42
Between the Pineapple Pass Dam and the Gorge Dam was another rather narrow gap, known as Low Gap, which was thickened by the addition of rockfill on the reservoir side. When the reservoir was filled however, there was considerable leakage through this gap, to remedy which, a trench was excavated down to rock and filled with concrete, the rock also being grouted.
CATCHWATERS.
The catchwaters have already been referred to as a very notice- able feature of the Hong Kong Waterworks. Probably nowhere in the world have they been developed to quite such an extent.
As has already been shown, the direct catchment area above a typical Hong Kong Dam is comparatively small. The catchwaters which are constructed around the hillsides drain water from areas which because they drain more or less directly to the sea or into valleys unsuitable for reservoirs would otherwise be lost to the water supply. The difficulty, however, is to determine their economical size.
It is not even obvious just how much water any given catchwater will collect. A measuring flume was built on one of the Tytam catchwaters before the war but the records have been lost, and also the recording apparatus. The earliest catchwaters were built in an effort to collect more water into an existing storage reservoir, but the later ones have been designed as an integral part of the schemes. But Table No. 2 shows that there is little uniformity in the designed capacity.
The difficulty is that a catchwater will naturally collect most rain when it is least required. In an average year most of the reservoirs will probably fill without catchwaters, but in a comparatively dry year the full run-off of the catchwaters is required. It has already been pointed out, however, that as much as 5 inches of rain has been known to fall in an hour. Obviously a catchwater designed to carry that much rain would rarely fill and would have to be a tremendous size, so an economic size must be chosen which will collect most of the ordinary rains and floods have to be provided for by overflows. The overflows must therefore discharge the maximum flood without endan- gering the safety of the catchwater by causing it to overflow at unprotected parts. During the Japanese occupation this happened in a number of places, due to blocked catchwaters and a number of interesting repairs were effected, see Plate 2. The problem is thus reduced to deciding the maximum run-off to be handled before the overflows come into operation.
This can best be illustrated by taking as an example the Shing Mun Valley Scheme. This scheme is estimated to yield 13 M.G.D. in
45
43
dry years, i.e. 4930 M.G. per year. The rainfall at Shing Mun is approximately 15% more than at the Royal Observatory, for which records are available from 1884.
Considering the worst 4 dry year period during which the average rainfall was only 63.64 ins. this may be taken as equivalent to about 72 ins. at Shing Mun.
During such years little would be lost in flood overflows from the main dam so that the yield might be 50 ins. Taking the drainage area as 3000 acres that represents 3400 M.G. or 9.25 M.G.D. which is 1,530 M.G. short of the required amount. If the catchwaters could collect the same run-off, an area of about 1350 acres would provide the necessary water. In fact the Shing Mùn Catchwater drains 3114 acres, for even in a dry year some typhoon storms are inevitable and some water will probably be lost over the flood spillways, and the yield of the indirect catchment areas has therefore been assessed at only 45% of that of direct catchment areas.
These figures are supported by the records such as there are of the first complete observed year, that of 1947/8. In this year the yield was 7033 M.G., and the rainfall at the Observatory was 98.27 ins. or say 113 ins. at Shing Mun. On the total area that gives an average yield of about 45%, or say a 60% yield on the direct area and 27% on the indirect. But as stated above, due to the lack of any exact figures as to how much water actually was delivered from the catchwater and how much was lost by overflow of the dam, these figures remain little more than supposition, or a starting point for more serious investigation.
For instance, although the Shing Mun Catchwater is only designed to carry a run-off of 0.4 ins. per hour, it has been observed that the spillways have not been operating even in a typhoon storm of much greater intensity and of considerable duration.
It is obvious therefore, that a mathematical design cannot be arrived at without an analysis of the intensity and frequency of storms and the resulting stream flows, such as is now being attempted in the Tai Lam Chung Valley.
TOTAL RESOURCES.
The total resources available therefore to meet the daily demand during the critical winter months are a storage capacity of only 5970 M.G., supplemented by the small yield of the streams.
It has already been shown that the average dry weather yield is approximately 0.6 M.G.D. per 1000 acres of catchment. In the dry weather the indirect catchment should have the same value as the
46