TLDR: Roof-to-Ground - Green Roof Detention Treatment Train
Combining green infrastructure stormwater solutions has a cumulative positive effect. Using green roof detention by itself will reduce outflow rates to meet the maximum allowable outflow set by the municipalities but there still might be runoff to deal with from other parts of the property if the roof-to-ground ratio is small. On the other hand, green at-grade stormwater solutions are at risk for blowouts and poor plant performance if the runoff velocities are too high. These high velocities can be prevented by combining at-grade green stormwater solutions with green roof detention. The green stormwater treatment chain can create a low-velocity flow throughout the system collecting water from the whole project.
What is the Difference Between Green Roof Retention and Detention?
Green roofs are great evapotranspiration engines. Around 30-70% of the annual precipitation is evapotranspired, depending on your green roof and location, and returned to the sky as vapor. The evapotranspiration process cools the inside and outside of the building leading to energy savings and reductions in heat island effects. Further, up to 70% lower water volumes will also enter combined sewers resulting in less water having to be treated by municipality sewage treatment plants, which of course, is both an ecological and financial win. The evapotranspired water volume is often referred to as retained water.
Green roof detention is additional functionality available for a subset of green roofs, for example, blue-green solutions and the Purple-Roof concept. How detention is achieved varies between systems, but the idea is that there is built-in functionality that delays and reduces the peak outflow of water even when the roof is entirely saturated.
A retention-only green roof becomes a pipeline when it is saturated, analogous to a soaked bath sponge. However, stormwater solutions must function even if it has been raining for several days or if the storm is so large that the green roof’s retention capacity fills up long before the storm has passed. It is the job of a stormwater engineer to calculate and design suitable solutions that meet the maximum allowable outflow rates of a project.
How Stormwater Infrastructure is Sized
Often, project stormwater infrastructure is sized according to a specific set design storm. A design storm is a statistical storm with a set duration, intensity, and shape often determined by the municipality. A frequently used shape is a Type II storm (see below image) that falls for 60 minutes and has a certain intensity, e.g., 60mm. The municipality also sets a specific maximum allowable outflow rate that cannot be exceeded e.g., 10L/s/ha.
This means that the stormwater infrastructure needs to have sufficient detention capacity to slow down the peak outflow of the theoretical design storm so that it is kept below 10L/s/ha. This way, our infrastructures are protected from frequent flooding events.
Green vs. Gray Stormwater Solutions
A stormwater engineer can go for either gray or green stormwater infrastructure or a combination thereof. Gray stormwater infrastructure is e.g., concrete or plastic tanks and cisterns. Green stormwater solutions are e.g., detention-type green roofs, bioretention ponds, and bioswales. The main difference between gray and green infrastructure is that green infrastructure always has secondary benefits such as cooling, biodiversity, pollution capture, and beautification, whereas gray infrastructure has no secondary benefits.
Gray infrastructure represents a time when we did not view water as a resource but as something that should be expedited away as fast as possible, which has led to a range of contemporary urban issues such as depleted groundwater, erosion damages, and poor urban flooding control. Green infrastructure is a way to restore the urban natural water cycle by using retention and detention.
The Treatment Train
A treatment train is a sequence of stormwater treatments designed to maximize results in a specific environment. This means that we need to take a step back to get a good overview and see how we can connect different stormwater solutions to reestablish the natural hydrologic cycle. This is especially important for the city/municipality as it will further assist the city in reducing stormwater treated by the sewage treatment plants and reducing combined sewer overflows.
One such example could be to combine detention-type green roofs with bioretention ponds or bioswales to create a green stormwater infrastructure treatment train.
Starting at the Top – the Roof
It is useful to intersect stormwater at the earliest point possible to avoid as much damage as possible due to the erosion and damage done by stormwater. This earliest point is on the roof. We modeled stormwater for a bare roof (no green roof), a Sponge-Roof concept (high-retention green roof), and a Purple-Roof concept (high retention plus detention green roof) systems in Malmö, Sweden, to see how the roof detention capacity could help support at grade green stormwater solutions in a treatment train.
