ELEMENTOS QUE INTERVIENEN

2.3 Hydrologic modelling of green infrastructure

To understand various aspects associated with urban stormwater, researchers and professionals have been using stormwater management modelling tools extensively (Jayasooriya and Ng, 2014). Software tools have been used for modelling urban stormwater quantity since the mid-1960s and models that are able to simulate stormwater quality and quantity started to emerge in the early 1970s (Zoppou, 2001). Components that can assess the effectiveness of GI practices were then incorporated into these models after the identification of GI practices being an important means for urban stormwater management. For most of these new models, the primary goal was to investigate the impact of GI practices on urban stormwater runoff quality and quantity (Jayasooriya and Ng, 2014).

Zoppou (2001) summarized a number of mathematical models that simulate urban stormwater, and also introduced many existing urban stormwater modelling packages/tools that are capable of simulating urban stormwater quantity and quality. However, the modelling tools do not incorporate GI practices (Jayasooriya and Ng, 2014). Elliott and Trowsdale’s (2007) review focused on the ability of models to represent GI practices. They identified approximately 40 models for simulating urban stormwater from various sources, and selected 10 models that were available and had not been superseded at the time of the review to discuss in depth their strengths and weaknesses of modelling GI in terms of stormwater quantity and quality. Ahiablame et al. (2012) investigated the effectiveness of GI for managing stormwater. They also discussed in details three different stormwater and GI modelling tools which addressed the quality and quantity of runoff. Jayasooriya and Ng (2014) reviewed 20 modelling tools that are able to simulate stormwater management and/or economics of GI practices. Ten models were then selected for detailed review due to their popularity among stormwater management professionals and had been widely used in research.

From the hydrologic models reviewed by Elliott and Trowsdale (2007), Ahiablame et al.

(2012) and Jayasooriya and Ng (2014), three models – SWMM, SUSTAIN, and L-THIA-LID – have been identified to be able to simulate hydrologic aspects of GI practices, are popular (depending on number of publications), and have updated versions with readily available information.

SWMM

The U.S. Environmental Protection Agency’s (EPA) Storm Water Management Model (SWMM) has been a popular urban and non-urban storm water modelling tool among water resource professionals and researchers (Abi Aad et al., 2010, Gironás et al., 2010, Jang et al., 2007, Khader and Montalto, 2008, Rossman, 2010). The public domain software which is known for its dynamic rainfall-runoff analysis is capable of simulating runoff quantity and quality from mainly urban areas (but with many applications in non-urban areas as well) in response to single event or long-term (continuous) precipitation input. SWMM was first developed in 1969-71 and has undergone major upgrades since then. The newest version is SWMM 5.0 (Rossman, 2010).

According to Jayasooriya and Ng (2014), the model simulates stormwater runoff quantity through several physical hydrologic processes, including time-varying precipitation (historical or synthetic storms); accumulating and melting of snow; standing surface water and ground water evaporation; depression storage intercepting rainfall; rainfall infiltration into unsaturated soil layers; percolation of rainwater into groundwater zone; ground water and the drainage system interflow; nonlinear reservoir routing of overland flow; and detention and retention of rainfall by various types of GI practices (Jayasooriya and Ng, 2014, Rossman, 2010).

The model estimates runoff based on a collection of subcatchment areas. Each subcatchment area receives rainwater and diverts the runoff either to another subcathcment, to storage devices or to nodes through conduits, while allowing for evaporation and infiltration losses.

During a simulation period which consists of multiple time steps, SWMM records runoff quantity and quality produced within each subcatchment; and the rate and depth of the flow and water quality in each pipe and channel (Rossman, 2010).

The physically-based, discrete-time simulation model, SWMM, uses several components to model stormwater runoff quantity, including surface runoff, infiltration, groundwater, flow routing, snowmelt, and surface ponding. Each subcatchment is treated as a nonlinear reservoir. Precipitation and water from designated upstream subcatchments are considered as incoming waters, and infiltration, evaporation and surface runoff are outflows. Surface runoff occurs only when the depth of water in the subcathcment exceeds the maximum depression storage, in which case Manning’s equation is used to estimate surface runoff. The model allows five ways to account for evaporation: a single constant value, a set of monthly average

values, a user-defined time series of daily values, values computed from the daily temperatures, and daily values read directly from an external climate file. SWMM offers three choices for modelling infiltration: Horton’s Equation, Green-Ampt Method and Curve Number Method. A water balance equation is numerically solved for the depth of water over the subcatchment (Rossman, 2010).

SWMM uses the two-zone groundwater model. The upper zone is unsaturated hence the moisture content varies. The lower zone is fully saturated and the moisture content is fixed.

