Tc 67/sc 4 n date: 2005-03-9 iso/wd XXXXXX ISO tc 67/sc 4/wg 6 Secretariat: Design of dynamic risers for offshore production systems Élément introductif — Élément central — Élément complémentaire  Warning

Service Life

1. Fitness-for-service

The flaw acceptance standards included in the majority of fabrication and construction codes are based on good workmanship practices. More recently, a number of fracture mechanics-based assessment procedures have been developed which enable the significance of weld discontinuities to be assessed on a "fitness-for-service" basis. Using this concept, a structure is considered to be fit-for-service provided it can be operated safely throughout its design life. The adoption of fitness-for-service concepts in several codes has resulted in the development of more rational flaw acceptance criteria.

Fitness-for-service assessment procedures can also be used to assess the significance of flaws or cracks in a structure which may have escaped detection during original fabrication or have developed in service (e.g., fatigue cracks or environmental cracks). These procedures enable operators not only to assess the significance of flaws in structures but determine remaining life, inspections intervals and inspection sensitivity requirements.

1. Fracture mechanics assessment procedures

• 
  size of discontinuity;
  
• 
  material toughness;
  
• 
  applied stress.
  

Figure 40 summarizes the different loading paths that can result in the failure of a statically loaded structure (assuming creep and corrosion effects are not significant). The loading paths range from brittle fracture under nominally elastic loading (applied stresses well below yield) to plastic collapse (overload of remaining ligament).

In cases where brittle fracture occurs at low applied stresses, the concept of linear elastic fracture mechanics (LEFT) can be applied, i.e., a stress intensity factor (K) approach. At the other extreme when the failure mechanism is plastic overload, assessments should be performed using limit load or plastic collapse analyses.

Between these two extremes, elastic plastic fracture mechanics (EPFM) methods can be applied to assess the integrity of structures. The two most common EPFM fracture characterizing parameters are the crack-tip opening displacement (CTOD) and J-integral (J). Fitness-for-service methodologies have been developed using both of these parameters.

A number of different fitness-for-service assessment methodologies for calculating allowable or critical flaw sizes are currently in use throughout the world. Nevertheless, the most widely used assessment methodology for offshore structures is BSI PD6493. This document includes detailed fracture and fatigue assessment procedures for welded structures.

Fitness-for-service design philosophies can be applied to welded structures in several different ways. The first and most widely used approach is to develop relaxed flaw acceptance criteria which are typically derived assuming a minimum toughness level which is incorporated in a weld procedure qualification. Nevertheless, fitness-for-service concepts can also be used to develop inspection criteria including sensitivity and probability of detection. In addition, fitness-for-service concepts can be used to demonstrate that flaws, which may escape detection, will not impair overall structural integrity even under accident or overload conditions.

2. Fatigue design

In the case of welded joints, there is practically no crack initiation period due to the presence of weld toe discontinuities which behave as preexisting cracks. As a result, the bulk of the fatigue life of a welded joint can be attributed to fatigue crack propagation.

The difference in the fatigue behavior of parent material and welded joints (i.e., the difference in the crack initiation phase) has significant effects on overall fatigue performance and fatigue design. In general, the fatigue strength of unwelded components increases with material tensile strength due to the increased initiation life associated with higher strength materials. In the case of welded joints however, the fatigue strength is relatively unaffected by material tensile strength because the bulk of the fatigue life of a welded joint is spent in the propagation phase, and although crack propagation rates can change from one material to another there is no consistent trend with regard to tensile strength.

1. S-N curve approach

   1. Parent material S-N curves

Parent material fatigue properties are influenced by mean stress. High mean stress will increase the fatigue damage and reduce the fatigue life. Commonly accepted methods used to account for the effect of the mean stress on the fatigue damage include the Goodman mean stress correction method and the Gerber mean stress correction method. Many fatigue curves have been adjusted for or inherently include the effects of mean stress and no further adjustment is required when using these curves. However, if the fatigue curve being used does not account for the effects of mean stress, then an adjustment of the mean stress effects should be included in the analysis.

2. Welded joint S-N curves

A classification system is used to relate details of the welded joint (i.e., weld joint geometry) and the appropriate design S-N curve. In general, the classification depends on the joint type, joint geometry, form of applied loading and the testing environment. Since these parameters will control the local stress distributions at welded joints, the different design S-N curves can be viewed as a method of accounting for local stress effects arising from weld joint geometry and the form of applied loading. Unlike parent material S-N curves, S-N curves for welded joints are in general, not influenced by mean stress, because the bulk of the fatigue life is spent in the crack propagation phase. In addition, weld residual stress always exists to some extent.

2. Fatigue crack growth assessment procedures

Fracture mechanics assessment procedures can also be used to predict the life of a joint with a weld discontinuity or a parent material component (e.g., casting of forging) which contains a flaw. The most widely used fatigue crack growth relationship is the Paris Law:

da/dN = C( K)^m ...(60)

where:

K = applied stress intensity factor range,

a = crack size,

N = number of cycles,

For design purposes, it is sometimes necessary to estimate the limiting flaw size which will not extend by fatigue during service. The limiting crack size below which fatigue crack growth will not occur can be calculated using fatigue threshold concepts. This information can be useful in defining the required sensitivity of NDT equipment and also drawing up inspection criteria.

