Subscribe to our newsletter to receive the latest news and events from TWI:

Subscribe >
Skip to content

Which procedures for fitness-for-service assessment: API 579 or BS 7910? (July 2003)

   
John B Wintle

Paper presented at International Conference on Pressure Vessel Technology, Vienna, 7 - 10 July 2003
Proceedings ICPVT-1 - July 7-10, 2003 Vienna, Austria

Abstract

This paper discusses the choice that engineers face when selecting which procedures to use to assess the fitness-for-service of pressure equipment containing defects or damage. Results from a recent survey of the use of fitness-for-service assessment in industry identify API 579 and BS 79210 as the two most commonly used procedures. The scope and organisation of these procedures is then discussed and comparison made of the treatments of corrosion damage and crack-like defects. Future developments in fitness-for-service assessment procedures are considered in the light of the evolving European framework and international market for pressure equipment.

Introduction

Procedures for assessing the fitness-for-service (FFS) of pressure equipment containing defects or damage have developed since the late 1960's and there are now many procedures available for engineers to choose from. Two of the most commonly used are the recommended practice for assessing fitness-for-service published by the American Petroleum Institute (API) in API 579 [1] and the guidance for the assessment of defects metallic structures published by British Standards in BS 7910 . [2]

For many engineers, the decision of whether to use fitness-for-service assessment procedures and which procedures to use can be difficult. While users and regulators across industry now increasingly accept defects and damage in equipment assessed as fit-for-service, the differences between the available procedures and the implied safety margins are not so well understood. There can be uncertainty about the data and technical skills required to make good assessments. As a result, the benefits from fitness-for-service assessment may not have been as widespread as might have been expected.

The aim of this paper is to review aspects of the API and BS fitness-for-service procedures that will assist engineers make an informed decision about which procedures to use. The historical background to the procedures is outlined and the results of a recent survey into the use of fitness for service assessment are given. The scope and organisation of API 579 and BS 7910 are reviewed, particularly with regard to the treatment of different levels of assessment.

A more detailed comparison is made of the procedures for the assessment of corrosion and crack-like defects. It is here that some significant differences in approach arise. These illustrate the importance of using FFS assessment in the context of the general design criteria of the equipment. Finally, future developments of these procedures are considered in the light of the standards set by the European Pressure Equipment Directive and the expanding international pressure equipment market.

Historical development

Within general manufacturing industry, the pressure vessel codes had always recognised the inherent occurrence of welding defects and had set standards on permissible defect levels to control the minimum weld quality. The achievement of these standards sometimes necessitated a large number of weld repairs that were not only time consuming and expensive but could also be detrimental to integrity. It was recognised that these standards of permissible defectiveness could, in some cases, be very conservative, particularly where the material was ductile and the stresses low.

In order to reduce the number of weld repairs during manufacturing, a procedure to assessing fitness for service of equipment containing welding defects was sought. Research at TWI and elsewhere had characterised the fracture behaviour of welds containing defects by means of crack tip opening displacement (CTOD). [3] This research, and developments in the theoretical understanding of the factors influencing fracture, led to the development of PD 6493 - a British Standard Published Document for the assessment of defects in fusion welded structures. [4]

The development of PD 6493 was fuelled by the requirements of the oil and gas industry for offshore jacket platforms to exploit the North Sea reserves. These platforms were of large tubular construction, similar to large pressure vessels, and contained a huge number of welded joints between plates and nodal connections. Not only was there a need to achieve high weld production rates with minimum numbers of weld repairs, the owners had also to assure the safety of the structures to the possibility of fatigue cracking in the hostile North Sea environment. There was therefore a move towards the assessment of fitness-for-service of welds containing defects generated in-service and new rules for the assessment of fatigue cracks were added.

