In the present work, as said, a significant application of Naples SUMP with specific reference to the design of a sustainable transport scenario for one of the highest density and more congested area of the city. The territorial area object of the analysis was the Municipio Square in the centre of Naples, where a new station ‘Municipio’ of the Metro Line 1 was under construction. This Metro Line, famous for its ‘art stations’ (‘Toledo’ station, for example, was defined as ‘The most impressive underground railway stations in Europe’ by The Daily Telegraph [10]), is an important example of high quality design aimed at achieving multiple objectives.

For the design of the layout of this area was considered both vehicular traffic flows (private and public transport) and the huge pedestrian flows. In fact, the attractiveness of this place is manifold. Inside the square there is the Maschio Angioino Castle; in addition, close to the square, within walking distance, there are other sites of historical interest (the Galleria Umberto I, the San Carlo Theatre, the Royal Palace and Plebiscito square) and some commercial sites (Via Toledo, Corso Umberto). Furthermore, this area is a link between the city and the sea: Municipio Square, that has a large extension, borders the marina and, daily, many tourists arrive in Naples with cruise ships. Finally, it can be noted that the square also represents a great attraction place for residents, given that the local government building, many other offices (e.g., banks, insurance companies) and an historical theatre (Mercadante Theatre) are located in. The new Metro station, also characterized by the presence of important archaeological finds (that will become a “compulsory” museum for all users who will pass through it), is further increasing the attractiveness of the area for pedestrians.

Also from the transport point of view, Municipio square plays a central role for the city of Naples. In fact, the adjacent roads (Via Acton, via Medina and via Cristoforo Colombo) are identified as primary urban roads. The two coastal routes (Via Caracciolo-Via Acton- via Colombo-Via Marina and Via Marina-Via Depretis -Piazza Municipio-Via Acton-Via Riviera di Chiaia) are a forced route for west-east-west urban traffic.

The final design layout of Municipio square is reported, with three different point of view, in the next Fig. (2).

The pedestrian area will be the central part of the square with an extension of about 12,500 m²; the vehicles will have two roads with two lanes (3.5 m wide). It will be realized a pedestrian underpass (equipped with escalators, lifts and treadmills) to connect the square both with marina and Metro Line 1, following the double aim of facilitating the pedestrian routes and reducing the traffic congestion by means of limiting the inferences between people and cars. Finally, the Metro Line 1 will have in this area 5 entry/exit points (Palazzo San Giacomo, via Medina, via Depretis, Porto, Scavi Archeologici) to ensure a wider accessibility all around the square. At the moment, the Metro Line 1 have, in the zone of interest, two already operating entry/exit points (Palazzo San Giacomo, via Medina) as illustrated in the following Fig. (3).

Fig. (2). Some details of design solution of Piazza Municipio Naples, Italy (source: Metropolitana di Napoli S.p.A.).

Fig. (3). Layout of Piazza Municipio with the two already operating entry/exit of the Metro station.

Using a complex zoning process, it was possible to concentrate, on a limited number of points, the trips with origin destination from/to the studied area. For example, with respect to the macroscopic model implemented, 399 traffic/ zones were defined (through different transportation criteria) of which 253 inside the city of Naples and 146 within the external areas. The zoning process, used in this work, is consistent with the supply model that is intended to simulate: more detailed near the city centre, less dense towards the peripheric area.

A specific survey campaign was performed aimed: i) to estimate the incoming and outgoing vehicles flows to/from the study area; ii) to estimate percentages of vehicles turning maneuverers at main intersections. A preliminary survey has allowed identifying the peak period (maximum vehicular and pedestrians interactions and congestion). The test-survey was performed both in winter and in summer of 2016. These allowed to identify that peak period for pedestrians and vehicles are the months of June and July. Starting from these considerations, 21 counting sections were monitored for 15 weekdays (10 business days and 5 holydays) in June and July 2017, both in the peak hours of the day and in the off-peak ones. Both vehicular and pedestrian flows were surveyed. For vehicles flows the following types of information were collected: geometrical characteristics of roads and intersections (cross sections, number and width of lanes, presence and dimensions of roadside parking and pedestrian crossings); type of intersections; characteristics of the traffic light (cycle time, number and duration of phases); number and characteristics of vehicles (cars, motorcycles, buses, taxis and light and heavy goods vehicles) in different hours of the day; percentages of turning maneuverers, per type of vehicles and intersections; traffic flows for each lane and queues lengths and propagation, through traffic surveillance cameras Through these monitoring, was also possible to (anonymously) read the license plates of the vehicles to estimate the origin-destination matrix (O/D matrix) and to evaluate the travel times for each origin-destination and vehicle categories identified.