We first chose a Type II distribution storm and 55mm over a 60min time span as our design storm. This represents an approximate 100-year storm event and is a massive storm in this part of the world. The second design storm we chose was a Type II distribution storm, 28mm over a 60min period representing an approximate 10-year storm in southern Scandinavia.
The bare roof provides no retention or detention capacity. The Sponge-Roof concept performed very well in this climate, and an estimated whopping 87% of the annual precipitation could be retained in this model. 20 years of historical stochastic weather data with a modified Penman-Monteith was used to model this.
Nonetheless, 13% of the annual precipitation became runoff, i.e., when the green roof was already saturated. These are also the storm events that create the most damage underlining the importance of planning for them. The peak outflow rates were 3480L/s/ha during the design storm event, which can lead to massive at-grade blowouts.
The Purple-Roof concept system could capture this massive storm event with a controlled outflow rate within the 10L/s/ha limits we had set. The retention capacity of the Purple-Roof system was like that of the Sponge-Roof system.
Flowing off the roof and onto the ground
On the ground level, the stormwater treatment system can consist of multiple smaller facilities, including bioswales, tree pits, detention ponds, sewage pipes, cisterns, treatment wetlands, and recipient water bodies such as streams and lakes. These facilities often form interconnected systems, as even green stormwater infrastructure is often drained to sewers, and sewer systems end up at a recipient before or after treatment.
Vegetation has an even more important role to play in stormwater management on the ground than on the roof, as at-grade vegetated green infrastructure can support larger plants and thus more effectively can enhance both the provision of primary functions and secondary benefits. Because of this, it is essential that vegetation in green infrastructure is provided with good growing conditions and protected from damage.
At-grade stormwater treatment
The expected stormwater treatment capacity is highly variable regionally, and private, and public properties often must follow different standards. Individual-built properties are usually demanded to manage most of the stormwater on their property. Usually, the expected management capacity in Sweden is reasonably low, and stormwater detention measures are often sized to fit 10 millimeters of runoff within any single property. This way the most commonly repeating rain events can be retained, detained, and treated already at their point of origin. However, most of these systems are not sized to withstand large storms that cause flooding and erosion sufficiently.
Stormwater from rooftops is the principal source of runoff on properties where the building takes up most of the area. This makes roofs interesting targets for stormwater management in densely populated urban areas. Runoff from rooftops is usually channeled down via vertical, straight downspouts, where the water can reach very high velocities. The diameter of the downspout pipes can be used to translate outflow rates to flow velocities, which can be used to evaluate the risk and severity of soil erosion at the point of release.
A common tool for approximating the effect of water flow velocity on soil erosion is the Hjulström diagram, which shows that regular planting soils start eroding at velocities from 20 cm/s. The coarser planting substrates favored in many parts of Sweden can resist erosion at velocities of up to about 50 cm/s. A dense plant cover with a high root volume, especially in the upper soil horizons, will, to some extent, protect from soil erosion. This means that newly established and sparsely growing plantings are at the highest risk for erosion damage.
In the modeled scenarios, we have calculated volumes and flow rates for a 1000m2 roof with four downspouts of equal size and share of the drainage work. We assumed that 80% of the roof area could be greened. Bare roofs and Sponge-Roofs release the stormwater at high velocities to interception bioswales even at 10-year storm events, leading to blowouts. Purple-Roof compliant systems, on the contrary, significantly reduced flow rates and thus flow velocities dramatically even during 100-year rain events, landing well below the threshold for soil erosion.
