Diagnostics of Regular Performance of Water Distribution Networks and its Relation to the Network Reliability - Pattern Recognition for Reliability Assessment of Water Distribution Networks - page 192

Table 7.11 Correlation patterns of GA-optimised network at the demand increase of 98% - ngt01-151

Name V j Q j v j h f,j S j P j T j pb j pb j /L j hl j PBI j
ngt01-15101 -0.02 0.45 0.45 0.66 0.64 0.76 -0.02 -0.08 -0.32 0.63 -0.75
ngt01-15102 -0.03 0.50 0.50 0.71 0.70 0.81 -0.02 -0.09 -0.34 0.69 -0.81
ngt01-15103 -0.04 0.56 0.56 0.77 0.76 0.87 -0.02 -0.11 -0.36 0.74 -0.86
ngt01-15104 -0.05 0.61 0.61 0.82 0.82 0.91 -0.02 -0.13 -0.38 0.78 -0.90
ngt01-15105 -0.07 0.66 0.66 0.86 0.87 0.95 -0.03 -0.14 -0.39 0.82 -0.93
ngt01-15106 -0.07 0.70 0.70 0.89 0.90 0.97 -0.03 -0.15 -0.40 0.85 -0.95
ngt01-15107 -0.08 0.73 0.73 0.91 0.93 0.98 -0.02 -0.16 -0.39 0.86 -0.96
ngt01-15108 -0.09 0.75 0.75 0.93 0.94 0.99 -0.02 -0.16 -0.39 0.87 -0.97
ngt01-15109 -0.09 0.77 0.77 0.94 0.96 0.99 -0.02 -0.16 -0.37 0.88 -0.97
ngt01-15110 -0.10 0.78 0.78 0.94 0.96 0.99 -0.03 -0.16 -0.36 0.88 -0.96
ngt01-15111 -0.10 0.79 0.79 0.94 0.97 0.98 -0.02 -0.16 -0.34 0.88 -0.95
ngt01-15112 -0.10 0.79 0.79 0.95 0.97 0.98 -0.03 -0.16 -0.31 0.88 -0.95
ngt01-15113 -0.10 0.80 0.80 0.95 0.97 0.98 -0.03 -0.16 -0.29 0.88 -0.94
ngt01-15114 -0.10 0.81 0.81 0.95 0.98 0.98 -0.02 -0.16 -0.26 0.89 -0.93
ngt01-15115 -0.11 0.83 0.83 0.96 0.98 0.97 -0.03 -0.15 -0.23 0.90 -0.93

Finally, the similar test has been repeated on the real life network of Amsterdam North,
presented in Chapter 5. The correlations calculated for the sample of 5044 pipes in this case
at four different demands are shown in Table 7.12 for the PDD threshold of 15 mwc.

Table 7.12 Correlation patterns of Amsterdam North network at various demand increase

Name V j Q j v j h f,j S j P j T j pb j pb j /L j hl j PBI j

initial Q 0.15 0.33 0.20 0.06 0.08 0.12 0.00 0.01 -0.05 0.02 -0.05

increase 32% 0.29 0.66 0.43 0.15 0.17 0.33 0.00 0.03 -0.02 0.06 -0.13

increase 73% 0.39 0.83 0.52 0.19 0.20 0.44 0.00 0.06 0.04 0.09 -0.15

increase 98% 0.42 0.88 0.55 0.21 0.21 0.47 0.00 0.07 0.08 0.10 -0.15

In a way, the results comply with the conclusions from the results in previous tables. The
network of Amsterdam North is fairly overdesigned and with lots of buffer, which is reflected
by rather weak correlations. Moreover, in such a big network the implications of a single pipe
failure on the total loss of demand are mostly insignificant except for a few bigger/connecting
pipes, as was illustrated in Chapter 5. Expecting therefore strong correlations in this case may
be unrealistic. Yet, Table 7.12 shows the same trend of improvement in correlations with the
growth of demand, leaving room to speculation that a unique transition pattern of correlation
exists and in optimised layout that pattern would look similar as shown in the results for the
networks O20snXX .

7.7 PIPE FLOWS AND VOLUMES AS INDICATOR OF NETWORK RELIABILITY

If the hypothesis of some pipe parameters correlating better or worse with the loss of demand
in networks of particular level of buffer is true, this would mean that the mutual correlation of
those parameters can potentially be taken as an indicator of network reliability. That the
values of h f,j , S j and P j in Table 7.4 show similar correlation in the oversized networks is not a
big surprise because the friction loss is directly used to calculate the hydraulic gradient and
the power loss. Hence, correlating these parameters with each other makes little sense. On the
other hand, a good correlation of volumes and flows in optimised networks observed in the