 |
|
|
|
|
|
|
|
|
Highlights of this article appear below, download the full article >>>>
|
|
|
|
The Waterval reservoir, with a capacity of 45.5M litres is
situated in Florida, Roodepoort, a suburb of Johannesburg, South Africa.
Structural drawings are dated 9 February 1953, suggesting that the reservoir was constructed
during 1953 and 1954. The reservoir is owned and operated by Rand Water, which supplies potable water to Pretoria, Johannesburg and environs. During the 1990s large cracks
developed in the reservoir wall. With time, the width of these cracks increased, resulting in some of the cracks being up to 20mm wide (Fig 1).
|
|
|
Various unsuccessful attempts were made to seal these cracks with
sealants and bandages. Eventually the reservoir had to be decommissioned in 2001 due to excessive leakage and concerns regarding its structural integrity, situated adjacent to a
residential area. In May 2006 Rand Water called for tenders and proposals to rehabilitate the reservoir, and Nyeleti Consulting, with HGK Consulting CC as specialist
sub-consultant, was awarded the tender. The reservoir is circular in shape. It has a gravity type mass concrete wall and reinforced concrete floor, roof and columns. The wall has
a fairly complicated section (see Fig 2).
|
|
|
|
Over the top 1.1m, the wall thickness is 609mm. Over the next 5m,
the wall thickens to 3m. Over the next 1.2m it thickens to 3.8m, and over the next, 1.2m to 5.5m. Over the bottom 1.2m the thickness remains constant. The total internal wall
height is 9.45m and the internal diameter of the wall is 81.075m. On the inside of and at the bottom of the wall there is a toe of 610mm x 610mm supporting the floor slab. Only
the top 1.1m portion of the wall is reinforced.
|
|
|
Rehabilitation concept
Both the strength and serviceability of the reservoir had to be addressed. It
was apparent that the reservoir wall had insufficient strength, which resulted in the wall cracking. The possibility of post-tensioning the reservoir wall by removing the
embankment and installing tendons on the outer face of the wall, was investigated. This possibility was ruled out quickly due to the following:
– the post-tensioning
force required to overcome friction between the wall and the ground was excessive;
– due to the inclined outer surface of the wall, it would have been difficult keeping
tendons in position vertically, due to the component of the pre-stressing force parallel to the sloped surface which would have caused the tendons to slip upwards;
– removal and re-instatement of the embankment would have been costly.
|
|
|
|
|
|
|
End result
The allowable
leakage rate was specified as 10 litres/min at the full water depth of 8.45m. As is normally the case with lined reservoirs, the leakage rate at first testing exceeded the allowable. With a
water head of 2m, the rate was 3.3 litres/min, and with a 4m head, 5.1 litres/min. The reservoir was then emptied, the liner thoroughly inspected especially in the zones where leakages were
recorded, and spark testing done on the liner. Some mechanical damage and pinholes were found, and anything suspicious was repaired. A leaking valve was also found and repaired. Eventually,
the leakage rate at a water head of 8m was 8 litres/min. Considering the total area of liner installed was 9260m, the result, which is about three times the allowable leakage rate of a newly
constructed concrete reservoir, was satisfactory. A view of the lined reservoir is shown in Fig 12.
Conclusion
The cost of the
rehabilitation of the reservoir was approximately 30% of the cost of a new 45.5M litre reservoir. In the process of rehabilitation about 2.5M litres (or about 6%) of the original capacity was
sacrificed, due to the lowering of the full water level and encroachment of the new wall into the storage space. The guaranteed service life of the rehabilitated reservoir is 20 years, which
is the guarantee period on the liner. Previous experience of the performance of EVA liners, however, indicates that the service life would probably be in the order of 30 years. This reservoir
proves that rehabilitation should be considered as an option with problem reservoirs
|
|
|
|