Soil Retention and Detention Capacities
Soil type, soil depth, and pooling depth of at-grade green stormwater infrastructure influence the available retention and detention volumes. Basically, most of the volume usually occupied by air can be occupied by water during a rain event. For soils, this is generally expressed through the concept of pore volume. A regular planting soil may have a pore volume of about 15%, whereas specific porous planting substrates may boast pore volumes of up to 30%. In practice, the pore volume will vary based on particle shape, degree of compaction, and plant root distribution, and some of the pore volume will be filled with air even during a flooding event. By replacing half of the soil depth in a bioswale with air, the theoretical detention volume could be doubled or even quadrupled compared to the soil-filled options. This, on the other hand, usually happens at the cost of the vegetation, especially on sites with non-permeable subsoils.
If instead, the pooling volume is added on top of a deep planting bed, both large vegetation and a good detention volume can be attained. Pooling water on top of the soil can to an extent also protect soils and plants from high flow speeds. On the other hand, the pooling water must be drained out of the bioswale within 24 hours to avoid rotting plants. Drainage can either be secured through highly permeable subsoils or piping. In the latter case, it is essential to consider where the pipes will connect, as many sewer systems may be overtaxed during more significant rain events, thus often taking the drainage pipes effectively out of function for a period of time even after a rain event.
Bioswale retention and detention during storm events
While bioswales are regularly outfitted with drainage, the common assumption is that bioswales will not drain out during larger storm events due to saturated subsoils and sewer systems running overcapacity. In our calculations, this has meant the assumption that the bioswales are in effect non-draining during the two model storms but that they will drain out slowly after the storm events as capacity in subsoils and sewer systems is restored.
In the modeled scenarios, 1-meter deep bioswales corresponding to 5% of the rooftop area were incapable of handling the stormwater volumes in either the 100-year storm event or the 10-year storm event, independent of the roof type or the soil type at the interception bioswales.
The runoff volume from a 10mm rain event, on the other hand, could be treated in the bioswales independent of the roofing solutions, provided that the interception bioswales used a more porous soil type. An improved interception bioswale that combines 1 meter of porous soil depth with 30 cm pooling depth could even detain a 10-year storm simply based on its volume but will suffer from erosion and severe vegetation damage if the outflow rate from the roof to the bioswale is not controlled.
Scaling up the Catchment Areas
Most individual built properties belong to larger stormwater-generating units, also called catchment areas, whose runoff collects at a single common point. Stormwater measures for specific catchment areas in Sweden are often sized with the rule of thumb that 5% of the total area of the catchment should be set aside for permeable and/or retaining structures with a minimum depth of 1 meter. The basic idea behind this rule of thumb is to provide enough retention and detention capacity to improve the quality of stormwater in a catchment area. For example, a 1-hectare catchment area should include a minimum of 500m2 of permeable area for treating stormwater at ground level.
Due to the complexity of stormwater management on larger catchment levels, it is difficult to make broad generalizations of the effect of scaling up green stormwater infrastructure. The detention and retention capacities of green roofs, bioswales, detention ponds and so on may vary wildly depending on their specific design. Additionally, the amount of runoff generated within a catchment depends largely on the canopy cover, surface materials and subsoils of the site. Catchment areas are also usually outfitted with pipe systems designed to transport water and counteract flooding, which can play a significant role in stormwater management during smaller rain events or towards the end of larger ones.
Despite the complexities and uncertainties, it is still clear that green stormwater infrastructure does provide opportunities for combining secondary benefits in addition to retaining, detaining, and improving the quality of stormwater to the extent that is not available with grey infrastructure.
From the simplified modeled scenarios, we can glean that the 5% rule of thumb is insufficient for sizing vegetated stormwater facilities for detention unless combined with additional detention volumes on top of plant soils. Increasing permeable surface to 15%-20% could provide enough detention volume for 100-year rain events of cloudburst type, especially when combined with rooftop detention in a treatment train.
Please don’t hesitate to contact the authors if you have any questions through our CONTACT FORM or via LinkedIn: Dr. Anna Zakrisson, biologist & Ella Uppala, Ph.D. student and landscape architect.