Fracture mechanics-based fatigue assessment procedures can also be used to define inspection criteria and in particular inspection intervals. This is achieved by assuming the maximum flaw size which may escape detection and specifying a critical crack size which should be detected to avoid unnecessary damage to the structure or the risk of failure.

3. Stresses for fatigue assessment

The loads used to estimate the fatigue damage generated by the low frequency wave drift are obtained from global analyses of the riser using the low frequency wave-induced motions of the FPS. These loads can be generated using static or dynamic global analyses, depending on the frequency of the motion. Dynamic analyses should be performed if the inertial loads generated by the low frequency wave-induced motions are significant. The loads generated by VIV must be obtained from a VIV analysis of the riser.

It is important to remember that all of the loads that contribute to fatigue damage of the riser components are cyclical in nature. A number of cycles or probability of occurrence for each type of load must be known to estimate the expected fatigue damage. For the primary wave cycle damage, this information is usually given in the terms of number of wave cycles for a deterministic analysis and number of storms for a spectral (stochastic) analysis.

When assessing parent material components or welded joints in the post-weld heat-treated condition, the stress range required for a fatigue assessment is the total stress range if the stress range is entirely tensile. In situations where the stress range is partly compressive, then the stress range for the fatigue analysis should be taken as the entire tensile stress range plus 60% of the compressive stress range. For welded joints in the as-welded condition, the stress range to be used in fatigue assessments should be based on the full stress range regardless of whether the stress range is partly or wholly compressive.

In the S-N fatigue approach, peak stress ranges are calculated for each bin in the fatigue weather scatter diagram. These peak stress ranges are equal to the product of the dynamic pipe wall stresses obtained from the global riser analysis and the stress amplification factors (SAFs) calculated for the riser components. The dynamic pipe wall stresses are calculated from the dynamic bending moments and the dynamic tension variations. The SAFs are derived by local finite element analysis of a structural component. The SAFs represent the stress increase caused by geometry, three-dimensional effects and load paths through the structural component.

In cases of variable amplitude loading, the Miner cumulative damage summation is used to sum the damage from the different loads. As long as all of the load cycles are included in the analysis, the total damage obtained using the cumulative damage summation is not affected by the sequence in which the loads are applied to the riser component. The method used to estimate the fatigue damage is different for the deterministic and spectral analysis methodologies. See API RP-2A for a description of the deterministic and spectral fatigue analysis methodologies. The estimated fatigue life is equal to the inverse of the annual damage generated by the fatigue loads. The total fatigue damage is equal to the combined damage generated by the fatigue loads (i.e., combined primary wave frequency, low frequency wave drift, and FPS offset-induced damage plus VIV damage) if the installation damage is negligible. (See 6.2.5.4.)

Care must be taken in calculating the stresses and SAFs to be used in the fatigue analysis. Relatively small changes in the stresses and SAFs can result in large differences in fatigue life. Fatigue life is proportional to the stress ranges and SAFs, each raised to the power of the S-N curve inverse slope (from 3 to 5). It can be demonstrated that, for an S-N slope of 5, doubling of either the stress range, SAF, or any product of these, decreases fatigue life by a factor of 32. For example, if the structural component had a fatigue life of 100 years, doubling the product of the stress range and SAF would reduce the fatigue life to 3 years.

3. Environmental cracking

When assessing environmentally-assisted cracking on a fitness-for-service basis, the user must first identify the cause of cracking or damage and assess the possibility of further growth or damage, i.e., is the cracking associated with original fabrication defects or previous operating conditions which were more severe than current conditions (e.g., upset conditions or a change in operating conditions). If the possibility of further growth or damage exists, the user must estimate remaining life and determine appropriate inspection intervals or develop an on-stream monitoring approach. The basic options for treating a component which has experienced environmentally-assisted cracking are as follows:

1. 
   prevent further cracking or damage;
   
2. 
   predict remaining life using appropriate crack growth law and determine appropriate inspection intervals.
   

1. Prevention of further cracking

The fatigue stress intensity factor threshold should also be considered when assessing the possibility of subsequent crack growth. It should be noted that the fatigue stress intensity factor threshold is also very sensitive to the environmental conditions.

2. Predict remaining life using crack growth law and determine inspection intervals

Depending on the material and the operating conditions, environmental crack growth relationships can take several forms. In most cases, the rate of crack growth is a function of the applied stress intensity factor although there are some forms of environmental cracking which are simply a function of time. With this in mind, it is essential that the user selects an appropriate crack growth relationship for the component and operating conditions under consideration. This can be a major problem since crack growth rates can be very sensitive to changes in the process environment. Nevertheless, assuming an appropriate crack growth relationship is available, the user can calculate the number of cycles or time required for the existing cracks to increase to a critical size. This information can then be used to set inspection intervals to monitor crack growth and enable the user to decide when to take remedial action or withdraw the component from service.

2. Wear