The High Speed Rail and the City – part 1

Routing High-Speed Rail in urban areas: the case of Florence

Over the past few weeks, I posted two threads on Bluesky: one about the history of planning the High-Speed Rail (HSR) route through Florence, and another about the conflicts surrounding the “fast quad-tracking” of a section of the Adriatic line between Bologna and Castel Bolognese, which is still in the early planning stages. I’ve been asked to condense them into shorter Substack posts for easier sharing, and I’m happy to do so, taking the opportunity to improve the content and broaden the discussion. Since these posts cover different but related topics, I’ve decided to give them a common title, “High Speed Rail and the City.” In the next paragraph, I will explain why I believe they are connected.

The High-Speed Rail and the City: A Fundamental Planning Problem.

One of the main advantages of HSR as an intercity travel mode is its capacity to reach city-centre destinations and, more importantly, to connect with the key nodes of local and regional transit networks, which are often situated centrally. This is something that planes, due to the airports' peripheral locations, struggle to do. However, it also presents a significant planning challenge. Routing a high-speed railway line, with its limited geometry and broad clearances, through a central urban area is both complicated and costly, especially if there is no suitable existing rail corridor or if it is already used by local or freight traffic. The case of Florence provides an insight into the difficulties in balancing trade-offs during the planning process.

The case of the Bologna-Castel Bolognese quad-tracking highlights another planning issue connected to the relationship between HSR and the City. Specifically, it concerns how to organise mixed-use traffic on urban and peri-urban rail corridors where there is demand for multiple levels and types of services. While high-speed rail infrastructure is often seen as a separate, self-contained system built solely to serve fast long-distance train services, it is not always wise to allocate the new rail infrastructure, suitable for higher speeds, exclusively to long-distance intercity services. Especially near major urban centres, where rail infrastructure can support high demand for urban, regional, and interregional trips, as well as freight, the additional capacity from extra tracks—regardless of their design speed—can be an opportunity to separate different types of traffic. High-speed can coexist with other services, but which ones? This topic will be explored in the second post, which will be published in the coming days or weeks. But now let’s delve into the case of Florence.

The Italian HSR and the City: the case of Florence.

One of the major controversial issues during the early planning of the Italian HSR North-South central spine, which spanned from the late 1980s to the late 1990s, was how it would handle cities. To avoid being overly detailed, we only briefly mentioned this topic in the recently published paper on Italian high-speed rail planning history, which I co-authored with Paolo Beria. Here, I will explore the case of Florence in more depth, as I believe it illustrates the trade-offs between constructability, costs, connectivity, and accessibility involved in planning how HSR will integrate into larger metro areas.

When the idea of a national HSR network gained momentum after the 1985 National Transport Plan, it was reflected in regional transportation planning efforts across Italy. The introduction of HSR was viewed as a major chance to rethink urban planning and mobility, and to promote public transportation, especially rail-based, under the slogan “The Rail Treatment” (La Cura del Ferro, in Italian). Tuscany was no exception. In 1989, as HSR was still in its early planning stages and alignment was being decided, the Regional Transportation Plan proposed a fully underground route through the city centre, with a new underground station at Santa Maria Novella (SMN), the city's historic stub-end rail terminal. This alignment would have ensured optimal connectivity with regional trains and direct access to the city's core, similar to what currently exists in Bologna, while allowing all high-speed services to flow smoothly along the main corridor. However, placing a large 500-metre-long C&C box for a four-track station around or below Santa Maria Novella (SMN), including very long approach tunnels under sensitive heritage and landscape-protected areas, as well as beneath the Arno riverbed, was considered too complex.

The route of the HS line through Florence, as proposed by the 1989 Regional Transportation Plan of Tuscany’s Regional Government.

The preliminary project presented in 1992 by TAV (Treno ad Alta Velocità), the Public-Private Special Purpose Vehicle established in 1991 to develop the HSR, mainly proposed instead routing the HSR above ground along the existing rail corridor of the then-closed Faentina local railway and the 1930s Western belt, utilizing large areas of Campo di Marte for a new at-grade through-running railway station serving both high-speed and regional trains. To maintain some service to the more centrally located SMN, TAV suggested using two of the four existing tracks of the railway belt for a spur to SMN. The city and the region did not support this idea, as it would have split the high-speed service between Campo di Marte and SMN, with most of it serving the more peripheral Campo di Marte, and it would have prevented the reopening of the Faentina railway for local service.

