Day: January 20, 2017

by

Subway Shutdown Plans for 2017

Updated January 22, 2017: Additional illustrations.

The TTC begins its 2017 round of subway shutdowns on the weekend of January 21-22 with the complete closure of the Spadina line north from St. George Station. This is the first of a series of monthly weekend tests of the new Automatic Train Control (ATC) system leading to a planned conversion from Dupont north to Wilson in the fall. This is a prelude to opening the Spadina extension, which has only ATC, not “conventional” signals, to Vaughan at the end of 2017. Most of the closures are on Line 1 YUS because this testing is on top of shutdowns for routine maintenance work.

Two three-day weekends, Easter and Thanksgiving, will see Line 2 BD shut down for work on the Prince Edward Viaduct.

2017ttcshutdowncalendar1

2017ttcshutdowncalendar2

There are important changes for 2017 in plans for shuttle bus operations.

For the Spadina shutdowns, the TTC will not attempt to operate a parallel bus service over the entire route as it is plagued by construction at some locations, and generally provides a long trip south to St. George Station. Instead, bus service will be improved on east-west routes to get people over to Yonge where they can make the trip downtown quickly on the subway. A “parallel” bus service will operate only on the north end of the Spadina line between Downsview and Lawrence West Stations.

2017ttcshutdownspadinamap

In past years, the scope of some shutdowns has been wider than the subway track layout and the locations of turnbacks would imply. The reason for this is that the power feeds to many potential turnbacks are not designed to support operations when power is available on only one side of the location. An example of how this was fixed is at Bloor crossover.

At Bloor, the power feeds are segmented into separate pieces:

  • Southward from the south end of the station
  • Bloor Station and crossover, plus the track to roughly halfway to Rosedale Station, one train length north of the crossover
  • Northward from the midpoint between the crossover and Rosedale Station

This allows trains to terminate at Bloor in either direction even if the power is off south to Wellesley or north to Rosedale. When the line was designed, the crossover itself was the gap between two power sections, and so this could not be a turnback point if power was cut on either side. This change, including the resignalling of the crossover, was installed as part of the North Yonge subway project.

Power supplies at crossovers at Rosehill (south of St. Clair), College and King have not yet been modified, but this is in the plans as part of the ATC project on the Yonge line. The new signal system will include the ability to manage these crossovers remotely as is done at Bloor.

A similar problem exists at some of the older turnbacks on the BD line. Whenever work is underway on the Viaduct, subway service has ended at Pape (although it physically operates west to Chester centre track) because power cannot be cut on the Viaduct without also cutting it at Broadview Station. The TTC will be installing a gap in the power feed just west of Broadview Station so that trains can use this as their terminal. Shuttle buses on Danforth will not be required (only from Broadview westward), and subway riders will have access to the downtown streetcar lines as an alternate route to the core on Dundas and King.

A planned shutdown at Union would normally have required shuttles south on Yonge below Bloor, but subway service will be operated using the crossover at King. This will be done on a manual basis because the new signals and controls for the crossover are not yet in place. This tactic will also be used in 2018 when work at Davisville would have required a shutdown north of Bloor. Instead, the crossover at Rose Hill will be manually operated so that shuttle buses can make their connection at St. Clair Station where there is an off-street loop.

2016 Reviewed

Shuttle bus operations have gradually improved, notably by actively changing the traffic environment in which the buses must operate rather than simply hoping for the best as the streets were. Parking restrictions, signal changes and the use of Paid Duty Officers were much needed changes, although one must wonder why it took the chaos of previous years’ events to show that replacing the subway with buses required a lot of accommodation by users of the streets.

Many types of work were undertaken during the shutdowns beyond the most obvious track and signal upgrades. These are described in the presentation to the TTC’s Board on January 18. Of particular interest was one project that is not illustrated in the report, the replacement of 900m of subway track on the western part of Line 2 with a new form of rail support.

All subway lines built before the Spadina line opened have the track mounted directly on the concrete tunnel floor with only a rubber pad that doubles for both vibration and electrical  insulation. Spadina saw a shift to the “floating slab” technique where the rail is mounted to large concrete sections that in turn sit on large rubber pads. This structure requires a deeper tunnel, but it provides better vibration isolation. Older lines cannot be retrofitted because adding a new layer of support would make trains too high for the tunnels.

The TTC has installed an experimental section of track where the rail is supported from the side rather than from below with rubber pads gripping the track below the head of the rail. The result is a roughly 20db reduction in vibration. This is a trial, and there are currently no plans or funding to retrofit other problem sections of the subway.

