submitted for ballot this fall. Prior versions of the draft have been reviewed by the committee, as well as outside vibration experts in the industry, and generally well-received.
How does Part 15 compare to the current Part 14? Part 14 is the general fatigue methodology and applies to thermal or mechanical fatigue assessment, typically seen in the low- to mid-cycle regime (< 107 cycles). A 1 Hz vibration for one year will accumulate 1.35 x 107 cycles, and the majority of piping vibration problems have frequencies higher than 10 Hz. To complicate this, very little experimental justification exists for any S-N fatigue model (allowable alternating stress [S] vs allowable stress cycles [N]) in the very high-cycle or giga-cycle fatigue regime (cycles > 107). The available test data has substantial statistical scatter and uncertainty, making remaining life predictions using S-N-based models difficult if not impractical without providing sufficient conservatism.
The existing welded joint fatigue curve of Part 14 and ASME B&PVC, Section VIII, Division 2, Part 5 has been developed by fitting of experimental test data up to 107 cycles using a straight-line extrapolation. However, the set of data used to fit the Master S-N curve primarily consists of cases where failure was observed in less than 107cycles. Few data points in the set have lives beyond 107 cycles, and no data points have lives beyond 108 cycles. As a result, the utilization of the Master S-N curve for vibration assessment can lead to overly conservative risk assessments and costly mitigation measures, so much so that performing piping vibration assessments per Part 14 can often suggest failure should have already occurred. Instead of evaluating a remaining life using an S-N curve, Part 15 will provide an assessment of fatigue based on the endurance limit concept. Endurance limit concepts for welded joints have been successfully used in European codes and standards for decades, such as BS 7608 and other published literature.5, 6, 7 While published experimental justification of these limits is still lacking, the concept of an alternating stress range low enough not to propagate a crack is a well-understood concept used in elastic fracture mechanics and also an accepted principle for crack-like flaw analysis in BS 7910 and API 579 Part 9.
Similar to other API 579 parts, Part 15 provides a three-tiered system of evaluation:
Level 1: screens for the need to perform further analysis, the Level 1 assessment is designed as a field-based evaluation not requiring extensive (or any) engineering resources. Allowable RMS velocity will be presented in table format with values provided for mainline piping systems, small-bore cantilevered piping, and piping subject to shell-mode vibration. Most importantly, the Level 1 allowable levels are frequency independent, eliminating the need to identify the dominant frequency, thereby reducing implementation complexity.
Level 2: designed for use by engineering personnel, this approach offers a more accurate analysis specific to the system in question. Level 2 is built on existing ASME OM-3 approaches with some improvements and further guidance on implementation.8 As with all API 579 Level 2 assessments, the approach uses algebraic calculations. The approach does require the identification of the piping component with the highest susceptibility for fatigue damage, so knowledge of the system and an understanding of the shape of the vibration, while not required, is helpful and will reduce the conservatism of the assessment. As compared to Level 1, a substantial reduction in conservatism is often achieved.
Level 3: an advanced FEA-based numerical analysis approach providing the most accurate assessment at the cost of increased effort. More advanced vibration data collection methods are required to identify the problematic vibration response shapes of the piping so they can be modeled using numerical analysis. Similar to existing Level 3 assessments in other parts, Level 3 is less proceduralized and is instead guidance-based. Level 3 approaches are most valuable when the most accurate assessment of failure probability is required, which is often the case when the alternative option to mitigate requires unit outages, extensive construction, or piping redesigns.
The flexibility offered by these three levels can lead to more accurate and targeted evaluations, allowing for more informed decision-making.
With the publication of Part 15, consideration of vibration into a mechanical integrity program is made more relevant. Inspection for vibration or vibration surveys of facilities can become a formalized inspection strategy using Level 1 and Level 2 approaches as needed. Part 15 is a highly anticipated improvement to our industry standards and reflects the industry’s continuous commitment to innovation.
Author: Michael F.P. Bifano, Ph.D., P.E., ISO VCAT-IV, Rotating Equipment, Vibration, & Dynamics Team Leader
This article was previously published on Hydrocarbon Engineering, August 2023.
Comments