The route of the HSR line through Florence as proposed by TAV Spa’s preliminary project in 1992.

As planning for the HSR advanced through the 1990s and hopes for private investment in a lean, "cheap" HSR project diminished, local stakeholders across Italy called for a high-speed rail project with a broader scope. This involved a major rethinking of urban hubs to support the expansion of suburban and regional rail services, as well as urban renewal. In 1994, under pressure from the regional and city governments, FS proposed a compromise solution that connected the urban section of the HSR through Campo di Marte to a broader redesign of the rail hub, similar to proposals being considered for Turin and Bologna. This redesign included a new urban boulevard with a tram (then in early planning) on the freed track space. The proposal positioned Campo di Marte as the city’s future central transportation hub. However, while it offers a more straightforward through alignment, Campo di Marte is poorly situated in a mainly residential area with limited redevelopment potential and is inadequately connected to the rest of the so-called Upper Tuscan urban area, stretching from Florence to Pistoia and beyond. Moreover, this proposal would maintain the split service between SMN and Campo di Marte of the 1992 TAV project and would add additional costs and disruption to put a section of the existing 1930s railway belt underground.

The 1994 compromise proposal advanced by FS, combining the HSR routing with an extensive urban renewal plan, similar to what Turin will eventually do with the “Spina centrale” masterplan along its HSR and suburban train combined “Passante”

The years between 1994 and 1999 were crucial for shaping the HSR project. In 1997, the quest for private funding was officially abandoned, and a more ambitious, integrated, and costly AV/AC (Alta Velocità/Alta Capacità, High-Speed/High-Capacity) concept was announced. For Florence, this meant agreeing in 1999 to a compromise solution, inspired by what Bologna achieved in the 1995 "Inter-Government Agreement" between local governments, TAV Spa, FS, and the Ministry of Infrastructure. Florence would have its own “Passante AV,” a through HSR tunnel with an underground station closer than Campo di Marte to the city centre and the existing Santa Maria Novella hub. This new urban route involved realigning the last section of the Florence-Bologna HSR, which would enter the metropolitan area to the North-West, near Castello, rather than along the Faentina railway, as initially proposed. It would use the existing 4-6 track rail corridor from Prato.

The issue is that Santa Maria Novella is probably the most centrally located main railway station in Italy, built within the city's walls (a rarity) and situated in a densely developed corner of the centre, just beside a significant medieval and Renaissance monastic complex from which it takes its name. For this reason, and due to better opportunities for redevelopment, the new HSR station was proposed in Belfiore, approximately 1.5 km north of SMN, in an area along the station's throat that was occupied by non-residential uses, which were deemed easier to relocate, as well as many abandoned buildings.

To address the lack of connectivity with the current station, TAV agreed to contribute to the construction of tramway line T2, which would connect SMN and Belfiore, providing a transit link between long-distance HS trains, calling at Belfiore, and local and regional trains, which would remain at SMN. However, this compromise left many stakeholders dissatisfied—those opposing any dedicated HSR infrastructure through the city due to concerns about the impact of the tunnel and the new large station, and those criticizing the unresolved issue of regional connectivity and the loss of direct city centre HS service.

The final design of the HSR route through Florence as approved in 1999, a compromise solution between the Regional and City governments, TAV Spa and the Ministry of Infrastructure.

Meanwhile, as the construction of the HSR Passante AV was delayed, a temporary solution was proposed in 2005 to address the bottleneck caused by routing the HSR through the existing tracks around the Florence rail node. This measure involves using the existing tracks to reach SMN from the newer out-of-town approach of the HSR north of Firenze Castello. A flyover (scavalco) would resolve some conflicts with local traffic on the busy Northern approach to SMN. At the same time, non-stop HS Milan-Rome trains would use the existing Wye bypass, sharing it with regional trains and freight, while most HS trains would still stop at SMN and reverse there. This is a suboptimal solution that limits the node's capacity due to multiple at-grade conflicts on the station’s throat caused by the in-and-out movements of HS trains.

