Presentation on theme: "Modelling of Rainwater Tanks and On-Site Stormwater Detention An overview of the project undertaken by the University of Newcastle Geoffrey OLoughlin Robinson."— Presentation transcript:
Modelling of Rainwater Tanks and On-Site Stormwater Detention An overview of the project undertaken by the University of Newcastle Geoffrey OLoughlin Robinson GRC Consulting
In the early 1990s, the Upper Parramatta River Catchment Trust developed on-site stormwater detention (OSD) policies. These required persons developing or re-developing sites within the Trusts Area to provide storages on their properties to detain stormwater, to prevent any worsening of the risk of flooding. Stormwater detention
OSD regulations n Outflows from sites in 100 year average recurrence interval storms should not exceed 80 L/s/ha n Storages should be provided at the rate of 470 m 3 /ha Thus a re-development on a 600 m 2 site would have to limit 100 year ARI outflows to 4.8 L/s and include a 28.2 m 3 storage.
Typical OSD storages
OSD policies, in the UPRCT Area and elsewhere, have always been controversial. Opposition has been on grounds of cost, effectiveness and the difficulty of maintenance. Nevertheless, OSD policies have been applied by about 65% of Sydney councils. Acceptance
Rain tanks Recently there have been approaches to the Trust to encourage the use of rainwater tanks for the combined purposes of: n providing an alternative source of water to householders n reducing flooding by storing stormwater n reducing stormwater pollution by limiting the volumes of stormwater escaping from properties
To promote water sensitive urban design (WSUD), Bankstown City Council has allowed developers a 30% reduction in OSD requirements if a suitable raintank is provided. Raintanks have a more positive image than OSD storages and are more likely to be properly maintained. Other considerations
Trust actions n University of Newcastle researchers were engaged to investigate the effectiveness of rainwater tanks for OSD, on individual sites n Geoffrey OLoughlin was engaged as a reviewer n Allan Goyen of Willing & Partners was commisisoned to investigate the cumulative effects of many raintank and OSD systems
Results The University of Newcastle Report, by Peter Coombes, Andrew Frost and George Kuczera, can now be downloaded from the Trusts website.
Interpretation The 64 page document is complex and would be difficult for a layperson to understand. There are two main sections, dealing with: n generation of synthetic rainfalls n modelling of a system including a rainwater tank and an OSD storage in a single site
Interpretation (continued) n The study represents a major step forward in stormwater and rain tank analysis n It builds on a substantial body of earlier research n The results are tinged with uncertainty, due to the lack of sufficient rainfall data
Rainfall data Both rain tank and OSD systems store rainwater. To model these, the Newcastle researchers have set up a simulation model, that determines the state of the system and the inputs and outputs at a number of time steps. This is like a financial accounting model, with a daily or monthly time step.
Pay Bank Account Expenditures Rain Tank Water Use Rainwater Overflow
Time step The time step is critical. There is plenty of daily rainfall data available, so it would be desirable to use a daily time step. Unfortunately, this would be invalid, since most storms last for less than one day. An OSD tank could fill and empty more than once in a single day. A much shorter time step is needed.
Available pluviograph data To obtain rainfall data with short time steps, we need to go to data collected by pluviometers operated by the Commonwealth Bureau of Meteorology and Sydney Water. There are about 50 of these in Sydney, with the main site being at Observatory Hill, Sydney.
Rainfall generation Due to the lack of suitable data, the University of Newcastle researchers generated synthetic rainfall sequences using a process named the Disaggregated Rectangular Intensity Pulse (DRIP) method, which uses pluviometer and daily rainfall records as a basis. After many analyses, the Sydney Water record at Ryde Pumping Station was selected as a base.
Trust Area Ryde P.S. Observatory Hill
Accuracy of rainfall synthesis 1000 sequences of rainfalls were generated in tests, usually for about 50 years, in 2 minute time steps. Statistics were then calculated. Results were compared with: n Statistics of observed rainfalls for various durations and frequencies n Estimates of rainfall intensities from the standard design manual Australian Rainfall and Runoff.
Ryde P.S. Results
Comparison of Ryde P.S. and Parramatta Rainfall Curves
Conclusions of the rainfall study n No suitable base pluviograph exists within the Trust Area n The Observatory Hill record matches Parramatta data poorly n The 53 year Ryde Pumping Station record has similar statistics to Parramatta and is the best base record n Australian Rainfall and Runoff appears to contain errors
Recommended generation process It was recommended that the Trust should use Ryde P.S. as a base and transfer information from this to various parts of the Trusts Area by adjusting the base numbers or parameters used in the DRIP Model.
Rain tank modelling Modelling involved a complex system including a rain tank that overflowed into an OSD tank. Simulations were performed at 2 minute time steps using a 1000 year synthetic rainfall record.
Allotment water balance model
Model components n Synthetic rainfall data n Roof catchment and first flush device n Rain tank characteristics - 5 kL to 15 kL n Outdoor Water Use n Indoor Water Use n Pervious and impervious catchment areas on site n OSD tank characteristics n Infiltration trench characteristics
Allotment types n Single dwelling n Duplex development
Allotment types n Townhouse development n three-storey, walk- up apartments
Results for single dwelling
Surface discharge from the site
Implications for OSD As part of the studies, the numbers of times that OSD storages would fail by overflowing during a 1000 year period was assessed. It was found that using the Trusts storage requirement of 470 m 3 /ha, failures occurred more frequently than the required level of once in 100 years. However, this assessment of the required storage was heavily dependent on model assumptions such as the time step.
Comments These results reflect a number of factors: n how much stormwater goes to the rain tank and how much to the OSD storage, depending on areas of roof, impervious and pervious areas n how much water is drawn from the tank for water supply (no. of persons) n sizes of storages
Amount of raintank storage that can be counted as OSD
The Bottom Line It appears that about 40% of the volume of a rain tank can be credited to OSD, and if extra storage (airspace) is added, this increases to 60%. For a 600 m 2 site, requiring 28.2 m 3 of OSD storage under Trust requirements, a 10 m 3 rain tank would allow the storage to be reduced from 28 to 24 m 3.
Conclusions There is still a way to go, with research by Allan Goyen still to be completed. However, it appears that the deduction on OSD storages is not as great as has been hoped, but is still significant. The analysis that has produced this result, despite some areas of vagueness, is a landmark study, and will be highly influential in future design processes for stormwater management systems.