Thread by @ChittiMarco

1/ Why engaging in the abstract, value-driven debate of LRT vs metro doesn't make much sense?

A little thread about the importance of average travel times and city size when planners choose a transit technology, and how RoW and station spacing impact it. A🧵

2/Let's pick Grenoble's line A, a typical Modern French Tramway: at-grade running with fully reserved right of way and traffic priority, level boarding etc. The 13.7 km of the line are covered in 47 minutes. That means an average speed of 17.5 km/h. Station spacing is around 450m

3/ If one goes through other examples of Modern European Tramways, pretty much finds a consistent average speed comprised between 16-19 km/h with station spacing in the 400-500 meter range.

4/ That average speed of 16-19km/h is the best is possible to get from a fully reserved at-grade right of way with frequent intersections regulated with full traffic priority for tramways. To break the 20km/h threshold you must reduce conflicts with further grade separation

5/ That is why it's common wisdom in France that for trips stretching more than 8-10km out fo the core the tramway looses its attractiveness because trip times become too long. In fact, almost no French modern tramway line is longer than 20km, with most being in the mid-10s range

6/ Some may say: yes but tram-trains!
Well, yes, but...

Take S5 in Karlsruhe. The "train" section between Pforzheim and Berghausen is some 21km and takes 27 minutes. That means an avg. of 46km/h, but with an average station distance of 2 km, 4-5 x longer than urban tramways

7/ Moreover, stops are not evenly spaced but there are longer gaps of up to 4 km, as the S5 line stops multiple times in villages along the line, then speed up between them. Would this pattern of uneven stop works in a evenly low density suburb in US or Canada? I Doubt so.

8/ For example, Most of the planned or u/c LRTs around Toronto metro area have consistent stopping pattern that follow the even grid of arterials. Finch LRT in Toronto has an avg stop spacing of 600m, and thus a planned average speed of 17.3km/h,the "magic number" of modern trams

9/ Some US LRTs, that in outer sections are somehow very similar to a tram-train type of alignement, ends up taking badly located old rail freight corridors to achieve tram-trains level of service speed, but without the clustering of density in compact villages that Karlsruhe has

10/ And, without surprise, the urban section of Karlsruhe's S5 has a more typical urban tramway speed of 19.5km/h, covering the 13 km within Karlsruhe and its inner suburbs in 40 minutes, stopping 22 times (590 m avg distance between stations)

11/To break the 16-19km/h rule of average speed for modern tramways, the only option is to increase station spacing and/or grade-separate the RoW as much as possible to avoid delays and slowdown occurring at intersections, even when light priority is in place

12/ Yet, the very problem of building LRTs designed and operated as modern EU tramways in the vastly sprawled suburbs of, say, Toronto is that they end up having performances similar to buses with reserved lanes along arterials, but with the rigidity of tramways

13/ Efforts to grade-separate the RoW further with tunnels, overpass, viaducts, in order to increase average speed, ends up bringing the RoW cost up toward the area of light automated metros, but without having the benefits in terms of capacity, frequency and lower operating cost

14/ Moreover, to really hit the >30 km/h average speed typical urban rail operation (most metros have average speeds in the 30-40km/h bracket, depending on station spacing) along suburban arterials, one ends inevitably up with a fully grade-separated tunnel or elevated alignement

15/ The reality is that, the vast distances of large metro areas, especially North-American ones, make tramways running at-grade on suburban arterials not very competitive, in the best case scenario, as beyond 8-10km, at-grade tramways loose any competitive advantage with cars.

16/ That said, Euro-style tramways may have their place in NA too, but mostly on heavy loaded corridors in the relatively dense inner core of large metro areas in the N-East or for growing cities aiming to strengthen and densifying their core, like Waterloo and Québec

17/ There is a reason why small metro areas like Calgary and Edmonton are increasingly grade-separating their LRTs new extensions, as this is the only way to achieve higher average speed, and not exclusively, as some advocates argue, because they don't want to upset drivers.

18/ On that line of thought, what ultimately killed many vast networks of interurbans, like LA one, was the incapacity to convert them in faster metro/suburban trains. That is instead what happened in another metro area that first boomed also around interurbans: Tokyo

20/ With the car being able to achieve higher average speed on longer trips, rail based transit that is unable to cope, by offering similar or at least globally competing travel times, taking into consideration all the other factors, is not going to succeed.

21/Transit advocates should try to go beyond simple heuristics like "slow is good", "sharing the space" to pick a preferred mode and realize the fact that either transit is competitive on average trips time (not top speed!) or it will always be marginal in a certain O/D market.

A little real real example as addendum: this is the planned speed graph for Bologna's future tramway line that will have a... 17.6km/h avg (magic number!).
Top speed is the same of cars (50km/h in the city, 25km/h in the city center). >>

>> traffic light priority still requires crossing intersections at 30km/h max (hence all the intermediate dips to 30km/h in between stops. A few larger intersections with particularly complex signal cycles will require up to 10seconds stop from tramways.

the whole report about the planned operations of the line is available here. It is a good example of typical assumptions made for urban tramways operations that make 16-19 km/h the typical avg. speed: drive.google.com/drive/u/0/fold…