Metro stations are designed around the walk, not the train

What follows is an argument about station design, and about where the received version of it stops being true.

Vertical movement is the bottleneck

A train delivers a few hundred people to a platform in seconds, and stairs, escalators and lifts remove them far more slowly. That mismatch is what produces platform crowding, and it worsens as service frequency improves. Adding trains without adding vertical capacity produces a station that fails at exactly the moment the service succeeds.

This is why station upgrade projects are so often about escalators rather than about tracks.

Pinch points determine the experience

Flow through a station is governed by its narrowest constraint, which is usually a gateline, a stair top or a corridor junction. Widening anything other than the pinch point achieves nothing measurable. Pedestrian modelling exists specifically to identify these, and it routinely finds them somewhere other than where staff expect.

Removing one constraint promotes the next, so station upgrades are sequenced rather than completed, and a project that widens a gateline without checking what sits behind it can move a crowd from a concourse onto a stair top where it is considerably more dangerous.

Counterflow is the expensive problem

People moving in opposite directions in the same space reduce effective capacity substantially compared with separated flows. Separating entry and exit routes is the standard solution and requires space that older stations do not have. This is why some stations operate one-way systems at peak times, which is a capacity measure rather than a crowd control one.

At street level, merging is the other half of the same effect, since two flows joining at a sharp angle interfere far more than two joining gradually, which is why the shape of a corridor junction changes its capacity more than widening it would.

Evacuation sets hard limits

Fire and evacuation standards specify how quickly a station must be cleared, and this frequently constrains permitted occupancy. Deep stations are hardest, because evacuation is upward and lifts are generally unavailable.

At street level, several station capacity decisions that look conservative are driven by these requirements rather than by comfort. The calculation generally assumes escalators are stopped and counted as stairs, and that one route is unavailable because that is where the fire is, which is why a station with two exits has considerably less than twice the capacity of one with a single exit.

Legibility reduces crowding

People who know where they are going move faster and stop less, which increases effective capacity at no construction cost. Clear sightlines to exits, consistent signage and simple geometry all measurably improve flow. Stations that require a map to navigate impose a permanent capacity penalty.

At street level, signage fails at the decision point rather than along the way, so the test is whether the next choice has been answered before a passenger arrives at it, and a sign that only becomes visible once somebody has committed to a stair produces exactly the stopping and turning that the crowd behind then has to absorb.

The exit lands somewhere, and that decides a lot

A station is a machine for putting people onto a pavement, so the width of that pavement, the crossing they meet and the direction the entrance faces determine whether the flow clears or backs up into the ticket hall. Multiple entrances spread that load and shorten walks at both ends, which is why closing a secondary entrance to save staffing costs can degrade a station more than losing an escalator would. Step-free retrofit is expensive because a lift shaft has to reach the platform through everything already built above it, so the binding problem is usually finding a route rather than paying for the lift.

Where step-free access exists at only some stations, the usable network for a wheelchair user is not the map everybody else is reading, and journey planners have only recently begun to show that difference.

The takeaway

Watch the escalator, not the platform. That is where the station's capacity actually is.

The design decision is visible long after the people who made it have gone.