2.2. Multi-Criteria Analysis (MCA)

The multi-criteria (or multi-attribute) analysis (MCA) involves the use of different types of variables aimed at providing a framework that allows to quantify preferences. This is particularly useful in the field of sustainability, where variables with different units are involved. Therefore, a MCA was used in this study to evaluate the most preferred alternative for the management of the CSO of Buccinasco. The literature framework of the MCA is discussed in the following sub-sections, in particular with regards to the use of value functions for the standardization of different units.

Criteria were defined to compare different alternatives for the management of the Buccinasco CSO (Table 1). The selected criteria permitted to cover the important elements for the decision, including the main objective of the intervention (water quality), the NBS side benefits (air quality, biodiversity, carbon reduction and sequestration, education health, recreation, and wastewater treatment), and the negative impacts (CAPEX, OPEX and administrative issues).

Criteria (with orientation and indicators) and weights defined for the sensitivity analysis, expressed in absolute (A: scale 0–10) and normalized (N: scale 0–1) terms.

W1 Expert-based (Case study). W2 Equal weights. W3 Citizens. W4 Environmentalists. W5 Grey infrastructure professionals. +: positive orientation, i.e., higher value of the indicator, higher preference of the criterion. -: negative orientation, i.e., lower value of the indicator, higher preference of the criterion.

Despite commonly used in MCA for CWs, the criteria for nuisance was neglected, since the proposed intervention is aimed at improving an already present nuisance condition, i.e., odor releases from the CSO discharge point. The areal footprint was also neglected, as it is already accounted for in the CAPEX criteria and is in line with EU policy of minimizing land consumption; indeed, the transformation of a non-natural area (urban or agricultural) into an NBS should be considered as an advantage, i.e., an occasion to provide ESs, rather than as a disadvantage as obstacle to future different anthropogenic uses of the land. For the same reasons, tax revenues were not considered as a criterion. Indeed, the alternatives were all located in areas not foreseen for urban development by the Municipality’s plans. Therefore, none of the alternatives were compromising future tax revenues from the land used.

Two groups of alternatives were defined, Alternative 1 for grey infrastructure and Alternative 2 for green infrastructure. Grey and green infrastructure are intended according to the definition given by Natural Water Retention Measures (nwrm.eu): grey infrastructure, solutions that use traditional methods to manage water, preventing any type of ecosystem from growing on it, and often built in concrete; green infrastructure, solutions that recreate natural or semi-natural areas to provide multiple services. Three alternatives were chosen for each group. The alternatives were sized in agreement with the recent Regional Regulations (R.R. 06/2019) as follows: the first flush tanks were sized on the basis of the impervious catchment, i.e., 29.3 hectares; in-line treatments (both green and grey infrastructures) were sized considering to continuously treat a CSO flow rate of 105 l/s, i.e., up to a dilution rate of 6 in comparison to the wastewater flow rate in the sewer during dry periods.

Grey infrastructure included both inline treatment and first flush tanks. As inline treatment, the option of primary treatment only (Alternative 1.1) was chosen, in order to consider the cheapest alternative but with the lowest water quality performance. Therefore, Alternative 1.1 included a preliminary automatic grid followed by a sedimentation tank of 375 m³, with an aerial footprint of 430 m² (minimum hydraulic retention time: 1 h, targeted removal efficiencies: 50% TSS and 25% BOD₅). Two first flush tanks were considered, following the minimum and maximum sizes indicated by the R.R. 06/2019, i.e., 25 and 50 m³ per impervious hectare. Consequently, Alternative 1.2 and Alternative 1.3 assumed a first flush tank of 730 m³ (areal footprint 540 m²) and 1460 m³ (areal footprint 880 m²), respectively.

Green infrastructures were chosen and sized following some of the most successful state-of-the art approaches used for CSO treatment with CWs [13]. Alternative 2.1 adopted the French approach, i.e., a single stage vertical subsurface flow constructed wetland (VF) with a net area of 3600 m² and a total gross areal footprint of 7300 m². Alternative 2.2 considered the Italian approach, i.e., a hybrid CW with a VF CW as 1st stage and a free water surface (FWS) system as 2nd stage; the net surfaces of the VF and the FWS were sized equal to 3600 m² and 1500 m², respectively, requiring a total gross areal footprint of 9550 m². Alternative 2.3 also considered the Italian approach, but it was designed in a new park, following the example of the Gorla Maggiore Water Park [10,11]; the sizes of the VF and the FWS were the same as in Alternative 2.2, while the required total gross areal footprint was estimated equal to 19,750 m². All the green infrastructures were assumed to provide the same high CSO treatment performance (TSS > 90%, BOD₅ 50–70%), in line with recent state-of-the-art evidence [13].

The plan views of the areal footprints are visible in Figure 1 for all the alternatives.

Plan view of the areal footprint and positioning of the different alternatives: (a) Alternative 1. 1, primary treatment only (primary only); (b) Alternative 1.2, first flush tank, minimum size according to local regulation (FFT—min); (c) Alternative 1.3, first flush tank, maximum size according to local regulation (FFT—max); (d) Alternative 2.1, constructed wetland with a single stage (VF); (e) Alternative 2.2, multistage constructed wetland (VF+ FWS); (f) Alternative 2.3, multistage constructed wetland sited in a newly developed urban park (VF + FWS park).

Five sets of weights were defined to perform a sensitivity analysis, which are shown in Table 1. For the first group, the weights were chosen to be representative of the case study with an expert-based approach (W1), i.e., considering the interview with the Major of the town of Buccinasco, a site visit, and the study of local context in terms of all the aspects of interest, such as environmental conditions, urban planning, and local legislation. The weights of the second group were assumed to be equally distributed, to investigate the best alternative in case of no preferences among the criteria. Finally, three sets of weights were defined to simulate the preferences of some particular stakeholders: (i) citizens (W3), more interested in social and welfare aspects; (ii) environmentalists (W4), who were assumed to prefer criteria related to environmental impact such as water or air quality; (iii) grey infrastructure professionals (W5), a general stakeholder group that was selected to highlight the “design as usual” preference, i.e., only considering costs in relation to the single main benefit of water quality.