Page 77 - Pipeline Risk Management Manual Ideas, Techniques, and Resources
P. 77

3/56 Third-party Damage Index
         along and either side of the ROW. The opportunity for early
         discovery  lies  in  the  ability to  detect activities before  the
         pipeline  ROW is encroached.                          Frequency distribution
           Note also the ability of certain aircraft (helicopters) to take   urve based on recent
         immediate action to interrupt a potentially dangerous activity.   historical data
         Such interruptions include landing the aircraft or dropping a   6
                                                     C
         container containing a message in order to alert the third party.   tn
                                                     3
           The  suggested point  schedule will  award points based  on   0-
                                                     2
         patrol frequency under the assumption of optimum effectiveness.   LL
         Ifthe evaluatorjudges the effectiveness to be less than optimum,
         he can reduce the points to the equivalent of a lower patrol fre-
         quency. This is reasonable because lower frequency and lower
         effectiveness both reduce the area of opportunity for detection.
           If practical, the patrol frequency can be determined based on
         a statistical analysis of data. Historical data will often follow a   0   1   2   3   4   5   6
         typical rare-event frequency distribution such as those shown   Number of Potential Threats Found on a Single Flight
         in Figures 3.8 and 3.9. Figure 3.9 is based on tabulated esti-
         mates shown in Table 3.3. Once the distribution is approxi-   Figure 3.8  Typical patrol data.
         mated,  analysis of  the  curve  will  enable  some  predictive
         decisions to be made. An analysis of the “opportunity to detect”   An example point schedule is as follows:
         various common excavation activities is presented at the end of
         this chapter.  Such analyses can be the basis of determining   Daily          15 pts
         patrol frequency or assessing the probability of detection for   Four days per week   12 pts
         any given frequency. For example, management may decide   Three days per week   IO pts
         that the appropriate patrol frequency should detect, with a 90%   Two days per week   8 pts
         confidence level, 80% of all threatening events. This might be   Once per week   6 pts
         based  on  a  costhenefit  analysis. Patrol  frequencies at  or   Less than four times per month; more than once
         slightly above this optimum can receive the highest points.   per month        4 pts
                                             Detection Opportunity
              1 .o

              0.9

              0.8
           8
           c
              0.7
           L
           0
           3
           6  0.6
           c
           0
             0.5
           n
           a
           n
           a  0.4
           >
           .-
           0)
           ;ri  0.3
           -
           E’
           5  0.2
             0.1
               0
                0        2         4        6        8       10        12       14       16
                                      Days of Proximal Activity and/or Evidence
                                      Figure 3.9  Patrol detection opportunities.
   72   73   74   75   76   77   78   79   80   81   82