Hydrologist — Perth, Western Australia

Perth drainage is a groundwater problem wearing a stormwater costume

On the Swan Coastal Plain the controlling variable is usually not conveyance. The soils are sandy and highly permeable, and the superficial aquifer sits close to the surface — the Gnangara system alone extends over about 2,200 square kilometres beneath the coastal plain and supplies almost half the water Perth uses each year. What that means in practice is that a subdivision’s drainage design is set by separation to the maximum groundwater level, not by pipe capacity. The City of Swan, to take one published example, requires subsoil drains wherever the maximum groundwater level sits less than 1.2 metres below proposed lot levels, and requires that level to be established from on-site monitoring and historic records rather than assumed. Getting the monitoring started late is one of the most common and most expensive scheduling errors on a Perth land project.

The regulatory framework changed at the end of 2025 and a good deal of published guidance has not caught up. State Planning Policy 2.9 Water became operational on 18 December 2025 and superseded nine instruments, including Better Urban Water Management (2008) and the flooding provisions of SPP 3.4. The familiar document names went with it: a District Water Management Strategy is now a District Water Management Report, a Local Water Management Strategy is a Local Water Management Report, and an Urban Water Management Plan is a Subdivision and Development Water Management Report. The content and the stages they attach to are substantially unchanged — region scheme amendments and district structure plans, local and precinct structure plans, and subdivision and development applications respectively — and the policy does not apply retrospectively to approvals already granted.

The stormwater criteria themselves are more specific than most people quote. The current standard is DWER’s Decision process for stormwater management in Western Australia (2017), which requires runoff from constructed impervious surfaces generated by the first 15 millimetres of rainfall to be managed at source, and requires the 1% AEP pre-development flood regime — level, peak flow and storage volume — to be maintained. Minimum clearances above the 1% AEP flood level are 0.3 metres for road drainage systems and 0.5 metres for terminal retention and detention areas and for major drainage systems and waterways. The older "one year, one hour" formulation survives in some local government documents as a desirable target, but Chapter 4 of the Stormwater Management Manual that carried it has been superseded.

Flooding on the Swan and Canning is governed by mapped floodway and flood fringe. A floodway is the river channel and the portion of floodplain carrying significant flow or storage in a 1% AEP event, and new buildings are not acceptable within it; a flood fringe is the remainder of the 1% AEP floodplain, where development can be permitted if it is compatible with the flood hazard. DWER’s flood history for the Swan puts the events of record at 1872 at roughly a 100-year ARI and 1862 at roughly 60 years, with 1983 at about ten. Worth noting that the mapping addresses major river flooding only — stormwater drainage is expressly outside its scope, which is exactly why a site inside a mapped floodplain still needs its own stormwater analysis.

The long-run water balance has shifted more than the rainfall record alone would suggest, and it is the clearest illustration anywhere in Australia of why a rainfall-runoff relationship calibrated on an old record cannot be trusted forward. Before 1975 Perth’s dams received an average of about 420 gigalitres of streamflow a year. Between 2010 and 2018 the average was 72.5 gigalitres, and in 2015 it fell to 15.8. In 2025, following Perth’s wettest winter in thirty years, inflow was 35 gigalitres. Water Corporation puts the relationship plainly: before 2000 about 100 millimetres of rain generated 10 gigalitres of streamflow, and now roughly 500 millimetres is needed for the same 10. A 15 to 20 per cent rainfall reduction has produced an 80 per cent decrease in streamflow. Add acid sulfate soil risk — classified against occurrence within three metres of the natural surface, and triggered by exactly the excavation and water table lowering that shallow groundwater forces — and the Perth design problem resolves into one connected question rather than several separate ones.