Another early driver for fitness-for-service assessment was within the nuclear industry where it was necessary to demonstrate high integrity and tolerance to welding defects of the safety critical pressure vessels. Subsequently, fitness for service assessment became vital for justifying the safety of nuclear vessels that were difficult to inspect or repair. These drivers led to the development of ASME XI and the R6 procedures. [5,6]

The wider application of fitness-for-service procedures to assess pressure equipment used in the refining and petrochemical industries has been a more recent development. The main drivers have been the need to extend the life of ageing equipment, to justify reduced inspection through risk based inspection, and to lower the high cost of repairs and replacement in terms of lost production. These and other factors prompted API to compile procedures and to publish recommended practice as API 579 in 2000.

In addition to defects in welds, the refining industry was also interested in assessing corrosion and locally thinned areas, and physical damage such as dents and gouges and overheating. Solutions for some of these types of damage had been derived from research work that had been published separately. Typical of these was the ANSI B31G methods for the assessment of locally thinned areas in pipelines. [7-10]

In the UK, work continued to develop PD 6493, and this led to other defect and damage mechanisms being considered. Work done by British Gas [11] had led to a procedure for the assessment of corrosion and locally thinned areas in pipelines and this was incorporated into PD 6493. After a significant period of world experience, consolidation and revision, British Standards recognised the standing of fitness for service assessment and PD 6493 became BS 7910.

Procedures for fitness-for-service assessment had also developed in other countries. In the late 1990's, within its fourth Framework Programme, the European Commission commissioned the SINTAP project [12] to review the available procedures and to recommend procedures that could be used. This started a process of harmonisation within Europe. It is now being continued through a thematic network called FITNET with the aim of determining if a European standard for fitness-for-service assessment can be realised.

Survey of FFS assessment

In 2001, TWI carried out a survey of the use of fitness-for-service assessment among its industrial members worldwide. All sectors of industry were represented, and many were major users of pressure equipment. The survey included mostly offshore oil and gas, petrochemicals, refining and nuclear and fossil power companies.

The response to the survey was good and informative. Of the respondents, 53% said that FFS procedures were used within their company. Whilst this figure may be regarded with some satisfaction, there are still a substantial number of companies that have apparently not accepted or are aware of the benefits, or perhaps simply do not feel capable of undertaking FFS assessment, preferring the established route of weld repair.

Another interesting statistic was that only 43% of respondents believed that the regulator of their operation of pressure equipment accepted FFS assessment. There is evidently a feeling among regulators that defects and damage in equipment should be repaired and concern with leaving them in situ. There could be many reasons for this reluctance to accept FFS assessment and further investigation to determine the barriers to acceptance and how these might be overcome would be of benefit.

Most companies (59%) using FFS assessment used published procedures, while a minority had developed their own procedures for dealing with certain damage types. The procedures most commonly used by general industry were API 579 andBS 7910. However, companies in the nuclear power sector tended to use procedures developed specifically for their industry such as R6 and ASME XI.

Companies gave many reasons for undertaking FFS assessment. A ranking of the reasons most frequently given gave the following results:

  • Determining the residual life of damaged plant
  • Ensuring safe operation beyond design life
  • Down-rating damaged plant below design
  • Demonstrating tolerance to defects within a safety case
  • Extending inspection intervals
  • Reducing duration of outage and shutdown

It is of interest that only some of these reasons involve actual defects and damage in equipment. FFS assessment is often made of postulated defects or damage so as to demonstrate the tolerance and safety margins in hand.

The ranking by frequency of type of equipment assessed gave the following results:

  • General pressure vessels
  • Process piping
  • Shell and tube heat exchangers
  • Transportation pipelines
  • Storage tanks
  • Fired heaters and boilers
  • Active equipment - safety relief valves, pumps, turbines and compressors

Whilst these results may reflect the experience of the respondents, it is significant that the use of FFS assessment for defects and damage in active equipment such as valves and rotary pumps is less than for passive equipment. Reasons for this could be that most procedures were developed with passive equipment in mind, and methods for moving parts where there may be inertial loads and dynamic effects are relatively undeveloped. Further development of procedures for assessing defects and damage in components of active equipment (e.g. shafts, blades) may be of benefit.

