A Very Bad Day on the Subway

Wednesday, December 9 was the first “snow day” for the TTC of the 2009-10 season.  Although I’m now retired and should have stayed in bed listening with glee to the traffic reports, I bundled up and rode over to Bloor-Yonge Station to watch the morning rush hour with the new crowd control setup.

It was not pretty.

The crowd control actually achieved its purpose in spreading out the load on the southbound platform, but the service was a complete mess.  The TTC had signal problems, service interruptions due to smoke at track level (more about this later) and a number of passenger assistance alarms (PAAs) brought on by people feeling unwell or fainting in crowded trains.

A log of my observations shows the wide gaps in service with headways rarely below 4 minutes. During the two-hour period from 0800 to 1000, the TTC managed to get only 26 trains through the station, slightly fewer than they would normally operate in the peak hour.

Traffic was heavier than usual with trains arriving southbound quite full of passengers.  However given the gaps in service, it was impossible to know if this was due to heavier demand on a snow day or simply the backlog of riders.  Passengers transferring from Bloor-Danforth made their way well down the platform, and the south end was often more crowded than the north end.  Even when the crowd was backed up on the platform beyond the pillars (roughly half of the platform depth), transfer passengers from BD flowed fairly freely behind them.

TTC staff adjusted their tactics to suit the changing situation and on one occasion sent passengers transferring from the Yonge to the Bloor line the “wrong way” through the passageway to the north concourse to avoid the congestion on the main part of the platform they would normally use.  This sort of flexibility and “on the spot” judgement about routing pedestrians is vital to the scheme, and will be part of the design considerations for any sort of “permanent” installation of barriers.

A sharp-eyed trainspotter can keep track of the approaching service using the “next train” time indications.  When these change infrequently, the next train is spending a lot of time at stations or crawling between them.  Given that signal problems slowed trains, and passenger congestion extended the dwell times, it’s hard to know which condition had the greater effect.

Dwell times at Bloor were appallingly long, and few trains achieved under one minute dwells.  The TTC has cut back on platform assistants, and this really showed because several trains had problems getting doors closed on the first attempt.  Moreover, some of the PAs held back from the crowd rather than being trapped between them and the platform edge.

A delay at Pape caused by a smoke observation shut down the entire BD line from about 0824 to 0841.  This choked off transfer traffic, and the YUS caught up with a “gap train” (empty train arriving express from Davisville) clearing the platform at 0830.  However, once the BD delay cleared, things on the YUS level became congested, and the platform was not cleared again until after 1000.  If the BD line had run normally, the platform at Yonge would likely have been overwhelmed.

These smoke delays are becoming quite common, although you would never know it from the TTC’s eAlerts.  Far more info is available on the TTC’s Facebook page.  According to that page, there have been three smoke delays so far today (1020), three yesterday, three Wednesday.  I am still waiting for the TTC to provide information on what is happening and why these delays are so frequent.

TTC’s eAlert system has been more or less missing in action.  The only alert on Wednesday was for a derailed streetcar at College and Ossignton, and it has been completely silent otherwise.

Overall, my impression of Wednesday’s operation was that the crowd control system together with the 20-minute shutdown of the BD line kept the situation at Bloor Station from completely falling apart.  This shows the importance of everything working as one system, and how badly things can go awry if any part of that system is unreliable.

A related issue is headroom, the spare capacity needed to absorb unexpected problems and surges in demand.  We hear a lot about the demand the YUS might carry, but that only works if the line is much, much more reliable.  All the signals and automatic train operation are worthless if there are regular delays caused by smoke or door problems or ill passengers from overcrowding.

We are trying to jam more and more people onto a system that was not designed for these loads, and whose maintenance philosophy appears to tolerate random service disruptions as a normal part of operations.  The more important any one component in the network becomes, the more important that it work reliably regardless of the weather.

The coming TTC operating budget debates will no doubt include the usual calls for belt-tightening, but we can already see that the TTC is falling behind in system reliability.  The debacle of the mid-90s must not be repeated, and the TTC must operate good, reliable service rather than falling back on “snow” as a catch-all excuse.

The Effect of Rapid Transit on Local Shopping

A few weeks ago, Stephen Rees Blog in Vancouver ran an interesting piece on the effect of the new Canada Line (the one connecting Vancouver Airport and Richmond to downtown) on local shopping neighbourhoods.  Since bus service on the former surface routes has been cut, merchants are concerned that they get less walk-in trade from bus passengers.

The comment thread following the article requires some knowledge of Vancouver’s neighbourhoods and geography, but includes a variety of viewpoints on this phenomenon.  One point comes right at the end of the thread in a comment about the reconstruction of Cambie Street, under which the Canada Line runs:

Cambie Street between 16th and King Ed was completely rebuilt at great tax payer expense. What an opportunity to create a “Great Street” and boost neighborhood identity!

Instead, the streetscape design is the worst possible for enhancing the Cambie Village experience. Six lanes of traffic without separating medians, with curb side parking taking up the outer two.

My favourite symbol of the lack of capacity shown in this area of design practice is the introduction of “park benches” on the sidewalks facing parked cars.

We hear a lot about urban design in Toronto and the improvements possible as an offshoot of the Transit City projects, and the Cambie experience should be a warning of what can happen.

Rees’ article talks about the problems of merchant pressure for parking taking precedence over transit, and this effect can be seen on St. Clair where parking was one of the car-oriented street functions that forced an uncomfortable design onto the street overall.

He makes an important point about transit, namely that it is part of a pedestrian experience:

Every transit trip is an interrupted walk. Transit stops and stations ought to be seen as key to retailing. Far too often in Greater Vancouver bus passengers are banished to remote, sterile areas like Phibbs Exchange, or the Ladner bus loop. Always this is forced by local merchants who have only contempt for what they see as the low income bus passenger, and who regard buses as noisy, smelly nuisances. Of course, transit’s selection of large diesel buses only confirms that view. We do have to learn from our experiences, and acknowledge our mistakes. Far too often, transit advocates are expected to be cheer leaders for a system which, sadly, often lets us down, and seems incapable of learning from its past mistakes. Let’s all learn from this when we design our next system change.

As one who is often expected to cheer for transit plans and hope that we will fix the design problems “later”, I can only say that the time for believing planners when they say “trust me” is long over.