Finally, for better calibrate the implemented models, speeds, free flow travel times and queues length at main intersections of the study area have been measured.

For pedestrian flows, a Revealed Preferences (RP) survey was performed, in order to estimate the mobility behaviours of a sample of users. Such surveys were carried out by interviewing pedestrians at specific points of the transport system (e.g., at the main pedestrian crossings of the project area). The submitted questionnaire to about 300 pedestrians allowed to collect: gender, age, purpose and frequency of the trip, number of “travel friends”, origin and destination of the trip and path followed in the square (to be chosen among those shown on a map proposed to the interviewed). In fact, the peculiarity of the square allowed to identify only a few possible paths (4-7 from the interview point) among which a pedestrian can choose how to move. Jointly with the interviews, pedestrian counts were also made in 21 sections. The combined use of the counts (sampling universe) and the results of the interviews (sample interviewed) allowed, through the application of consolidated statistical inference technique, to estimate the number of pedestrian crossing over the time and estimate the O/D matrix for both the morning and the evening peak hours.

Through the estimated models, was possible to compare the impacts of different design scenarios through the evaluation of both Global Indicators (e.g., total time for crossing the study area [h], maximum queue length [m], number of vehicle per km and number of vehicles simultaneously present in the study area/vehicular density), and Local Indicators (e.g., average length of queue at each road lines [m]).

The measurement campaign results show that the study area is affected by a daily vehicle traffic of about 80,000 Passenger Car Equivalent (PCE), equally divided during the morning time period (8 a.m. to 2 p.m.) and afternoon ones (2 p.m. to 8 p.m.). In term of hourly trends (Fig. 4), the morning peak hour (from 8 a.m. to 9 a.m.) presents 7,500 equivalent cars (equal to about 10% of the daily flow: respectively 52% cars, 5% taxis, 37% motorcycles, 5% heavy good vehicles, 1% buses).

Fig. (4). Example of count sections for estimate vehicles flows and the Origin/Destination matrix (passenger car equivalent, PCE, for morning and afternoon peak hours).

Regarding the pedestrian survey, the results indicate that 52% of users move for tourism (of which 50% occasionally or once a week) and the remaining 48% for work or study reasons (of which 50% every day or 2/3 days a week). The employed people are 47% of the total, the students 18%, the unemployed 10% and the tourists 25%. Almost 3,000 are the pedestrians that move in the whole studied area (of which 60% are the movements in the marina area, while the remaining 40% regards via Toledo shopping street and surrounding areas). The following Fig. (5) shows the O/D matrix in the case of maximum pedestrian load: the assumption of maximum load is the sum of all transport modes, also that of cruise ships. This last is strongly variable and depends on the number of cruise ships moored at the seaport. In minimum load situation, the pedestrian movements are 30% less.

Fig. (5). The pedestrians’ O/D matrix estimation results (morning peak hour).

Different layout design scenarios regarding Municipio Square were tested. They can be summarized as follows. For the wide area around Municipio square, 3 basic traffic schemes (that differ with each other for driving directions, number of lanes, lanes width and pedestrian paths) have been compared. In each of these scenarios, in the west part of the area has been foreseen by the design of a new roundabout at the intersection among the following roads Via S. Carlo, via Gongaza and Via V. Emanuele III. This new roundabout (Fig. 6a), allowing to facilitate the manoeuvres at the intersection and moderate vehicles speeds, was considered a fixed point for all the scenarios.

On the contrary, in the east zone for the area called Largo Immacolatella, two different solutions have been evaluated (see Figs. 6 b1 and b2). These ones differ from each other for the presence/absence of traffic lights and, consequently, for the different number of lane and their different widths.

Fig. (6). The most significant simulated traffic scenario for the wider area around Municipio square with: a) the new roundabout in the west zone; b1) and b2) the two alternatives for the east zone.

Overall, 6 different design scenarios were simulated and just the results relative to the most significant one (whose main features are generally illustrated in Fig. (6) is discussed in the present paper.