After accounting for the water fluxes, a mass balance is written to compute a new water table depth and unsaturated zone moisture content. Three choices are available for flow routing:

steady flow routing, kinematic wave routing and dynamic wave routing. When excessive water exists in a system, it can be chosen either to be transported further downstream or be stored in a ponded fashion (Rossman, 2010).

In each subcatchment, the snowmelt component considers snow accumulation, snow redistribution and removal operations to update the state of the snow packs. Any melted snow is considered as an additional rainfall input onto the subcatchment (Rossman, 2010).

SWMM continues to serve as a tool to plan, analyse and design related to stormwater runoff throughout the world. It has been applied to thousands of sewer and stormwater studies worldwide. Particularly, it is capable of evaluating the effectiveness of several GI practices such as bio-retention cells, infiltration trenches, pervious pavement, rain barrels/cisterns, and bioswales (Rossman, 2010, Jayasooriya and Ng, 2014, Abi Aad et al., 2010). SWMM was also applied to simulate green roofs (Khader and Montalto, 2008, Roehr and Kong, 2010). GI practices provide some amount of rainfall or runoff storage, infiltration into soil and evaporation of stored water (except for rain barrels/cisterns) (Rossman, 2010).

In SWMM, GI practices are modelled by a group of vertical layers. The layers are characterized on a per-unit-area basis, which allows the same GI practice design but with different areal coverage to be easily placed within different subcatchments of a study area.

SWMM computes a moisture balance that tracks water movements between the layers and the amount stored in each layer. Figure 2-2  shows the layers used for modelling a bio-retention cell and the flow pathways (Rossman, 2010).

Figure 2-2 Conceptual diagram of a bio-retention cell (Rossman, 2010)

Layers available in SWMM-05 are Surface layer, Pavement layer, Soil layer, Storage layer, and Underdrain System. These are now defined separately in the following. Surface layer is the ground/pavement surface which treats direct rainfall and runon from upstream land areas as inflow. Excess inflow is stored in depression storage. Surface outflow produced either leaves the system through the drainage system or flows onto downstream land areas.

Pavement layer is a layer of concrete or asphalt that is pervious which is used in continuous porous pavement systems. This layer can also be paver blocks and filler material that are used in modular systems. Soil layer contains engineered soil mixture that is used to sustain vegetation growth in bio-retention cells. Storage layer provides water storage in bio-retention cells, porous pavement, and infiltration trench systems by using crush rock or gravel. The function of the underdrain system is to transport water out of the storage layer of bio-retention cells, porous pavement systems and infiltration trenches into common outlet pipe or chamber (Rossman, 2010).

SUSTAIN

The System for Urban Stormwater Treatment and Analysis IntegratioN (SUSTAIN) has been developed by the U.S. Environmental Protection Agency (EPA) since 2003 in response to the challenge faced by decision makers which is to select the best combination of GI practices for implementation among many other options available to achieve most practical and cost-effective for the location interested. In other words, SUSTAIN assists watershed and stormwater practitioners to develop, evaluate, and select optimal GI options for various watershed scales based on cost and effectiveness. SUSTAIN is a decision-making framework intended to be used by knowledgeable model users who are familiar with technical aspects of

watershed modelling. The objectives SUSTAIN addresses are to determine the effectiveness of GI practices in reducing runoff and pollutant loadings; to identify the most cost-effective solutions that meet water quality and quantity objectives and to select the location, type and size of GI practices (United States Environmental Protection Agency, 2015, Lai et al., 2010).

SUSTAIN integrates modelling techniques, cost management practices, and optimization tools that are publicly available in a geographically-based platform, ArcGIS, to achieve the planning objectives. From multiple, distinct sources, the most applicable algorithms for simulating urban hydrology, pollutant transport, and mitigation processes were packages together in SUSTAIN (Lai et al., 2010).

The user accesses the framework modules in SUSTAIN through a base platform interface (Lai et al., 2010). There are several framework modules in SUSTAIN – Framework Manager (including BMP siting), Land simulation, BMP simulation (including BMP cost estimation and aggregation of distributed BMPs), Conveyance simulation, BMP optimization, and Post-processor (Shoemaker et al., 2009, Lai et al., 2010). Functions performed in each module are described in Table 2-27.

SUSTAIN can be used to address various stormwater management practice planning questions faced by decision makers, e.g. to determine optimal GI strategies to reduce runoff volume and peak flows; to evaluate the water quantity and quality benefits of distributing GI practices in urban streams (United States Environmental Protection Agency, 2015).

Many GI practices are supported by SUSTAIN, including bio-retention cells, rain barrels/cisterns, constructed wetlands, dry ponds, grassed swales, green roofs, infiltration basins, infiltration trenches, pervious pavements, sand filters (non-surface and surface), vegetated filter strips, and wet ponds (Shoemaker et al., 2009).