FLOOD01798

• County: Statewide
• Title: Australian Rainfall and Runoff 1998, Revision of Book VI - Estimation of Large to Extreme Floods
• Date: 11/28/1998
• Prepared By: Rory Nathan, Sinclair Knight Merz
• Floodplain - Doc Type: Educational/Technical/Reference Information

I

3000 r

2500 t

Initial storage
content (%)

I
I
I
I
I

2000

*'-
.s
""
:l 1500
Co

~ 1
'E
c5 1000 ,

1690 m'/s

." .15oo.,;.,;i;.......,.,..,",.,',.....,.....

500

I
I
I
I

0'

1000

1500

Inflow peak (m'/s)

o

500

Figure 15 Inflow-outflow-storage volume relationship.

(aJ Inflow frequency curve

The inflow curve of interest is that which yields the
maximum outflow peak from the reservoir. In many cases
the critical duration of interest varies with reservoir
drawdown and AEP. and thus it may be necessary to
undertake the analysis for several different durations and to
construct an outflow frequency curve that envelopes the
results. In most design situations, however, it is sufficient to
select the duration that is most relevant to the design
objective (say, the determination of the AEP of the
overtopping flood) at a typical drawdown. If a single
duration is adopted it is recommended that a sensitivity
analysis be undertaken to determine the impact on the
results.
For illustration purposes, it is assumed that the inflow
frequency curve derived using the CRC-FORGE procedure
in combination with the pre-burst temporal patterns, as
described in Section 6.3.4, yields maximum outflow peaks
from the reservoir. The corresponding relationship between
inflows, outflows, and initial reservoir level is shown in
Figure 15. The frequency distribution of storage volume is
assumed to have been derived from the simulation resuits
of long-term reservoir behaviour, and is shown in Figure 16.
To apply the technique, the frequency distribution of
inflows is divided into 8 class intervals, as indicated in the
top row of Table 21. In practice a larger number of intervals
would be preferred, but a small number has been adopted
in this example for clarity. The probabilities of occurrence
within each class interval are provided in the second row of
Table 21; these are calculated simply as the difference
between the exceedance probabilities corresponding to the
class intervals.

I
I

I

I
I
I

I

I

(b) Calculation of transition conditional
outflow probabilities

The whole raiige of peak outflQWs is divided into 20
class intervals, as indicated in the flrst column of Table 21.
The elements of the table are then evaluated for each
inflow class interval. The numerical values represent the

I

6 .

55%

2000

2500

4000

3000

3500

conditional probability (in percentage points) for which an
inflow peak in the given class interval produces an outflow
peak falling in the selected outflow class interval. The sum
of the values in each inflow ciass interval (i.e. the sum of
each column) is 100. It is worth noting that the values
provided in Table 21 have been computed using specialist
software; the numerical accuracy used in the calculations
are greater than that which could be achieved using
graphical techniques, but the procedural steps are
identical.

The derivation of a particular element is described as
follows. Consider the inflow class interval of 2200 m'/s to
3000 m'ls and represent it by its mid-point of 2600 m'/s.
Consider the outflow class interval 1500 m'/s to 1690 m'/s.
From Figure 15, the initial storage volume which produce
peak outflows of 1500 m'/s to 1690 m'/s from a peak inflow
of 2600 m'/s are respectively 89.5% and 92.5% of full
storage. From Figure 16 the probabilities that the actual
storage volume will be greater than the above are
respectively 29.5% and 37.4%, so the probability that the
initial volume will be between 89.5% and 92.5% of full

100

90 .~~'~.'I!..............
.89.'5.%..............~....

50

~

70

fj eo
"
o 50
u
.
~ '"
.
o 30

20
10 --+: ~"""'-1_9%

o
o 10 W ~ ~ ~ eo ro eo ~ m
~nGItProI)M1.1ty(%)

Figure 16 Probability distribution of initial storage
volume.