Page 132 - Pipeline Risk Management Manual Ideas, Techniques, and Resources

Risk variables and scoring 51109
lems. the inregrip verification variable receives a score of 30.5 One challenge often faced by evaluators is the requirement
out of 35 possible points, based on this ILI program. that they use results from different types of ILI tools. Different
These points will be reduced over time until either the infor- tools will often have different detection capabilities and accura-
mation has aged to the point of little value or the ILI is repeated. cies. Even similar tools used at different times can have signifi-
For instance, if a fixed 5-year deterioration is assumed the cant variations due to the evolving technologies. To make use of
score after 3 years will be: all available information, it might be necessary to establish
equivalencies between indications from various tools. An indi-
(5-3)/5x30.25= 14points cation from a low-resolution tool should be weighted differ-
ently from one from a high-resolution tool. given the different
Scoring ILI results The previous discussion focused on scor- uncertainties involved in each.
ing the ILI process-how timely and robust was it? It did not
take into account the use of the results of the ILL That aspect is Approach 1
discussed here.
ILI results provide direct evidence of damages and, by infer- An example system for generalizing ILI results is outlined here.
ence. of damage potential. Such evidence should he included in Under this scoring scheme, pipeline segments are character-
a risk assessment. The specific use of direct evidence in evalu- ized in terms of past damages that might reduce pipe strength
ating riskvariables is discussed in Chapter 2. and indicate possibly active failure mechanisms. Data from the
ILI results provide evidence about possibly active failure most recent 1LI runs for every pipeline are collected. The
mechanisms, as illustrated in Table 5.4. pipelines are divided into fixed length segments-perhaps 100
or 1000 ft long. For each segment, all ILI indications are accu-
lritegriv assessment andpipe strength When integrity assess- mulated and characterized based on their frequency and sever-
ment information becomes available, it can and should become ity. Each type of anomaly is counted and weighted and then
a part of the pipe strength calculation. All defects left uncor- used in setting five variables, discussed in the following sub-
rected should reduce calculated pipe strength in accordance sections, that characterize the relative amount and severity of
with standard engineering stress calculations described in this damage to the pipe wall.
chapter. Defects that are repaired should impact other risk
model variables as direct evidence of failure mechanisms (see External damage This variable represents the relative quan-
Appendix C). Even if no defects are detected uncertainty has tity and severity of dents, gouges, and other indications of out-
been reduced with a corresponding reduction in perceived risk. side force damage. It is created by using the counts of dents,
If the information is from very specific portions of the dents on welds, and top side indications from recent ILI results.
pipeline-such as after a visual or NDT inspection of an exca- Each is weighted according to its possible impact on pipe
vated section of pipe-a zone-of-influence approach (see strength. Higher weightings are assigned to anomalies on welds
Chapter 8) or ideas taken from statistical sampling techniques andor those more likely to be related to third-party damage and.
can be used to expand integrity information for scoring longer hence, possibly involving a gouge or a more severe contour or
stretches ofpipeline. dent. As an overall adjustment to risk scores, this variable
Full characterization of the impact of 1LI indications on pipe reduces the previously calculated third-pry inch by up to 10%
strength would involve statistical analysis of anomaly measure-
ments, considering tool accuracies. But even without detailed Corrosion remaining strength This variable represents the
calculations. the effective actual wall thickness should be relative remaining strength. from a pressure-containing view-
reduced depending on the nature of the anomalies detected in point, of the pipe after allowing for metal losses due to corro-
the pipeline segment being scored. For example, a severe corro- sion. It represents the relative severity of metal loss by
sion indication might warrant a 50 to 70% reduction in effective accumulating the lengths and depths of metal loss indications
pipe wall thickness. in each pipeline segment. Greater emphasis is given to lengths
This direct consideration of ILI results presumes that spe- in keeping with commonly accepted formulas for calculating
cific anomalies have been mapped to specific pipeline seg- the remaining strength of pipe. As an adjustment to risk scores,
ments and that anomalies are few enough to consider this variable reduces the previously calculated sufeh~,fuctor by
individually. Ifthis is not the case, ILIresults can also be used to up to 30%.
generally characterize the current integrity condition. This can
be done either as a preliminary step pending full investigations Corrosion metal loss This variable represents the relative
or as stand-alone evidence. quantity and severity of corrosion damages. It measures the

Table 5.4 Interpretation of ILI results
1L1 unotmilv Failwe mechanism
Geometric anomalies (dents. wrinkles. out-of-round pipe) Third-party damages (normally on top and sides); improper support;bedding
(normally on bottom); excessive external loads
Metal loss (gouging and general. pitting, and Gouge = third-party damage; metal loss = external or internal corrosion
channeling corrosion)
Laminations. cracks. or cracklike features Fatigue and/or manufacturing defects