Scope and organisation of API 579

The American Petroleum Institute prepared API 579 specifically for assessing equipment in the refining and petrochemicals sectors designed to ASME codes. The procedures and supporting data relate to ASME design specifications and materials and are consistent with the design philosophy in terms of allowable stresses and factors of safety. A wide range of defect and damage types typically found during in-service inspection of refinery and petrochemical equipment are covered, with corrosion and locally thinned areas given prominence. Defect and damage types specifically considered include:

  • General metal loss
  • Local metal loss and gouges
  • Pitting corrosion
  • Blisters and laminations
  • Weld misalignment, dents and shell distortions
  • Crack-like flaws
  • Creep damage
  • Fire damage

API 579 has modular organisation based around each defect/damage type. The procedures are largely self contained within each module and derived from recognised authoritative sources. There are extensive annexes containing materials data, design formulae and reference solutions. Each module generally has three levels of assessment.

  • Level 1 is aimed at inspectors for use on site for quick decisions with the minimum of data and calculation.
  • Level 2 is intended for qualified engineers and requires simple data and analysis.
  • Level 3 is an advanced assessment requiring detailed data, computer analysis and considerable technical knowledge and expertise in FFS assessment procedures.

API 579 recognises the need of plant inspectors and engineering personnel on site to be able to undertake a quick initial assessment of defects and damage detected during plant examination. The level 1 procedures are designed for this purpose. Personnel with a broad engineering knowledge and experience can use these procedures with ease, although they may be simplistic and very conservative in some cases.

A more refined FFS assessment can always be made using the level 2 or 3 procedures. The degree of conservatism becomes progressively less as levels increase but this is compensated by the increased knowledge that is available aboutthe equipment, the defect/damage and the margins in hand. Application of level 2 and 3 procedures is usually a more complex process requiring greater specialist knowledge and experience.

Accordingly, API gives guidance for the knowledge and experience of engineers considered competent to undertake FFS assessments to each level. It recognises the need for adequate education and training in FFS assessment so that companies may have confidence in their staff making safe and correct judgements. Whilst qualifications and accreditation of welders, non-destructive testing personnel and plant inspectors have been in existence for some time, there is now perhaps a need to extend these schemes to cover fitness-for service assessment in a more formal way.

Scope and organisation of BS 7910

BS 7910 is published by British Standards for application to metallic structures across a range of industries and is therefore more general in its approach than API 579. From its origins From its origins in PD 6493, BS 7910 is strongly orientated towards the assessment of defects in and around welds, and its most detailed procedures are for the assessment of fatigue and creep crack growth and the proximity to fracture. Other failure modes, such as corrosion in pipes and pressures vessels, are covered at a guidance level.

BS 7910 comprises 10 sections and 15 annexes. Sections 1 to 6 describe the information required for assessment in terms of defect characteristics and dimensions, stresses and material properties. Section 7 to 10 give the proceduresfor assessment of fracture, fatigue, flaws under creep conditions and other modes of failure. The annexes contain normative procedures for dealing with certain situations (e.g. combined direct and shear stresses, determination of fracture toughness from variable materials data) and informative data (e.g. residual stress distributions for as- welded joints, weld strength mismatch, and proof testing and warm pre-stressing). This information is maintained at a state of the art level and is one of the most useful features of BS 7910.

BS 7910 gives procedures for assessing fatigue crack growth based on quality factors and crack growth calculation. A single procedure is given for assessing flaws at high temperature and corrosion, with advice given on further assessment if initial results are not favourable. There are three levels for the assessment of fracture based around the failure assessment diagram concept.

  • Level 1 is a screening procedure and the most conservative.
  • Level 2 is material specific and estimates the interaction between fracture and plasticity.
  • Level 3 involves a direct calculation of plasticity effects.

In general, BS 7910 is intended for use by qualified engineers trained in fracture mechanics, and significant computation of stresses and fracture parameters is often necessary. Because BS 7910 is intended to apply to equipment manufactured to different design codes and materials, (unlike API 579 which is based around ASME design and materials), specific stress and materials data is required even for level 1 fracture assessment. As a result, use of BS 7910generally requires personnel experienced in FFS assessment with access to appropriate data and/or testing facilities.

