FIT Wellington: Making Tram-Train Work

The potential benefit of a Tram-Train solution is that it would eliminate the need for people travelling through the city centre to change at the railway station. How could a tram-train service work in practice?

FIT estimates that if passengers show up and catch the next train, as people do on a frequent network, over 85% of southbound travellers wishing to go beyond the railway station will have to transfer. Alternatively, if people wishing to go beyond the railway station wait for the next through train to avoid a transfer, it will be full.

Here’s why.

The maximum capacity of an on-street light rail service is about 10,000 passengers per hour. Anything higher requires grade separation, to avoid creating unacceptable delays for other road users. The practical maximum frequency for an at-grade light rail line is about 20 vehicles per hour.

Railway ridership is currently 15,000 passengers per hour and growing. There are 30 trains per hour on the main line and 4 per hour on the Johnsonville line.

A two-car Matangi can carry 1090 passengers, or 930 on an 85% full basis, 2790 passengers in a six-car set.

Assume that 72 metre tram-trains are about the same as a Siemens Avenio, carrying 540 passengers, and run 85% full on average (460 passengers, crowding slows the route too much). This gives an on-street capacity of 9200 passengers per hour. In practice, the maximum length may be 63 metres, as Auckland has adopted, but let us be generous; the maximum vehicle length under the German BOStrab regulations, the on-street basis of tram-train, is 75 metres.

A logical operating approach would be one tram-train carrying 460 people, one Matangi carrying 2790 passengers. The tram-trains run through; the Matangis terminate at the station. This is beginning to resemble the old recipe for rabbit pie: one rabbit, one horse.

Some arithmetic shows that the Matangis are carrying 86% of the passengers. This means that:

• either 86% of people wishing to go beyond the railway station will have to transfer;
• or they can wait for the next tram-train, which will be full, because that’s what happens when you mix high-capacity and low-capacity units on the same line.

However, Matangis are capable of coupling up into 8 car sets: capacity 3720 passengers. So another option is 8-car Matangis and longer platforms. This is a cheap way to increase capacity, because the Matangis are already available. It pushes the rabbit pie ratio down a bit, but not too far, and it pushes capacity up by another 930 passengers on each of 15 Matangis an hour: say 14,000 pass/hr.

In this case, about 89% of passengers wishing to go beyond the railway station will have to transfer.

In principle, we could schedule the shorter tram-train to “shadow” the longer Matangi, say running 2 minutes behind. This would make it more attractive for people going beyond the railway station to wait for a through-running train, without it filling up with other passengers. In practice, this is easier said than done and impractical on the Johnsonville line.

These difficulties all arise from the constraints an on-street light rail line imposes. There is a solution, but it’s expensive: build a metro-style, grade-separated light rail line on its own dedicated right-of-way. This would run underground through the city centre, then elevated through the inner suburbs, at grade in the outer suburbs to the airport. We could then make a tram-train solution work, but at a price: grade-separated light rail costs about twice as much per km as at-grade.

That is, for the price of a metro line from the railway station to the airport, we could build an at-grade line to the airport and convert the Johnsonville line to low-floor light rail.

Copyright © 2024 the Affinity Projects wiki and its authors — Attribution ShareAlike