Page 126 - Pipeline Risk Management Manual Ideas, Techniques, and Resources
P. 126

Riskvariables and scoring 5/103
              sures. A 100-psi pressure cycle will have a potentially greater   Example 5.6: Scoring fatigue potential
              effect on a system rated for 150 psi MOP than on one rated for
               1500 psi. Most research points to the requirement oflarge num-   The evaluator has identified two types of cyclic loadings in a
              bers of cycles at all but the highest stress levels, before serious   specific pipeline section: (1) a pressure cycle of about 200 psig
              fatigue damage occurs.                     caused by the start of a compressor about twice a week and (2)
                In many pipeline instances, the cycles will be due to changes   vehicle traffic causing a 5-psi external stress at a frequency of
              in internal pressure. Pumps, compressors, control valves, and   about  100 vehicles  per  day. The  section  is  approximately 4
              pigging  operations  are  possible  causes  of  internal  pressure   years old and has an MOP of 1000 psig. The traffic loadings
              cycles. The following example schedule is therefore based on   and the compressor cycles have both been occurring since the
              internal pressures  as percentages  of MOP. If another type of   line was installed.
              loading is more severe, a similar schedule can be developed.   For  the  first  case,  the  evaluator  enters  the  table  at  (2
              Stresses caused by vehicle traffic over a buried pipeline would   startdweek x 52 weekdyear x 4 years) = 416 cycles across the
              be an example of a cyclic loading that may be more severe than   horizontal axis, and (200 psig/lOOO psig) = 20% of MAOP on
              the internal pressure cycles.              the vertical axis. This combination yields a point score of about
                This is  admittedly  an  oversimplification  of this  complex   13 points.
              issue. Fatigue depends on many variables as noted previously.   For the second case, the lifetime cycles are equal to (100
              At certain stress levels, even the frequency of cycles-how  fast   vehicles/day x 365 daysiyear x 4 years) = 146,000. The magni-
              they  are occurring-is   found to affect the failure point. For   tude  is equal  to  (5 psig/lOOO psig)  = 5%.  Using  these  two
              purposes of this assessment, however, the fatigue failure risk   values, the schedule assigns a point score of 7 points.
              is  being  reduced  to  the  two  variables  of  stress  magnitude   The worst case, 7 points, is assigned to the section.
              and  number.  The  following  schedule  is  offered  as  a  pos-
              sible simple way to evaluate fatigue’s contribution to the risk   Cracking: a deeper look
              picture.
                One cycle is defined as going from the starting pressure to a   All materials have flaws and defects, if only at the microscopic
              peak pressure and back down to the starting pressure. The cycle   level. Given enough stress, any crack will enlarge, growing in
              is measured as a percentage of MOP.        depth and width. Crack growth is not predictable under real-
                In this example of assessing fatigue potential, the evaluator   world  conditions.  It  may  occur  gradually  or  literally  at  the
              uses the scoring protocol illustrated in Table 5.2 to analyze   speed of sound through the material. (See also discussions on
              various combinations of pressure magnitudes and cycles. The   possible failure hole sizes in Chapters 7 and 14.)
              point value is obtained by finding the worst case combination   As contributors to fatigue failures, several common crack-
              of pressures and cycles. This worst case is the situation with   enhancing  mechanisms  have  been  identified.  Hydrogen-
              the lowest point  value. Note the “equivalents” in  this table;   induced cracking (HIC), stress corrosion cracking (SCC), and
              9000 cycles at 90% of MOP is thought to be the equivalent of   sulfide stress corrosion cracking (SSCC) are recognized flaw-
              9 million cycles at 5% of MOP; 5000 cycles at 50% MOP is   creating or flaw-propagating phenomena (see Chapter 4). The
              equal to 50,000 cycles at 10% of MOP, etc. In moving around   susceptibility of a material to these mechanisms depends on
              in this table, the upper right corner is the condition with the   several variables. The material composition is one of the more
              greatest risk, and the lower left  is the least risky  condition.   important variables. Alloys, added in small quantities to iron-
              The upper left corner and the lower right corner are roughly   carbon  mixtures,  create  steels  with  differing  properties.
              equal.                                     Toughness is the material property that resists fatigue failure. A
                Note also that Table 5.2 is not linear. The designer ofthe table   trade-off often occurs as material toughness  is increased but
              did not  change point values proportionately  with changes  in   other important properties such as corrosion resistance, weld-
              either the magnitude or frequency of cycles. This indicates a   ability,  and  brittleductile  transitions  may  be  adversely
              belief that changes within certain ranges have a greater impact   affected. The fracture toughness of a material is a measure of
              on the risk picture. The following example illustrates further   the degree of plastic deformation that can occur before full fail-
              the use of this table.                     ure. This plays a significant role in fatigue failures. Much more
                                                         energy is required to fail a material that has a lot of fracture
                                                         toughness, because the material can absorb some of the energy
                                                         that may otherwise be contributing directly to a failure. A larger
              Table 5.2  Fatigue scores based on various combinations of
              pressure magnitudes and cycles             defect  is  required  to  fail  a  material  having  greater  fracture
                                                         toughness. Compare glass (low fracture toughness) with cop-
                                 Lifetime cycles         per (high fracture toughness). In general, as yield strength goes
                                                         up,  fracture toughness  goes down. Therefore.  flaw tolerance
                                                         often decreases in higher strength materials.
                                                           Another  contributor  to  fatigue  failures  is the  presence  of
               IO0     7      5       3      I      0    stress concentrators. Any  geometric  discontinuity  such  as a
              90       9      6       4      2      1    hole,  a crack, or a notch, can amplify the  stress level in the
              7s       10     1       5      3      2    material. Coupled with the presence of fatigue loadings, the sit-
              50       11     8       6      4      3    uation can be further aggravated and make the material even
              25       I2     9       1      5      4    more susceptible to this type of failure.
               10      I3    IO       8      6      5
              5        14     II      9      7      6      The process  of heating  and cooling  of steel  during initial
                                                         formation and also during subsequent heating (welding) plays
   121   122   123   124   125   126   127   128   129   130   131