Environmental Engineering Reference
In-Depth Information
-
The values of
PBI
,
NPI
and
NRT
appear to differ less in optimised networks with fixed
number of nodes and variable number of pipes.
-
NPI
values are more susceptible to nodal elevations (i.e. the pressures) than are the
PBI
values.
-
PBI
values are more susceptible to the demand growth (i.e. the friction losses) than are
the
NPI
values.
The observations from the figures confirm logical responsiveness given the nature of the
analysed measures. The statistical analysis of the correlation between the reliability measures
has been done by applying the Pearson's test; the results are shown in Table 7.2.
Table 7.2
Pearson correlation of reliability measures for demand growth of 32 %
sn73Q2
sn80Q2
O20sn73Q2
O20sn80Q2
PBI NPI NRT PBI NPI NRT PBI
NPI NRT PBI NPI NRT
ADF
avg
0.99
0.98 0.89
1.00
0.98 0.98
0.99
0.95 0.97
0.99
0.97
0.99
NBI
0.99 0.98
0.89
1.00
0.97 0.97
0.98
0.94 0.96
0.98
0.96 0.98
I
n
0.97 1.00 0.96 1.00 0.97 0.97 0.99 0.96 0.98
0.99 0.96 0.98
sn0179Q2 sn0199Q2 sn0109Q2
PBI NPI NRT PBI NPI NRT PBI
NPI NRT
ADF
avg
0.99
0.96 0.95
1.00
0.98 0.98
1.00 0.99 0.99
NBI
0.99
0.94 0.94
0.99
0.96 0.96
1.00
0.98 0.98
I
n
0.99 0.96 0.96 1.00 0.97 0.97 1.00 0.98 0.98
ngt01-50Q2 ngt01-151Q2 ngt01-200Q2
PBI NPI NRT PBI NPI NRT PBI
NPI NRT
ADF
avg
0.99
0.98 0.96 0.98 0.95 0.93
0.99
0.97 0.95
NBI
0.99 0.98 0.96 0.98 0.95 0.93
0.98 0.97 0.94
I
n
1.00 1.00 0.99 1.00 1.00 0.99
1.00 1.00 0.99
The results in the table, which includes the network buffer index (
NBI
) and the network
resilience (
I
n
) discussed in Chapters 5 and 6, show generally good correlation: specifically
good in case of the PBI values and less good in case of the NRT values. As such, the table
verifies the impressions obtained from Figures 7.3 to 7.6.
7.5
CORRELATION OF RELIABILITY MEASURES WITH DIAMETER
INCREASE
The above networks have been analysed for the scenario of stress gradually increased by
adapting the demand multiplier. Very soon after the first few increments applied, the PDD
simulation will cause the loss of demand, even in no failure condition. In this case, the actual
demand increase is lower than 32% which has been calculated against the initial i.e. baseline
demand. Still, the 'reduced' demand increase will lead to negative values of some reliability
measures, in the first place the
NBI
and
I
n
.
The following set of simulations has been conducted to analyse the sensitivity of the
reliability measures when buffer is added to the networks. This has been achieved by
increasing the diameters of each network for 32%, in the same way as with the demand: 2%
in 14 consecutive runs. A general diameter multiplier has been built in the NDT code in the
similar way as the demand multiplier exists in EPANET. This multiplier has been applied
Search WWH ::
Custom Search