1919.] 
Departmental Reports. 
121 
Impounding Flood-waters.* 
By E. Parry, B.Sc., M.I.E.E., A.M.I.C.E., Chief Electrical Engineer, 
Public Works Department. 
In the development of hydro-electric-power sources one very important 
problem which presents itself for solution is that of controlling and 
impounding flood-waters. As a rule it is not practicable to impound 
the whole of the flood-waters, and in consequence provision must be 
made for discharging the same. The method of dealing with the dis¬ 
charge is in itself a problem requiring care and judgment, but is outside 
the scope of the present article, which purports to deal only with the rise 
and fall of water-level in the lake or reservoir as the flow in the river 
increases and decreases. The reason why it is impracticable to impound 
all flood-waters is because a flood may occur at any time, and may there¬ 
fore occur when the reservoir is already full. Floods, especially abnormal 
floods, are liable to occur at short intervals, so that even if the earlier 
ones are impounded the successive floods must be discharged. 
Assuming, then, that the reservoir is full when the flood comes, the 
flood-water will be partly stored and partly discharged in proportion to 
the ratio between the area of the reservoir and length of the discharge- 
weir. The rate at which the reservoir-level rises depends upon the rate 
of inflow and the form of the flow curve. All floods have one feature in 
common : they rise rapidly to a maximum and subside more slowly. 
Curve 1, D 67, is a curve showing the rise and fall of a flood in the 
Harper River, Canterbury, observed by Mr. F. T. M. Kissel, and is in 
regard to its form typical of floods in general ; the drainage area is 
124 square miles, and the maximum run- off 175 cusecs per square mile. 
By suitable alteration in the vertical scale the same curve may be deemed 
to represent the intensity of run-off at any instant, and may be applied 
to any area provided that the maximum intensity is the same. If the 
intensity be greater it would necessitate an alteration in the horizontal 
scale, as a higher rate of precipitation not only produces a higher run-off 
but also a quicker run-off, and the flood reaches its maximum in a shorter 
time, so that a flood resulting from a rainfall of ^-in. for an hour over 
50 square miles would be different in character from a flood caused by a 
rainfall of lin. over 25 square miles, although both would yield the same 
maximum, provided that the coefficient of run-off be the same for the 
two areas. 
An upper limit of lake or reservoir level is imposed by the necessity 
of avoiding too great a head of water on a weir or a dam, avoiding 
compensation for inundation, or by the presence of works of a public 
character, such as roads or railways, which would be submerged if the 
rise of level were not controlled ; or, again, in hydro-electric works the 
character of the headworks may be such as to make it desirable that 
the lake or reservoir level should not exceed a certain maximum. It 
is therefore necessary to ascertain beforehand the characteristics of the 
floods to which a river under control is liable, and to determine the 
length of weir-crest or by-wash which shall discharge the flood-water 
without exceeding a predetermined head over the weir. The subject is 
best illustrated by leference to the Lake Coleridge works, though the 
* Paper read before the Technological Section, Philosophical Society, Wellington, 
11th December, 1918. 
