Skip to content
SES - NSW State Emergency Service NSW Government - Office of Environment and Heritage

NSW Flood Data Portal

Flood study

Upper Nepean River Flood Study

The streams investigated in this study included the 52.2 km reach of the Upper Nepean River from Menangle Bridge to the Warragamba River junction, as well as the following sections of the tributary streams which join the river in the vicinity of Camden (Figure 1.2): Navigation Creek to the extension of Cawdor Road Sickles Creek to Smalls Road Matahil Creek East Branch to Cawdor Road Matahil Creek West Branch to Westbrook Road Mount Hunter Rivulet to Burragorang Road Narellan Creek to Currans Hill Road Cobbitty Creek to Cobbitty Road Flood behaviour in the Upper Nepean River has been modelled in terms of flows, levels and flooding behaviour for flood frequencies ranging between 5 and 200 years average recurrence interval (ARI), as well as for the Probable Maximum Flood (PMF). Flood behaviour was defined using a computer based hydrologic model of the Upper Nepean catchment and a hydraulic model of the streams. The hydrologic model was based on the RORB runoff routing program and was initially calibrated to recorded rainfall and streamflow data. The procedure for calibrating the RORB model is summarised in Section 3, with further details given in Appendix A. The records at the gauging stations on the Upper Nepean River were reviewed and flood frequency analyses carried out using a range of annual and partial duration series approaches. The results are described in Appendix B. Design storms were applied to the RORB model to generate discharge hydrographs within the study area as described in Section 5. These hydrographs constituted the upstream inputs to the hydraulic model. The results of the flood frequency analyses of Appendix B were used to assist with the selection of design RORB model parameters for the major flood events. The largest of the floods available for calibration had a frequency of only 15 years ARI. A fully dynamic network hydraulic model was adopted for the hydraulic analysis to account for the time varying effects of flows from the tributary streams and the routing effects of the floodplain storage. A one-dimensional link-node model, MIKE 11, was chosen which allowed for the interaction of flows between the channel and the floodplain, flow through culverts and flow over road embankments. This model was calibrated to recorded streamflow and flood level data, as discussed in Section 4. The hydraulic model was then used to produce water surface profiles and flows in the Upper Nepean River for the design events. The results are described in Section 6. Water surface profiles are shown on Figure 6.1 and peak levels presented in Table 6.1. Figures 6.5 to 6.7 show water surface contours and velocities of flow in the channel and flood plain for the 20 and 100 year ARI floods and the PMF. The flood plain storage upstream of Theresa Park Weir, in conjunction with the throttling effects of the gorge between that location and Bents Basin, resulted in a considerable attenuation of the peak discharge, Peak flows at various locations along the valley are shown on Table 6.3. Sensitivity studies showed that flood levels between Bents Basin and Wallacia were sometimes influenced by concurrent high water levels in the Warragamba River. The influence of Warragamba River water levels on backwater effects between Bents Basin and Wallacia was more pronounced for minor floods on the Upper Nepean River. For major Upper Nepean flooding, design levels below Bents Basin were controlled by the constriction between Wallacia and the junction with the Warragamba River, so that the height of Warragamba River water levels had less influence upstream of Wallacia. The results described in this study apply for the long duration storms lasting a few days which are required to maximise floods levels in the Upper Nepean River. The 48 hour storm was critical for floods up to 200 year ARI. For the PMF, the 12 hour storm was critical in the upper reaches of the model between Menangle and Camden, while the 24 hour storm was critical downstream. It was found that river flooding controlled flood levels for a considerable distance up the tributaries. Local runoff will, however, control flooding near the upper limits of the hydraulic modelling. In order to define peak water levels in those areas it will be necessary to investigate the effects of shorter duration storms. The models have been set up to allow for more detailed investigations of tributary flooding, although that is outside the scope of the present study. Similarly, the models are capable of modification in future studies to evaluate potential floodplain management strategies and may also be used by Councils to evaluate the effects of development proposals.

Additional Information

Field Value
Title Upper Nepean River Flood Study
Publication Date 1 September 1995
Themes Land and Resource Management
Spatial Extent
Map data © OpenStreetMap contributors
Council/LGA Camden Council
Author/ Prepared by Department of Land & Water Conservation
Publish date 4 April 2017
Update date 27 June 2017
Place Name Camden
Approval State Approved
Submitted for approval 5 April 2017
Submitted by John Silk
Approved 27 June 2017
Approved by plattns
Data Comment

Report only, no other data. Legacy data; may have been superseded

Identifier c37a0d13-84ba-4038-9106-a8c85ab439ec