Comparison of procedures for corrosion assessment

Both API 579 and BS 7910 provide procedures for the assessment of various types of metal loss in pressure parts due to corrosion and other causes. API 579 has separate procedures for dealing with general metal loss (Section 4),local metal loss (Section 5) and pitting (Section 6). The BS 7910 Appendix G procedure can cover both general and local metal loss in pipes and pressure vessels and is similar but subtly different to that used by API 579 for local metal loss.

The API procedure for assessing general metal loss determines the average minimum thickness t am from a grid of spot thickness measurements around the corroded area. This procedure is illustrated in Figure 1. The part is assessed as fit-for-service if t am (minus any future corrosion allowance) is more than the ASME code minimum design thickness t min for the part and the minimum measured thickness t mm within the grid is greater than the larger of 0.5t min or 2.5mm. The approach is essentially to show that the part still falls within the original design basis of the code while ensuring there is adequate thickness for practical purposes and wear and tear.

t am - c > t min

and

t mm - c > max[0.5t min or 2.5mm]

The API procedure for assessing local metal loss determines a remaining strength factor from which a revised maximum working pressure with the metal loss is calculated as a fraction of the original maximum working pressure. The basis of the procedure is to treat the locally thinned area as a part through wall defect and to use the form of the equations developed by Battelle for local bulging failure through plastic limit mechanisms (the Folias factor). [13] According to API 579 Level 1 procedure for assessing local metal loss, the remaining strength factor RSF given by:

spjbwjuly2003e1.gif

where

spjbwjuly2003e2.gif
spjbwjuly2003e3.gif
spjbwjuly2003e4.gif

and

t mm = minimum measured thickness
t min = minimum (ASME) code design thickness
FCA = future corrosion allowance
s = length of corroded area
Fig. 1. Assessment of general metal loss in corroded pipe
Fig. 1. Assessment of general metal loss in corroded pipe

The procedure in BS 7910 Appendix G for the assessment of corrosion in pipes and pressure vessels is derived from research on pipelines carried out for British Gas. [12]

Based on a reserve strength factor, it uses the same form of the equations as API for the remaining strength factor for assessing local metal loss. The differences between the equations for reserve/remaining strength factor are:

(a) The constant in the formula for M is 0.31 instead of 0.48

(b) The wall thickness used in the expressions for R and λ is the nominal wall thickness of the part instead of the minimum (ASME) code design thickness.

The reserve strength factor is defined as the reduction in the failure pressure as a result of the metal loss. It has been extensively validated by tests and finite element analysis. Although consistent with failure controlled by plastic flow, its basis is essentially empirical.

In both cases the remaining/reserve strength factor is used to calculate a reduced/rerated maximum allowable working pressure (MAWPr) from the original code maximum allowable working pressure (MAWP), according to the API is

MAWP r = MAWP (RSF/0.9) for RSF<0.9

(API recommends the factor of 0.9 while BS 7910 leaves the choice of safety factors to the user)

As an example of the differences in the remaining/reserve strength factors predicted by the two procedures, consider a pipe made of SA 516 grade 70 with the following dimensions.

D = 762mm
T min = 9.0mm
t nom = 9.8mm
s = 1000mm

Figure 2 shows the predicted remaining/reserve strength factor calculated according to each procedure as a function of the minimum remaining pipe wall thickness. For this particular pipe, the remaining strength factor predicted by API 579 is higher than the reserve strength factor predicted by BS 7910 Appendix G, a difference primarily due to the use of code minimum thickness as opposed to nominal thickness in the formulae. With differing estimates of the remaining strength, use of the assessment procedures may give rise to different judgements when appropriate safety factors are applied.

Fig. 2. Comparison of remaining/reserve strength factors as a function of minimum remaining wall thickness
Fig. 2. Comparison of remaining/reserve strength factors as a function of minimum remaining wall thickness

Comparison of procedures for fracture assessment

API 579 and BS 7910 both define three levels of procedures for the FFS assessment of equipment containing crack-like defects liable to fracture. A comparison of these procedures is given below.

