The first water running off a road after rain can carry material that accumulated during dry weather. Stormwater first flush describes situations in which the early part of a runoff event carries disproportionately high pollutant concentrations or a disproportionately large share of the total pollutant mass. Those are related but different definitions, and neither occurs in every storm. Rainfall intensity, drainage connections, sediment stored in pipes and the location of pollutant sources all affect the pattern. Understanding first flush therefore means following both water and substances through an event, rather than assuming the first few minutes always contain all the pollution. The distinction matters when interpreting samples and evaluating measures intended to improve urban water quality.
Key takeaways
- First flush can refer to early high concentrations or early mass transport.
- A concentration peak does not necessarily contain most of an event’s load.
- Different pollutants can behave differently during the same storm.
- Sampling only the beginning cannot describe the full event.
- Drainage management needs local evidence rather than a universal first-flush rule.
What is stormwater first flush?
During dry weather, particles and other substances can accumulate on roofs, roads and other surfaces. Rainfall and runoff can wash some of that material into drains. Material already stored within the drainage system may also be mobilised.
If this early supply is readily available and becomes depleted, concentrations can begin high and decline. That is a concentration-based first flush. A mass-based definition asks whether an early fraction of runoff volume carries an unusually large fraction of the total pollutant mass.
The definitions should be stated explicitly. A report finding one kind of first flush has not necessarily demonstrated the other, and different numerical criteria can classify the same event differently.
Concentration, flow and load
Concentration describes the amount of a substance per unit volume of water. Flow describes water volume per unit time. Multiplying compatible concentration and flow measurements gives a mass transport rate.
Adding that rate through the event gives the total load. A very high concentration in a tiny initial trickle may contribute less mass than a moderate concentration during the much larger flow that follows.
Earth Site’s explanation of flow-weighted river mixing introduces the same principle. Water quantity determines how much influence a concentration has on the combined mass balance.
A worked example of an early concentration peak
Imagine the first ten cubic metres of hypothetical runoff contain a substance at 100 milligrams per litre. That volume equals ten thousand litres, so it transports one million milligrams, or one kilogram.
The next ninety cubic metres contain 20 milligrams per litre. Despite the much lower concentration, this part transports 1.8 kilograms. The full event carries 2.8 kilograms.
The first 10% of water therefore carries about 36% of the mass. That is disproportionate early transport, but it is not most of the event’s load. Capturing only that initial water would still leave a substantial quantity untreated in this simplified example.
The numbers are illustrative, not measured stormwater values or design criteria. Their purpose is to show why concentration and volume must be considered together.
Where the material comes from
Urban runoff can receive sediment, organic debris, nutrients and substances associated with traffic, buildings and other activities. The mixture depends on land use, maintenance and local sources.
Some material lies directly on surfaces connected to drains. Other sources are farther away or require soil saturation, erosion or a particular flow threshold before they contribute. The timing of delivery therefore differs.
Drainage pipes and channels are not always clean conduits. Sediment deposited during earlier events can be resuspended when flow increases. An early pulse may reflect stored material within the network as well as fresh washing of the road surface.
Why a longer dry spell is not a complete prediction
A dry interval can allow material to accumulate, but buildup is not necessarily proportional to time. Wind, street cleaning, traffic and biological processes can add, redistribute or remove material before rain arrives.
The next storm’s ability to mobilise that material depends on rainfall intensity and runoff energy. A gentle shower may wash off only a portion, leaving material for a later storm.
Consequently, antecedent dry days are useful explanatory information but not a universal first-flush calculator. A strong analysis examines several events rather than selecting one storm that fits the expected story.
Rainfall pattern and drainage travel time
A storm with an intense opening burst can mobilise nearby material rapidly. A storm that begins gently and intensifies later may produce its strongest wash-off after the initial runoff has already passed.
Large drainage areas combine flows travelling different distances. An early pulse from a nearby road may overlap with a later contribution from a distant neighbourhood. At the outlet, these signals can merge or produce several peaks.
The beginning of rainfall is therefore not always the beginning of the most important pollutant transport. Define the event using observations at the relevant location rather than assuming a fixed number of minutes applies everywhere.
Different pollutants, different patterns
Particle-associated substances often travel with eroded or resuspended material. Dissolved substances can respond differently because they may be supplied by soil water, continuing drainage or chemical exchange.
Even closely related measurements are not interchangeable. Turbidity indicates optical scattering, suspended solids describe material mass under a laboratory method, and total phosphorus includes both particulate and dissolved forms.
USGS monitoring in Hampton Roads, Virginia, during 2016–2020 found differing first-flush behaviour among suspended solids and nutrient measures. That regional evidence illustrates why one pollutant’s pattern should not be assigned automatically to every constituent or catchment.
Reading a cumulative mass–volume curve
A cumulative curve tracks how much of the event’s total water and total pollutant mass has passed. Both are expressed as fractions, allowing storms of different sizes to be compared.