The temporary solution approved in 2005 for routing the high-speed rail service through Florence pending the construction of the “Passante AV” and Belfiore station.

While the temporary solution was completed on time in 2011, the construction of the Passante AV and Belfiore station, designed by Norman Foster, was delayed during the financial crisis and was ultimately halted around 2015 when the main contractor went bankrupt and costs increased. In the late 2010s, as construction slowed and austerity measures were implemented, critics of the project called for several modifications, and a few alternative options are briefly (but unconvincingly) explored by RFI, including reverting to the 1989 proposal of a Santa Maria Novella underground station, using Campo di Marte as initially proposed in 1992, building a new HSR station at Statuto, or even cancelling the project altogether and making the temporary solution permanent. The institutional actors, including the Region and the City, eventually rejected all these alternative scenarios.

Finally, as austerity eased after 2020, a new, updated project emerged in 2021 to address some of the criticisms of the 1999 project. The Passante AV will be built as planned, but a new local train station, called Circondaria, will also be constructed beside Belfiore to improve direct connections with local and regional trains. Additionally, an airport-style 1 km people-mover will eventually connect Belfiore to SMN. Work resumed in 2023, and the Florence HSR node is now expected to be completed by 2028, about 30 years after the original 1999 project was approved and 40 years after planning began, going through multiple iterations and controversies. The project has now reached a price tag of about €2.7 Bn.

The updated version of the 1999 plan. It is currently under construction and is expected to be completed by 2028.

What does this planning story teach us?

Integrating HSR into cities is one of the most challenging aspects of building high-speed rail, especially in areas with tight physical constraints and no easily available corridors. The final segments, sometimes dozens of miles long, are often the most complex and costly to construct and also the most contentious. The British HS2 project’s staggering price tag reflects the difficulties of carving out a new alignment from Euston through London's sprawling suburbs and green belt, necessitating a costly underground route for more than 20 km since surface corridors are heavily used for local rail traffic. The early 2010s Australian HS rail study envisioned an extraordinary 50 km urban tunnel to reach Sydney Central Station, an even more problematic scenario than London's. The Ahmedabad-Mumbai HS line will traverse the peninsula via a 21 km undersea tunnel. South Korea's HSR has eventually constructed a 50 km tunnel to reach the centrally located dedicated HSR Suseo station. France avoided building expensive urban tunnelled routes by either bypassing intermediate cities altogether or by utilizing existing corridors at the expense of local and regional rail services, which are typically weak outside of Île-de-France. Canada’s still undefined HSR Alto plan has yet to officially address how it envisions connecting to the heart of Canada’s two largest metropolises, with no straightforward corridor to reach Toronto’s Union Station and Montréal’s Gare Centrale, raising the risk that more peripheral, less regionally connected locations might be chosen for the respective HSR stations. Bringing California HSR into San Francisco via the Upper Peninsula, probably the NIMBY-est place on earth and the one with the highest construction costs, making tunnelled solutions impossible, is so challenging that the shortcomings of the compromise “blended” solution have inspired one of the best into-the-weeds transit-focused blogs around.

I can go on forever, citing additional examples from Spain, Turkey, and Germany, or exploring China's strategic decision to establish new, large out-of-center HSR hubs, with the expectation that city centers will shift towards these new hubs, similar to the Mountain and Prophet. To conclude, selecting a viable HSR route within larger metro areas and choosing an optimal urban station location in terms of regional accessibility is a challenge that all aspiring HSR countries should fully understand and openly discuss during the early planning phase, because it is by far what will determine the feasibility and financial success of any HSR project, especially in countries with high construction costs and flawed planning processes that get sidetracked by minutiae while neglecting the fundamentals, a specialty of the Anglosphere. Should we opt for the easier corridor or the one that provides the best regional connectivity, even if it involves higher costs and more disruption? What are the implications of each option? Can we openly discuss these implications with the various stakeholders? Essentially, this is a classic “planning” problem—one of devising possible and technically sound scenarios, transparently weighing the pros and cons of each, and making these trade-offs understandable for the public and the elected decision-makers.