API 579 Section 9BS 7910 Section 7
Level 1 - Based on a maximum allowable length of defect related to the minimum design temperature Level 1 - Requires the use of a simple failure assessment diagram and fracture mechanics analysis
Level 2 - Uses a simple failure assessment diagram similar to BS 7910 Level 1 with refined fracture parameters Level 2 - Generalised and material specific failure assessment diagram, similar to R6
Level 3 - More refined FAD similar to BS 7910 Levels 2 and 3 Level 3 - Allows for ductile tearing and plasticity effects through a direct computation of J

Apart from API Level 1, the procedures are based on plotting a point on a failure assessment diagram (FAD) relating K r, the stress intensity factor/fracture toughness, and L r, the plastic limit load. The API acknowledges the use of these concepts from the BS and is similar in its approach. The use of a FAD requires computation of the K r and L r parameters from the stress distribution and reference solutions and therefore it should be applied by suitably trained engineers with a knowledge of fracture mechanics.

Fig. 3. Failure assessment diagram according to BS 7910 Level 1
Fig. 3. Failure assessment diagram according to BS 7910 Level 1

In contrast, the API Level 1 procedure is designed as a screening tool that inspectors can use on site. The procedure uses a diagram, Figure 4, that relates the maximum allowable length of flaw to the minimum design temperature of the equipment calibrated according to the reference temperature of the material. Important conditions for the application of the procedure are that the equipment must be designed to an ASME code and be made from a range of ASME specified materials, since these effectively define a maximum level of reference stress and a minimum assumed fracture toughness transition curve.

Fig. 4. API Level 1 screening curves for longitudinal defect in a cylindrical section
Fig. 4. API Level 1 screening curves for longitudinal defect in a cylindrical section

Different curves are provided for defects in base metal, welds with post weld heat treatment and welds without PWHT, and for defects in flat plates, cylinders and spheres. The most conservative assessment uses curves based on the assumption of a through thickness defect. These curves are applicable when the defect depth cannot be accurately determined by qualified non-destructive testing (NDT) or when the defect depth exceeds 6.3mm wall thickness in wall thicknesses between 25mm and 38mm. When NDT can accurately determine the depth of the flaw, curves based on a quarter thickness defect may be used for depths up to 0.25t in wall thicknesses less than 25mm, and for defects less than6.3mm depth in wall thicknesses between 25mm and 38mm.

The advantage of the API Level 1 method is that it can be used in conjunction with radiography and penetrant non-destructive testing methods when defect depth is not determined. Apart from the defect length, it just requires knowledge of the material, the minimum design temperature and the wall thickness. The method is therefore easy to use and avoids computations, and in many cases will be sufficient to assess FFS.

Both API 579 and BS 7910 provide reference solutions for the computation of stress intensity factor and limit load for defects in flat plates and cylinders. In a comparison exercise, differences were noted between the limit load solutions for an internal defect in a cylinder and the correction for plasticity, Figure 5, although these are not quite as much as the Figure would indicate with the false zero and unity extremes. The solutions for flat plates are very similar. It is not the objective of this paper to say which is right; simply to note that there are differences that could affect the outcome of an assessment.

Fig. 5. Comparison of fracture parameters calculated from API 579 and BS 7910 reference solutions
Fig. 5. Comparison of fracture parameters calculated from API 579 and BS 7910 reference solutions

Which procedure to choose?

The industry survey shows FFS assessment of pressure equipment containing defects and damage is now well accepted, although there is still a degree of reluctance to rely on FFS assessment among some companies and regulators. The reasons for this are not clear, but the endorsement of FFS assessment by the American Petroleum Institute and British Standards should give its use more confidence and impetus. More use of FFS assessment can be expected as plant owners extend the life of ageing equipment and apply risk based inspection.

For pressure equipment in a non-nuclear context, API 579 and BS 7910 are the most commonly used procedures for FFS assessment. Both are recognised as representing best practice and safe, although they may not always give the same results. In many applications both API 579 and BS 7910 will be suitable. The choice may depend on company policy and the attitude of the national regulating authority and access the necessary data and sources of information, training and support.

