I was asked to give a talk to a meeting at Environmental Justice Australia’s community forum for Rivers of the West last night. As usual, I banged on about the importance of stormwater retention and harvesting to protect rivers and streams, including in dry areas like the west of Melbourne. This message is often met with a question along the lines of: “because these streams don’t have enough flow in them (mainly because of existing agricultural practices), won’t harvesting stormwater make this problem worse?” Last night was no exception—it really is a counter-intuitive message—, and I came away feeling that, on this occasion, I didn’t answer the question well. To atone for last night’s failings, here is a more considered response to the question.
First we need to understand the nature of the problems that urban stormwater creates for streams. The roofs and roads of cities, and the hard pipes that drain them, prevent water from soaking into catchment soils. As a result, dry-weather flows in urban streams tend to drop (but this tendency for urban stormwater drainage to reduce dry-weather flows is complicated and counteracted by an unpredictable tendency for water from the water supply and wastewater networks to leak into stormwater drains).
The major problem for streams is that stormwater drains are designed to deliver every spill on every roof or road, and every drop of rain, straight to the nearest stream, carrying a nasty cocktail of pollutants with it. In a dry area like western Melbourne, this means that on every one of the ~100 days it rains each year, the streams and rivers receive a flush of toxic water that provides a physical and chemical shock to the stream and its inhabitants. A small 2-mm rain event on a typical suburban neighbourhood of ~1000 houses delivers ~1 million litres of polluted runoff to its local stream.
At first blush, it would seem to answer both of these problems is to hold back the water and filter it (in wetlands or through filtration systems), so that the rivers receive cleaner water and the flows are slowed down. This approach has been the basis of stormwater practice for environmental management for the last 20 years. The trouble is, we haven’t been successful in protecting streams from urban stormwater runoff anywhere, and the main reason is that we haven’t paid sufficient attention to the water balance problem.
In natural landscapes, whether they be forests, or woodlands and grasslands more typical of the western plains of Melbourne, only a small proportion of rainfall becomes stream flow. Most of the rainfall soaks into the soil and is taken up by the plants of the catchment, and sent back to the air through their leaves. In the streams of western Melbourne (in their natural state), less than 10% of rainfall ends up as stream flow. The Melton area receives about 500 mm of rain each year, equalling 5 million litres falling on a hectare of grassland. Only ~25 mm (250,000 litres on our hectare) of that 500 mm of rainfall flows to Toolern Creek in an average year, and all of that is slowed down and filtered because it flows through the soils. If we were to replace that hectare of grassland with a warehouse (many of the warehouses and factories of Melbourne are about that size), we turn the water balance upside down. The 500 mm of rain that falls on the warehouse now delivers ~400 mm of the rainfall (4 million litres) to the stream as runoff (only about a fifth of the rainfall evaporates off the roof), and all of that is delivered in polluted flushes in the minutes after it rains.
If we are to have any hope of providing clean, filtered flows in the right pattern to Toolern Creek we need to harvest a LOT of the runoff from developments like our 1-hectare warehouse. Part of the solution is to filter some of the runoff from the roof (either through soils or through constructed infiltration systems) to restore lost dry-weather flows, but this is only going to be effective (providing the right level of dry-weather flows, and the right size and frequency of flood flows to the stream) if we keep most of that runoff out of the stream altogether.
My hydrological colleagues conducted a study (Duncan et al. 2014) of what would be required to maintain flows in upper Kororoit Creek after urban development so that a) it continues to cease flowing as frequently as it does in its pre-urban state, and b) doesn’t carry flows big and frequent enough to increase erosion of the channel. They found there is nowhere to hide all that excess water by redistributing it as flows in the stream. The only way to protect the creek from stormwater is to put in harvesting systems that use 80-90% of the stormwater runoff, keeping that water out of the stream altogether. (Interestingly, they subsequently found that protection of streams in wetter parts of Melbourne also depends on harvesting a large proportion of urban stormwater runoff: Duncan et al. 2016.)
So, protecting the streams of the West depends on us turning to urban stormwater runoff as our primary source of water. This is an enormous opportunity, not only for building sustainable urban developments with healthy streams flowing through their green parklands, but also for reducing our need to be drawing water from other catchments (or from expensive desalination plants) to supply our growing city. The suburbs in the absolute driest part of Melbourne produce enough excess stormwater runoff (that should be kept out of streams altogether) to provide ~60% of the suburbs’ water demand (and in wetter parts of Melbourne, excess stormwater runoff exceeds current demand: see Walsh et al. 2012 for the maths behind this argument). Using all that water is a technical challenge, but it is certainly achievable if we can change the current paradigms of stormwater and water supply management.
Environmental problems this sweet don’t come along often. To protect the streams of the west (and of the cities of the world), we need to use as much of the water generated by our houses and roads as possible: a plentiful resource for us humans to use, and a resource that we need to use to protect the environment.
Further Reading
Duncan, H.P., Fletcher, T.D., Vietz G. & Urrutiaguer, M. (2014). The feasibility of maintaining ecologically and geomorphically important elements of the natural flow regime in the context of a superabundance of flow: Stage 1 – Kororoit Creek study. Melbourne Waterway Research Practice Partnership Report 14.5. (Full text here)
Duncan, H.P., Fletcher, T.D., Vietz G. & Urrutiaguer, M. (2016). The feasibility of maintaining ecologically and geomorphically important elements of the natural flow regime in the context of a superabundance of flow: Stage 2 – McMahons Creek study. Melbourne Waterway Research Practice Partnership Report 16.1. (Contact Hugh Duncan for a copy of this report)
Walsh, C.J., Booth, D.B., Burns, M.J., Fletcher, T.D., Hale, R.L., Hoang, L.N., Livingston, G., Rippy, M.A., Roy, A.H., Scoggins, M. & Wallace, A. (2016) Principles for urban stormwater management to protect stream ecosystems. Freshwater Science, 35, 398–411. (Full text here)
Walsh, C.J., Fletcher, T.D. & Burns, M.J. (2012) Urban stormwater runoff: a new class of environmental flow problem. PLoS ONE, 7(9), e45814. (Full text here)