From Complexity to Practice: Testing the Hydrological Module of a Simplified Tool for Multiple-Benefit Assessment of Best Management Practices ^†

Abstract

Best Management Practices (BMPs) are key instruments for improving the resilience of urban environments to climate change and land-use pressures. They mitigate pluvial flooding and heat waves by restoring natural soil processes and providing multiple co-benefits at both the building and urban scale. Urban planning increasingly requires comprehensive assessments of the multiple benefits provided by BMPs, which extend beyond their hydrological function. Traditional hydrological models such as SWMM5 are robust and widely used for simulating drainage performance, but they are not designed to evaluate wider co-benefits or to be easily applied in planning contexts. For this reason, simplified tools have been developed to offer quicker and more accessible assessments, although their reliability, especially in reproducing hydrological outcomes, remains uncertain. This study examines the Green Values Stormwater Management Calculator (GVC), which has been developed to combine hydrological and co-benefit evaluations within a single, easy-to-use framework. In this preliminary analysis, we tested the hydrological module of the GVC on a 290-hectare mixed-land-use catchment in the metropolitan area of Milan, where two calibrated SWMM5 drainage models were available as benchmarks: one representing current conditions and another including a retrofitting design with BMPs. The scenarios were simulated with the GVC and compared under selected storm events in terms of total runoff volumes. The results show that the GVC reproduces current-condition runoff with good accuracy, but tends to underestimate BMP efficiency.

1. Introduction

Urban areas are increasingly challenged by the interplay of climate change and land-use pressures, which results in more frequent pluvial flooding and heat waves [1,2]. Best Management Practices (BMPs) offer effective responses, restoring natural soil functions and providing co-benefits at building and city scales [3,4,5]. Scientific research has extensively assessed BMP performance using advanced hydrological models such as SWMM5, which provide robust simulations of drainage behaviour [6,7]. Yet these models are primarily designed for detailed hydrological analysis and are not suited to the holistic evaluation of the multiple benefits that BMPs are expected to deliver in planning practice. For this reason, simplified multi-benefit evaluation tools have been increasingly adopted in planning contexts, as they can provide assessments that are comprehensive while also timelier and more accessible.

This raises the question of whether such tools, developed primarily for non-specialist use, can nonetheless offer reliable insights into both hydrological performance and the wider benefits of BMPs. Among the available tools, the Green Values Stormwater Management Calculator (GVC), developed by the Centre for Neighborhood Technology in Chicago (US), has gained recognition for its capacity to estimate the hydrological performance of BMPs as well as their additional co-benefits across multiple spatial scales [8,9,10].

The present study explores the effectiveness of the GVC through the Sesto Ulteriano case study, a mixed-land use district within the metropolitan area of Milan affected by pluvial flooding. The analysis compares stormwater runoff volumes estimated by the GVC with outputs from calibrated SWMM5 models under both current conditions and a retrofitting scenario with BMPs, considering rainfall events of different severities. The objective of this study is to assess the robustness of the hydrological module of GVC, providing evidence on the accuracy of its outputs. In doing so, the study begins to explore the extent to which such tools may complement advanced modelling within integrated urban planning processes.

2. Materials and Methodology

The case study is located in the municipality of San Giuliano Milanese (Metropolitan City of Milan, Italy) and extends over approximately 290 hectares. The catchment is served by a combined sewer network and is predominantly industrial, with impervious surfaces covering about 94% of the area. These conditions make the district highly prone to pluvial flooding and representative of many urbanised contexts. The hydrological module of the GVC estimates stormwater runoff using site characteristics (impervious surfaces, landscaped areas, roofs) and local rainfall data. Runoff volumes are derived using Curve Numbers (following the USDA-SCS method) assigned to different land cover types, adjusted when BMPs are introduced. For each BMP intervention, the model quantifies capture and retention volumes through simplified mass-balance calculations. Results can be obtained on an annual scale, based on long-term rainfall statistics, or on an event scale, allowing comparison of runoff with and without BMP implementation [11]. In order to test the hydrological module of the GVC, two event-scale scenarios were simulated: first, the Traditional Scenario (GVC-TS), reproducing the existing drainage system of the Sesto Ulteriano district, and second, the BMP Scenario (GVC-BS), representing a retrofitting strategy with 24 ha of BMPs (rain gardens, roadside swales and pervious parking lots) implemented according to

Table 1. Description and parameterization of BMPs.