In terms of the advantages and applicability of the two procedures, readers may find the following points helpful.

  • API 579 is intended for equipment designed using the ASME code and materials and gives results consistent with the original ASME design safety margins.
  • API 579 may be used for equipment designed to other codes but users should be prepared to interpret the procedures in an appropriate manner.
  • BS 7910 is applicable to all metallic structures and materials and is written in a more generalised manner without reference to a particular industry, design code or material thereby allowing users to decide safety margins.
  • API 579 covers a wide range of damage types typically found in refining and petrochemicals application, and gives procedures for different types of metal loss, physical damage, low and high temperatures, and crack like defects.
  • BS 7910 deals comprehensively with fatigue and fracture of defects in and around welded joints and gives annexes covering advanced aspects such as mismatch, mixed mode loading , residual stress effects and leak before break.
  • API 579 is designed at level 1 for use by plant inspectors and plant engineering personnel with the minimum amount of information from inspection and about the component.
  • BS 7910 requires some technical expertise in fracture mechanics and access to fracture parameter solutions and toughness data at all levels.
  • API 579 is supported by a number of organisations based in the USA where most experience resides.
  • BS 7910 was developed in the UK where TWI is the main source of expertise, training and software.

Future developments

Both API 579 and BS 7910 will continue to be developed and updated. The latest European development in fitness-for-service is FITNET, a thematic network set up under Framework V. This has the objective of selecting, developing and extending the use of FFS procedures in Europe. It will review the best procedures currently in use and consider their application for pressure equipment meeting the new harmonised standards and the essential safety requirements of the PED.

The safe use of FFS assessment must depend on having an adequate level of competency, training, information and support necessary to make technical judgements about potentially hazardous equipment. Industry will always like quick simplified procedures that can be used on site without detailed information, analysis and specialist knowledge. Expert systems may be the means to reconcile these aims.

References

  1. API Recommended Practice 579, Fitness-for-Service, API Publishing Services, First edition January 2000
  2. British Standard 7910, Guide on methods for assessing the acceptability of flaws in metallic structures, 1999 incorporating amendment No 1
  3. Wells A A, IIW Houdrement Lecture, Brit Welding J., 12, No 1, 2, Jan (1965)
  4. British Standard Published Document 6493, Guidance on some methods for the derivation of acceptance levels for defects I fusion welded joints, 1980
  5. ASME Boiler and Pressure Vessel Code Section XI, Rules for in-service inspection of nuclear power plant components,, The American Society of Mechanical Engineers, New York, 2001
  6. British Energy, R6 Assessment of the integrity of structures containing defects, Rev 3, February 1997.
  7. Kiefner J F, Duffy A R, criteria for determining the strength of corroded areas of gas transmission lines, American Gas Association Conference 1973
  8. ANSI/ASME B31G, Manual for determining the remaining strength of corroded, A supplement to the ASME B31 code for pressure piping, The American Society of Mechanical Engineers 1984
  9. Kiefner J F, Vieth P H, A modified criterion for evaluating the strength of corroded pipe, Final report for Project PR 3-805 to the pipeline supervisory committee of the American Gas Association, Battelle Ohio, 1989
  10. ANSI/ASME B31G, Manual for determining the remaining strength of corroded pipelines (revision of ANSI/ASME B31G-1984), A supplement to the ASME B31 code for pressure piping, The American Society of Mechanical Engineers, New York, 1991
  11. Fu B, Batte A D, Advanced methods for the assessment of corrosion in linepipe, UK Health and Safety Executive Report OTO 97-065, HSE Books 1999
  12. Bannister A C ed, Structural integrity procedures for European industry (SINTAP), Final report BE95-1426/FR, British Steel plc (now Corus), Swinden Technology Centre, September 1999
  13. Kiefner J F,Maxey W A,Eiber R J, Duffy R, The failure stress levels of flaws in pressurised cylinders, ASTM STP 536, ASTM, Philadelphia, 1973

For more information please email:


contactus@twi.co.uk