Page 113 - Pipeline Risk Management Manual Ideas, Techniques, and Resources

4/90 Corrosion Index
again, scores are conservatively assigned in the absence of more corrosion types, depending on whether the ILI indica-
more evidence. tions are internal or external wall loss. For example, suppose
that, prior to the ILI, an evaluator had assessed the coating con-
dition, CP effectiveness, etc., and had assigned the segment a
Adjustments subsurface corrosion score of 55 out of 70 (higher points indi-
cate more safety). If the ILI score, based on the recent inspec-
Several adjustments to previously assigned scores are appropri- tion, indicates that some damage might have occurred
ate. Some have already been discussed, such as adjustments for (suspicious indications), then the subsurface corrosion score
age of surveys or equipment malfunction potential. Additional would be capped at 60% x 70 (maximum points possible) = 42
adjustments will often be warranted when direct evidence is and the previously assigned 55 would be temporarily reduced to
included. The corrosion variables use mostly indirect evidence 42, pending an investigation. In other words, the previous
to infer corrosion potential, as is consistent with the historical assessment based on indirect evidence has been overridden by
practice of corrosion control in the industry. Because the scores the results ofthe ILI. The segment would be reassessed after an
will ideally correlate to corrosion rate, any detection of corro- investigation had determined the cause of the damage-how
sion damages or direct measurements of actual corrosion rate the mitigation measures may have failed and how the risk
can be used to calibrate the scores and/or tune the risk model. assessment may be incorrect.
Where a corrosion rate is actually measured, the overall corro- An ILI score that indicates no damages puts no limitations on
sion score can be calibrated with this information. The reverse corrosion scores. This is discussed in Chapter 5, if the direct
is not always implied, however. Caution must be exercised in inspection score is based upon un-verified ILI results, it can
assigning favorable scores based solely on the non-detection of eventually be improved through ‘pig digs’, that is, excavation,
corrosion at certain times and at limited locations. It is impor- inspection, and verifications that anomalies are indeed dam-
tant to note that the potential for corrosion can be high even ages. The limitation on corrosion scores can also be reduced
when no active corrosion is detected. even if the direct inspection score does not improve by damage
repair. This can happen if a root cause analysis of the detected
damages concludes that active corrosion is not present, despite
Previous damages a poor inspection score. For example, the root cause analysis
might use previous ILI results to demonstrate that corrosion
Results from an in-line inspection (ILI) or other inspections damage is old and corrosion has been mitigated.
may detect previous corrosion damage. When there is actual A critical aspect is the determination ofwhether the damages
corrosion damage, but risk assessment scores for corrosion represent active corrosion or are past damages whose progress
potential do not indicate a high potential, then a conflict seem- has been halted through increased corrosion prevention meas-
ingly exists between the direct and the indirect evidence. Such ures. Replacing anode beds, increasing current output from
conflicts are discussed in Chapter 2. rectifiers, eliminating interferences, and recoating are all
Sometimes we will not know exactly where the inconsis- actions that could halt previously active corrosion.
tency lies until complete investigations have been performed. This type of adjustment should be only temporarily
The conflict could reflect an overly optimistic assessment of employed. It will not give satisfactory long-term support for the
effectiveness of mitigation measures (coatings, CL etc.) or it risk model since it is, in effect, overriding risk information
could reflect an underestimate of the harshness of the environ- rather than finding and correcting discrepancies of evidence.
ment. Another possibility is that some of the information might
be inappropriately used by the risk model. For example, detec-
tion of corrosion damages might not reflect active corrosion.
As a temporary measure to ensure that the corrosion scores
always reflect the best available direct inspection information, Table 4.11 Temporarily limiting corrosion scores on the basis of
recent inspections
limitations can be placed on corrosion scores, in proportion to ~~~~~~~ ~
the direct inspection results. This will force the risk model to %of maximum
preferentially use recent direct evidence over previous assump- Interpretation of direct inspection score score.
tions, until the conflicts between the two are investigated.
Techniques to assimilate ILI and other direct inspection infor- Severe corrosion damages are identified. 0
mation into risk scores are discussed in Chapter 5. If such direct Significant corrosion damages are identified. 10
inspection scores are created, they can be used as input to the Possibility of some damages has been identified. 30
corrosion scores. Basically, a ‘ceiling’ is created that uses the Suspicious results suggest that damages might
have occurred.
inspection information (adjusted for age and accuracy) to over- Direct evidence has verified that no corrosion 60
ride scores derived from the more indirect evidence. This is has occurred.
illustrated in Table 4.1 1. 100
In this sample, the worst ILI scores, indicating the most ‘Use this value or current corrosion score, whichever indicates higher
extensive corrosion damage, limit the risk scores for one or corrosion threat.