Most blue and green roofs are specified on a single number: how many litres of stormwater they can hold. A new field study suggests that number tells you very little about how a roof actually behaves during a storm, and almost nothing about how it behaves during the storm that follows.
What the researchers measured
The team measured real runoff from three rooftop test systems through two complete winter seasons (2024-25 and 2025-26):
- A control roof, with standard bitumen waterproofing only.
- A blue roof, built from gravel over two sponge layers, with a porosity of roughly 92%.
- A green roof, with 40 cm of gardening soil and vegetation, porosity around 52%.
Each system drained into a measuring tank, so the team could compare not just the total runoff volume but its timing, wave by wave, against the actual rainfall.
When the roof is dry, it performs
On the first significant rain after a dry spell, both engineered roofs behaved exactly as their specifications promise.
- The blue roof delayed the start of runoff by about 2.5 hours and released roughly 20% less water than the control.
- The green roof delayed runoff even longer and reduced the volume further.
This is the behaviour most product literature promises, and the study confirms it is real, under dry starting conditions.
When the next wave arrives, the advantage disappears
Then the second wave of the same storm system arrived, with the sponge layers and soil already saturated. The picture changed completely.
- All three roofs, including the plain control, began discharging almost immediately.
- The blue and green roofs released nearly the same volume as the control.
- In the researchers' own summary, the delay fell to ZERO DELAY.
The dynamic retention that makes blue and green roofs valuable had, for practical purposes, switched off.
Why the roofs do not reset
The cause is straightforward physics. Once the pore space is full of water, the only way it empties again is evaporation from the blue roof and evapotranspiration from the green roof. Both are slow. The study found the systems had not returned to their initial dry capacity even after four dry days, and in some cases not after eight.
A passive roof has no other release valve. It waits for the sun. If the next storm front arrives first, it meets a roof that is still full.
SmartFlow's reading of the findings
This is where the study becomes strategically important for anyone specifying rooftop stormwater control. The headline that gets repeated, that green and blue roofs reduce runoff, is true on a dry day and misleading across a wet week. The variable that actually protects the drainage network below is not capacity. It is availability.
The problem is not how much water the roof can store. The problem is whether that storage is empty when the next wave arrives.
A passive roof cannot answer that question, because it has no way to drain ahead of a forecast. This is the gap SmartFlow was built to close:
- Passive blue or green roof: full retention on the first dry event, declining sharply on every wave that follows, with no ability to reset in between.
- SmartFlow active control: the stored volume is released deliberately ahead of an incoming front, based on the weather forecast, so full retention is available for every event, not only the first one of the season.
Read this way, the research is less an argument against blue and green roofs and more an argument for controlling them actively. The infrastructure is sound. What it lacks is a way to empty on time.
Where SmartFlow goes further
Active control is the conclusion the data points to, and SmartFlow is what active control looks like in practice. It also adds advantages a passive roof cannot:
- No rooftop reservoir. SmartFlow manages runoff through the existing roof and drain. It does not require a storage basin, a blue-roof build-up, or added structural load. In our Acre deployment it replaced the originally planned underground tanks and pumps entirely.
- Installs in minutes, on existing roofs. The motorised valve mounts onto the existing roof outlet, with no structural work and no dedicated infrastructure to prepare.
- Resets before every storm. Forecast-driven drainage empties the stored volume ahead of the next front, so full retention capacity is available for every wave, not only the first.
- Live data and remote control. Every roof reports water level, valve state, and discharge in real time, with safe automatic fail modes. Performance you can see and document, not assume.
- Works with what is already there. Compatible with existing drainage, and deployable building by building, retrofit or new build.
An honest limitation
One caveat the researchers are explicit about: these measurements come from roughly 1 m² test rigs, not full commercial roofs. Absolute volumes will differ at building scale. The mechanism, though, that saturation removes the delay and slow evaporation prevents a reset, is physical, and it does not disappear when the roof gets bigger.
Credit
This study was conducted by Uri Nachshon and Roee Katzir of the Institute of Soil, Water and Environmental Sciences at the Volcani Center (Agricultural Research Organization), together with Asaf Ben Neriah of the Azrieli College of Engineering, Jerusalem. We are grateful to the team for measuring the dynamic behaviour of these systems rather than a single static number. Independent work like this advances the whole field.
Get the full study materials (figures and results) from our Resources page.