From the transportation point of view, the main simulation results (reported in Fig. 7) show: i) a significant reduction in the queue length (form -35 m to -205 m) due to a better distribution of flows along the new available paths; ii) an increase in the number of vehicle per km due to new longer paths of about 5%; iii) a general decrease in total delay (an increase is showed just for the peak hour in the evening); iv) a significant decrease in the number of the vehicles that cross the Municipio square (-22% in the morning peak hour, while -31% in the evening). Furthermore, among the alternative solutions tested for the east zone of the study area, the best solution appears to be that reported in the previous Fig. (6), producing significant improvement during the peak hours both in the morning and in the evening.

Fig. (7). Screenshot of simulation results for the peak hour in the morning: on the left the current scenario, on the right the best design scenario. Comparison in absolute and percentage values is reported in the table both for the peak hour in the morning and for the in peak hour in the evening.

Results relative to the other design scenario tested (not detailed for brevity) show an average increase (with respect to the current scenario) of the length of queues (e.g. in Largo Immacolatella all day, in Via C. Colombo in the morning, in Via Marina in the evening) and the delay, so these scenarios were discarded.

Furthermore, another design scenario was tested considering a vehicle restricted area within the Municipio Square. The results of this scenario were a modest decrease in traffic flows with a consequent of increasing both the overall number of vehicles per km (due to longer paths, see for example the flows from B, C, D to F in the Fig. (5) and the average length of queues, also this scenario has been excluded.

In the morning peak hour (from 8 to 9 a.m.), the pedestrian flows were estimated as follows: the sum, of systematic demand (about 2,000 pedestrians), the cruise tourists (about 1,000), and the generated demand produced by the new station ‘Municipio’ of the Metro Line 1 as detailed in Fig. (8a). Overall more of 3,000 pedestrians that simultaneously pass through the study area were estimated. The following Fig. (8b) shows the simulated pedestrian flows along the available paths in the study area and the ones entering/exiting the new Metro station.

Fig. (8). a) The five entrances and exits of the new metro station L1 ‘Municipio’ that will be coming into operation in the end of 2018; b) the main simulated pedestrian flow relative to the best design scenario implemented.

As said, one of the aims of the paper was to perform a cost-benefit analysis aiming in improving transport, urbanistic, artistic/cultural, aesthetic, economic and environmental aspects as well as liveability for citizens, transport users (public and private) and tourists of the study area. The analysis was performed according to the European guidelines [23] and a recent research study [24].

The main infrastructures implemented relative to the best design scenario is the pedestrian underpass reported in Fig. (9).

The comparison among the project alternatives was performed through the difference between costs and benefits over the years. Defined and quantified (in monetary terms) impacts related to the new infrastructures (for each design scenario), some Measure Of Effectiveness (MOE) were estimated. Net Present Value (NPV) is the measure of the profitability of an investment that is calculated by subtracting the present values of cash outflows (including initial cost) from the present values of cash inflows over a period of time. Pay Back Period (PBP) is the period of time required to recoup the funds expended in an investment, or to reach the break-even point (return of the investment). The selection of the best project scenario to develop was defined comparing the MOE indicators estimated. Three different underpass scenarios were compared: i) no underpass but rationalization of pedestrian crossings; ii) “short underpass”, which is an underpass that allows to cross only half square in underground (with about half benefits produced); iii) “long underpass” that allows to cross all the square in underground. Table 2 shows the comparisons between the simulated scenarios in terms of estimated MOEs.

Fig. (9). Some details of the pedestrian “long underpass” implemented in design solution of Municipio Square (source: Metropolitana di Napoli S.p.A.)

As reported in Table 1, the best solution is the “long underpass” that, although it is the one with the highest implementation costs, will produce the greatest benefits that are (Fig. 10):

• An average reduction of vehicles queued by -35%, compared to the scenario without a pedestrian underpass;
• A reduction in average waiting time at the intersections of 2.4 - 3.2 minutes per vehicle;
• A reduction of fuel consumption of about 0.20 € per vehicle;
• A reduction in pollutant emissions of about 500 g of equivalent CO₂ per vehicle crossing the study area.

All these benefits will produce more than 40 million € for 6 years and more than 70 million € for 20 years. Comparing them to the construction costs was estimated a Pay Back Period lower than 3 years and a positive Net Present Value. Finally, a sensitivity analysis was also performed, confirming the robustness of the design solution implemented.

Fig. (10). The social benefits due to the construction of the pedestrian “long underpass” (2017 price).