Since World War Two, some cities have sought to extend rapid transit into their suburbs by leveraging legacy commuter rail lines. Building on prewar examples from Berlin and Tokyo, they initiated a variety of treatments to modernise their commuter rail: electrification, integrated fares, high all-day frequency, and cross-city connections.
All this turns commuter rail into an express metro line. The city that has done the most in this direction is Paris, which since the 1970s has built a network called the RER, with five lines labeled A through E.
It is the cross-city connections that are the costliest to provide, since they almost always involve new tunnels under city center. Cities can build cross-city tunnels in two ways. One approach involves high investment: the tunnels are longer and involve several stations, often in difficult-to-construct locations. The main example is the RER A, whose construction involved about 17 km of new tunnel and seven underground stations, running on an east-west axis through central Paris.
The other approach is lower-investment: tunnels are the shortest possible connecting commuter rail terminals. The main example is the RER C, whose construction involved just 1 km of new tunnel and no new stations, creating an southeast-to-southwest line on the Left Bank of Paris.
A geographically accurate map of the RER network in central Paris. RER A is in red; RER C is in yellow. Image: Wikimedia Common.
This is a spectrum rather than a binary division: RER lines B, D, and E are intermediate between the high investment that went into the RER A and the low investment into the RER C. In layout, the RER B is quite similar to the RER A, but managed to leverage a legacy line reaching within 2 km of city center.
The same division between the two approaches holds outside Paris, too. In London, Thameslink is similar to the RER C, whereas Crossrail, with its long new tunnels, is like the RER A, as is the planned Crossrail 2. Berlin's North-South Tunnel from the 1930s, creating a new axis in the city complementing the older east-west Stadtbahn, is like the RER A.
North American projects, including the SEPTA Regional Rail tunnel in Philadelphia and the ongoing Toronto RER project, are both like the RER C. The Regional Rail tunnel connected two commuter rail terminals to create a mainline the shape of an inverted L, with some lines self-intersecting. Toronto is fortunate enough not to need new tunnels at all, since all commuter lines serve Union Station, some coming from the east and some from the west.
The main benefit of the RER C style is that it is much cheaper. It involves less tunneling, and the city can choose to build fewer stations. When tunneling deep underground, the stations are the most expensive element: for example, in New York's Second Avenue Subway, built deep to avoid street disruption, the tunnels cost $415m whereas the three new stations cost $2.2bn total. The central segment of the RER A cost about 5bn francs, corresponding to about €600m per kilometer in 2016 prices; no other rail tunnel in the world has cost so much except some New York lines and Crossrail. Crossrail, the other major modern example of this type of construction, is even costlier, perhaps £750m per kilometer.
The main benefit of the RER A style is that it lets commuter rail act as an express metro line. Such tunnels do not follow the shortest path between legacy terminals: both the RER A and Crossrail were designed as express east-west lines through city center, with stations connecting to most intersecting Metro or Underground lines. And they are not just commuter rail schemes but also relief lines for the busiest metro lines, namely Metro Line 1 and the Central line. RER C-style lines do not necessarily provide this: the RER C is parallel to Metro Line 10, the least busy in Paris.
Another metro-like property of the RER A is that it has a long trunk segment providing high frequency. This is also true of the RER C, but not necessarily of other RER C-style lines elsewhere. Thameslink's shared trunk is short, just between King's Cross and Blackfriars, and SEPTA's trunk is only a few kilometers long. This happens when a short tunnel connects to many commuter rail branches.
Evidently, the RER A style leads to higher ridership: current ridership on the RER A is about 1.1m per weekday (see page 24 of this PDF); that on the RER C only 540,000. This is despite the fact that the sprawling, many-branched RER C is almost twice as long as the RER A.
The proposed New York-New Jersey Crossrail.
In North America, proposed regional rail modernisation projects fall on the RER C side. In New York, the Regional Plan Association has proposed using the planned new tunnels under the Hudson River to build a New York-New Jersey Crossrail project. The RPA is not planning on any new stations to connect to subway lines that have no connections to the existing Penn Station.
And in Chicago, the Midwest High-Speed Rail Association has proposed reactivating through-tracks at Union Station to create a Crossrail Chicago. The plan only includes one new urban station and has no transfers to the busiest L lines. In both cases, the Crossrail name does not imply service levels comparable to Crossrail: the routes are awkward, kludged together from the available commuter rail lines.
In Boston, plans for the North-South Rail Link are more mixed. This project would provide new tunnels connecting the city's two rail terminals, North Station and South Station, which are about 2 km apart. One RER A-style feature of the plan is that, in addition to these two stations, there are plans for one intermediate station, called Central Station (Boston's central business district stretches roughly from South Station to the planned new station). North and South Station together connect to three of Boston's four subway lines, and Central Station would connect to the fourth.
The proposed Boston North-South Link.
For a city planning to modernise its commuter rail network with new tunnels for through-running, there are merits to both models: evidently, Paris built the RER C and not just the RER A. However, it is a mistake to assume that short tunnels could provide the benefits of the RER A or Crossrail. In New York and Chicago, if there are plans to through-run trains, their respective transit agencies should at least consider adding stations to intersect more subway or L lines, or even the busiest bus corridors. For example, New York could open a commuter rail station at Astoria and, when the new Hudson tunnels are built, at Bergenline Avenue. Toronto is fortunate not to need tunnels, but it should consider adding infill urban stops on the planned RER to relieve the city's two main subway lines.
The biggest cities should probably plan on at least one RER A-style commuter line. London came to this conclusion when it began the Crossrail program; despite the high cost, it is now very likely to build Crossrail 2. The largest North American cities should learn from this and consider some truly metro-like commuter lines rather than just lines in the mold of the RER C.