Environmental Engineering Reference
In-Depth Information
8.5 Conclusion
This chapter demonstrates the usefulness of the DSS model. The cost and frequency
functions for three different cities were estimated by two methods and then used to
determine the optimal renewal period via the DSS for each method. In City A, the
cost function behaved in an intuitive manner. However, the linear estimation
techniques resulted in a considerably different solution from the nonlinear tech-
niques. Linear estimation results in an optimal renewal period of 48.7 years,
whereas the nonlinear estimation results in a renewal of 35.9 years. In City B, the
cost functions again behaved in an intuitive manner. This time, however, the linear
and nonlinear estimation techniques produced very similar results in terms of the
optimal renewal time. Linear analysis resulted in an optimal renewal of 26.7 years
whereas nonlinear analysis resulted in an optimal renewal of 24.4 years.
The linear estimation techniques resulted in illogical results when used in City C.
The DSS in this city was minimized at a renewal period of 30.2 years. However, the
present value of the cost of service was negative at
an
impossible result. The nonlinear techniques, however, did not exhibit this illogical
behavior in City C. Interestingly,
this renewal period
—
the optimal renewal period under nonlinear
techniques is found to be 35 years.
There are, however, limitations that need to be mentioned. The data collected
from the Regional Municipality involved only a 20-year period. Therefore, the
behavior of the relevant functions before and after this period cannot be estimated
via regression analysis. This behavior can only be inferred based on theory or the
behavior during the period for which data exist. Another limitation of the analysis
stemming from the data involved the incomplete nature of data. At times, only a
portion of the relevant information was available. This seemed to occur more
frequently with the earlier data. Also, there were years where no failures were
recorded and these years sometimes fell in between years of numerous failures. It is
possible that due to the large costs of data collection, or due to personnel changes,
some data were not collected for these years. The last data limitation that should be
mentioned pertains to what the data includes. It was not possible to determine if the
cost included peripheral damages, and on occasions the cause of failure was not
known. This model assumes that failure occurs as a result of
“
normal
”
wear-and-
tear use under
conditions. However, if a pipe had been installed incor-
rectly or if a 100-year storm had occurred, then the optimal solution may not be
correct. The age at which failure occurs was also not given in the dataset and a
benchmark had to be established for each city to approximate a date of construction
for each distribution network as a whole. This is only an estimate and assumes that
the whole network was built at the same time.
As with all applied economic research, the data limitations identi
“
normal
”
ed above
should be taken into account. Nevertheless, the DSS demonstrates its power to
incorporate risk in managing water infrastructure assets.
In this application, the DSS was used to determine the optimal renewal for the
whole network. This is not the only way in which the DSS may be used. If enough
Search WWH ::
Custom Search