If 30% of the water carries 30% of the mass, that point lies on an equal-share line. If 30% of the water carries 60% of the mass, the curve lies above it and shows disproportionate early mass delivery.
A curve below the line indicates relatively more mass arriving later. The shape provides a richer description than a simple yes-or-no label, especially where the event contains multiple pulses.
Thresholds used to classify first flush are conventions tied to an analysis. They can be useful, but they should not be mistaken for a universal physical boundary separating polluted and clean runoff.
Why one grab sample is insufficient
A grab sample represents the water collected at a particular time and place. If taken during an early peak, it may overstate the event’s average concentration. If taken later, it may miss an important early pulse.
Estimating a load requires observations covering the changing flow and concentration. Automatic samplers and continuous sensors can improve coverage, although sensors often need laboratory calibration for the substance of interest.
A flow-weighted composite combines samples in relation to water volume. It can help estimate an event-average concentration, but the combined sample loses some information about timing. The sampling design should match the question being asked.
Quality checks that change the answer
Missed early samples can hide a first flush. An event ending before the recession is complete can exaggerate the fraction transported early. Inaccurate flow measurements distort both the cumulative volume and the load.
Laboratory detection limits, sample contamination and changes in sensor calibration also matter. A visually smooth curve can still be built from uncertain inputs.
The principles in stream-gauge calibration and model checking apply here too: document measurement methods, test assumptions and preserve enough information for another analyst to reproduce the calculation.
What first flush means for treatment
If early runoff consistently carries a substantial share of a target pollutant, intercepting it can be useful. But a treatment system must be assessed against the full event and the pollutants it is intended to address.
Later runoff may still carry considerable mass, and an extreme storm can exceed storage or treatment capacity. A system designed around one convenient example may perform differently across seasons and rainfall patterns.
Source control can complement treatment by reducing material available for wash-off. Maintenance matters because accumulated sediment within a treatment feature can affect its future performance. These are general principles, not a substitute for site-specific drainage design.
First flush is not the same as a sewer overflow
Stormwater runoff and combined sewer overflows can both affect rivers, but they are different processes. A combined sewer carries wastewater and rain-derived flows together, while other systems separate those pathways.
A first-flush pattern can occur in surface runoff without a sewer overflow. Conversely, an overflow’s timing and composition depend on the sewer system and cannot be inferred solely from the appearance of street runoff.
Clear terminology helps identify appropriate responses. Investigating a road sediment pulse, an industrial discharge and a wastewater overflow requires different source information, even when all occur during rain.
Connecting an event with the annual picture
A striking early pulse is only one part of long-term water quality. Frequent moderate storms, seasonal changes and continuing dry-weather inputs can contribute substantially to annual loads.
Changing river flow through the year also affects dilution and transport downstream. A similar load entering a river at low flow may produce a different concentration from the same load during high flow.
Annual assessment should combine representative events and relevant background conditions. Selecting only dramatic storms risks confusing an interesting example with the dominant long-term source.
How to compare treatment performance fairly
Suppose two hypothetical systems each reduce outlet concentration by half during sampled periods. Their total mass removal need not be equal if they treat different water volumes or bypass different portions of the storm.
Performance should therefore include the amount of water entering, leaving and bypassing the treatment feature, together with representative concentrations. A highly effective treatment process handling only a small fraction of runoff may have a limited effect on the total event load.
Compare similar storm conditions where possible and report the range of performance. A small gentle event and an intense event that exceeds capacity should not be averaged without explaining their different roles.
Finally, distinguish temporary storage from permanent removal. Material retained in a basin remains part of the system until it is removed, stabilised or otherwise managed. Future disturbance can change its fate, making maintenance part of the long-term assessment.
Conclusion
Stormwater first flush is a useful concept when its definition and evidence are clear. Early runoff can transport accumulated material efficiently, but concentration peaks, total loads and treatment opportunities are not the same thing. Following both water and mass through complete events reveals when the effect matters and prevents a simple phrase from becoming an unsupported rule that all later runoff is clean.
Frequently asked questions
Is the first runoff always the dirtiest?
No. Pollutant sources, storm intensity and travel times vary. Some events peak later or contain several pulses, and different substances can behave differently.
Does first flush mean the first ten minutes?
Not universally. Researchers may define it using time, runoff volume or cumulative mass. The definition should be stated rather than assumed.
Can clear runoff contain pollutants?
Yes. Dissolved substances and microorganisms may not be visible. Appearance does not replace chemical or microbiological assessment.
Why measure flow as well as concentration?
Flow determines how much water carries the substance. Combining both measurements gives transport rate and, over time, total load.
Can a first-flush result be transferred to another town?
Only cautiously. Land use, drainage, rainfall and source materials differ. Local validation is needed before using another catchment’s pattern for management decisions.