BMP Name in GVC and SWMM5	BMP Total Area (ha)	BMP Layer in GVC	cm	-
Rain Garden in GVC and “Bioretention Cell Type 1 and Type 4—Bior.cell 1, Bior.cell 4” in SWMM5 [6]	2.86	Ponding Depth	10
		Amended Soil	25
		Amended Soil Porosity		0.35
		Aggregate Depth	25
		Aggregate Depth Porosity		0.35
Roadside Swale in GVC and “Bioretention Cell Type 2—Bior.cell 2” in SWMM5 [6]	9.91	Ponding Depth	20
		Amended Soil	40
		Amended Soil Porosity		0.35
		Aggregate Depth	40
		Aggregate Depth Porosity		0.35
Park Lot Swales in GVC and “Bioretention Cell Type 3—Bior.cell 3” in SWMM5 [6]	4.84	Ponding Depth	70
		Amended Soil	45
		Amended Soil Porosity		0.35
		Aggregate Depth	45
		Aggregate Depth Porosity		0.35
Park Lot Swales in GVC and “Permeable Pavement—Pe.pav” in SWMM5 [6]	6.61	Bedding	3
		Bedding Porosity		0.25
		Base	4
		Base Porosity		0.3
		Sub-Base	15
		Sub-Base Porosity		0.3

and following the parameterization proposed by D’Ambrosio et al. [6] in a previous SWMM5-based study. For both GVC-TS and GVC-BS, runoff volumes (V) were simulated under 9 h design storms with return periods (T) of 2, 5, and 10 years, respectively, corresponding to cumulative depths of 44.7, 61.9, and 74.8 mm. From these results, a performance indicator, R-BMPs_GVC (Equation (1)), was defined to quantify the relative reduction in runoff volumes achieved through BMP implementation compared with the baseline. To verify the robustness of these results, the same land use and climatic scenarios were simulated using SWMM5, benefitting from a previously calibrated model of the study area drainage network [12]. Two parallel scenarios were thus defined: the Traditional Scenario (SWMM5-TS) and the BMP Scenario (SWMM5-BS). Based on their outputs, the same performance indicator was calculated and denoted as R-BMPs_SWMM5 (Equation (2)), allowing for a direct comparison with the R-BMPs_GVC values obtained from the GVC.

R-BMPs_GVC = [(V_GVC-TS − V_GVC-BS)/(V_GVC-BS)]·100

(1)

R-BMPs_SWMM5 = [(V_SWMM5-TS − V_SWMM5-BS)/(V_SWMM5-TS)]·100

(2)

3. Results and Discussion

The comparison between total runoff volumes (V) in the GVC and SWMM5 produced noteworthy results (

Table 2. Comparison of GVC and SWMM5 hydrological performance in the analysed scenarios.

	Total Runoff Volumes—V (m3)			BMP Performances (%)
	GVC-TS	GVC-BS	SWMM5-TS	SWMM5-BS	R-BMPsGVC	R-BMPsSWMM5
T = 2	104,080.93	69,219.03	107,857.1	61,510.15	33	43
T = 5	153,190.30	113,124.42	147,132.7	91,274.19	26	38
T = 10	189,580.35	146,848.69	176,249.9	113,919.4	23	35

). Under the Traditional Scenario (GVC-TS and SWMM5-TS), total runoff volumes were broadly consistent across the two models.

Differences were limited to approximately 3.6% for the 9 h storm with a 2-year T and to 3.9% and 7% for the 5- and 10-year T precipitation events, respectively, with higher volumes predicted by GVC-TS in the latter cases. When BMPs were introduced (GVC-BS and SWMM5-BS), the discrepancies became more pronounced. The runoff volumes (V) simulated by GVC-BS exceeded those obtained from SWMM5-BS by 11.1–22.4%, depending on precipitation severity. This outcome was expected, given the different structural assumptions of the models. The GVC treats BMPs as isolated storage units, where retention capacity is defined by the void volume of the structural layers, without accounting for possible hydraulic interactions. By contrast, SWMM5 allows BMPs to be interconnected with each other and with surrounding pervious areas, enabling a more realistic representation of their cumulative hydrological response. It should also be noted that infiltration processes are modelled differently in the two tools. In SWMM5, BMP infiltration typically adopts a Green–Ampt formulation [13] that accounts for soil hydraulic properties and dynamic storage, whereas the GVC relies on the SCS Curve Number method, which provides a more simplified estimation of runoff based on soil characteristics. The impact of these modelling differences is evident when considering the performance indicators (R-BMPs_GVC and R-BMPs_SWMM5). In the GVC, BMPs achieved reductions in runoff volumes (R-BMPs_GVC) of 33%, 26%, and 23% for T of 2, 5, and 10 years, respectively. In SWMM5, the corresponding reductions (R-BMPs_SWMM5) were higher (43%, 38% and 35% respectively), suggesting that a simplified tool may underestimate BMP effectiveness, probably due to their more schematic representation of hydrological processes.

4. Conclusions

This study aimed to assess whether the GVC hydrological module can provide assessments consistent enough to support practitioners and local authorities in preliminary planning and feasibility analysis involving BMP retrofitting. The preliminary outcomes indicate that this simplified tool can generate acceptable hydrological results within a short computational time frame, although with a tendency to underestimate BMP performance compared with advanced hydrological modelling. Nevertheless, further investigations across different urban contexts, application scales, and intervention types would be beneficial to better define the potential of the tool, including analyses that extend to cost–benefit evaluations. In addition, future work should consider the ability of the tool to capture the temporal variability in BMP performance, since nature-based systems may experience efficiency losses over time due to ageing or insufficient maintenance [14,15]. Addressing these aspects would significantly enhance the relevance of the tool for long-term planning and monitoring.