The title of this post is slightly misleading. It should rather be “Lack of Level Crossings”.
Metro Railways and Level Crossings are not necessarily “natural enemies” – they just don’t fit together very well. It’s a bit like wanting to have a pedestrian crossing on a freeway. Two things happen, or maybe three:
- It is dangerous for pedestrians to cross the road.
- The performance (speed) of the freeway is seriously compromised.
- When trying to avoid performance loss of the freeway, it gets all but impossible for pedestrians to cross.
That’s actually a pretty good analogy. A main feature of metro railways is high frequency of train services. If you have trains coming every two minutes or so, and the same in the other direction, there will be not enough time to block the railway line for crossing car traffic without disrupting the rail services. In Perth, for example, a notorious four-track level crossing close to the city centre will be closed for special events as trains will have to run so often that the crossing would be almost permanently closed for cars anyway.
Metros that are built as metros don’t have level crossings in the first place. Think of underground railways in tunnels – no level crossings, obviously. But what about existing “brownfield” railways to be upgraded to metro-style service? The best way is to get rid of any existing level crossings. This is just happening in large numbers in Melbourne, where the rail network is most notorious for its many level crossings. The more trains are running, the shorter the opening times for car traffic become, so it’s no fun for either party.
Understood that the removal of a level crossing, replacing it which a bridge or underpass, is a costly exercise. But it is probably still less expensive than building an entirely new metro railway without level crossings.
A word on keeping level crossings. Because there are hardly any examples (actually none I can think of) of metro railways with level crossings, suppliers of metro-suitable technology (e.g. CBTC signalling) don’t have experience with integrating level crossings, and they don’t like it and probably don’t do it very well when forced. The best solution here is to simplify the “integration” of level crossings, avoiding complicated interfaces in an attempt to “improve” level crossing operation by means of a signalling technology which was neither conceived nor developed for dealing with level crossings in the first place.
