Page 185 - Reliability and Maintainability of In service Pipelines

Case Studies on the Application of Structural Reliability 169

1
0.9 [DS]
0.8 A
b/P'
0.7
k
0.6 j
Range 0.5 u
0.4
0.3
0.2
0.1
0
0 50 100 150 200 250 300
Time (year)
Figure 5.38 Comparison of range values of basic random variables.

In addition to the assessment of existing pipes, the reliability analysis methods
used in this case study can also show how changes in design parameters of sewers
(such as concrete cover) can affect the service life of the pipeline.
Sensitivity analysis was undertaken to identify factors that affect the probabil-
ity of pipe failure due to corrosion. The analysis showed less significant contribu-
tion of some variables in failure functions. Therefore, it would not be necessary
to consider those parameters as random variables and they can be treated as deter-
ministic constant values for further studies.

The results showed that among six random variables, [DS], b=P, and A have
the most effect on the probability of sewer failure. This effect is more consider-
able for lower values of time, which means special attention should be taken for
accurate determination of these variables for new pipelines.
The methods in Chapter 4 were applied to the reliability analysis of a UK con-
crete sewer case study. The results were verified by using a Monte Carlo simula-
tion, and more realistic results from multifailure assessment were achieved. The
degree of sensitivity to different variables was also investigated.

References

Ahammed, M., Melchers, R.E., 1994. Reliability of underground pipelines subject to corrosion. J.
Transport. Eng. 120 (6), Nov/Dec.
Ahammed, M., Melchers, R.E., 1997. Probabilistic Analysis of Underground Pipelines Subject to
Combined Stress and Corrosion. Eng. Struct. 19 (12), 988 994.
American Water Works Association. (2005). “Installation of Ductile-Iron Water Mains and Their
Appurtenances.” AWWA C600, Denver.
Anon. 2002. Office of Pipeline Safety, Available from: www.ntsb.gov and www.ops.dot.gov.