The new track suspension system is described on the vendor (Pandrol) website, and illustrated below in photos provided by the TTC.

Updated January 21, 2017

The following illustrations are from a presentation provided by the TTC. The first gives the rationale for use of the Panguard track mounting system and includes a good cross-section of the track suspension technique.

ttc_nandv_whypanguard

The following diagrams show before (yellow) and after (blue) vibration measurements.

ttc_nandv_panguard_results1

ttc_nandv_panguard_results2

by

TTC Vehicle Reliability

Delays in the arrival of the new Bombardier Flexity streetcars, together with last summer’s sauna conditions on the Bloor-Danforth subway, make for ongoing concern about the condition of the TTC’s fleet. Statistics in the January 2017 CEO’s Report triggered media reports and a discussion at the recent TTC Board Meeting.

The numbers, although presented in what is supposed to be an “industry standard” format, lead to much confusion for a variety of reasons:

  • The basic standard is that any fault causing a delay of five minutes or greater counts, while all others do not.
    • A fault that might delay a bus or streetcar (doors not working) may not count against the subway because there is so much redundancy.
    • There is no distinction between a fault that represents a severe failure of a component or a minor annoyance that simply caused a long enough delay to be counted. Similarly, the cost and effort needed to repair faults does not contribute to the metric.
  • Faults are reported “per vehicle kilometre”, but many subsystems fail more on the basis of hours in operation (how long has an air conditioner been running), or number of cycles (how many times did doors open and close).
    • For a specific fleet and type of operation, hours and kilometres are interchangeable because the fleet operates at a consistent average speed within its frame of reference.
    • Fleets (or even subsets of fleets) operating under different conditions (average speed, frequency of stops, loads and grades) will not have the same ratio of hourly-based to distance-based faults. Direct comparison of distance-based statistics between these conditions is meaningless. For example, a well known problem in comparing streetcars with buses is that bus routes tend to operate in suburban conditions at relatively high average speeds. When they shift to more congested, densely used routes, their operating characteristics change. (It is self-evident that fuel consumption is affected by route conditions, and operator wages are paid per hour, not per kilometre. Slower buses run fewer kilometres. Time-based wear and tear, and associated reliability stats will rise when expressed on the basis of distance.)
  • Some fleets are a uniform age, while others are diverse.
    • Toronto’s rail fleets have major vehicle groups each of which was sourced as a single large order: The T1 (BD) and TR (YUS) subway car fleets; the CLRV, ALRV and Flexity streetcar fleets; and the SRT.
    • The bus fleet has a wider range of ages and technologies, and so its statistics are the combined effect of vehicles over a range of ages and conditions.
    • For a list of the TTC fleet by type, see the last page of any Scheduled Service Summary such as the one for January 2017. These are available on the TTC’s Planning webpage.

In the figures reported by the CEO, these issues are not explored in detail, but are at best mentioned in a few footnotes. Unsurprisingly, the media and politicians (even transit pundits) can jump to the wrong conclusion about what the stats actually mean.

To ensure that even without taking these factors into account, we are dealing with similar methodologies for each fleet, I asked the TTC whether the same principles apply across the system.

SM: Is it correct that there is a different set of criteria for a “defect” charged to the streetcar fleets and to the subway fleet? Are the criteria used for buses yet another way of measuring defect rates, or are they the same as for streetcars?

TTC: Same principle applies.  In principle, the calculation of MKBD is the same for each mode.   Overall vehicle reliability is dependent upon component and systems reliability. 

MKBD is calculated from the number of chargeable Road Calls and Change Offs (RCCO) during service.  The definition of a chargeable RCCO is any disruption to revenue service caused by a preventable equipment failure.  This definition is applied to all modes of operation.  It should, however, be noted that there are slight differences to the criteria of RCCO for each mode.   For example, a failure to a set of doors on a subway train may not cause a disruption or a delay to service.  Line mechanics may respond to the failure and barricade the inoperable doors. This may happen with no impact to customer or to service. This is due to the fact that subways have multiple sets of doors that customers can enter or egress from. Transit Control, therefore,  may decide not to remove a train from service if one set of doors is inoperable.  For a 40’ bus, however, the option to continue in service with a set of inoperable doors is not an option. Passenger flow on and off the bus will be significantly impacted.  Therefore, in this case … the same equipment failure may be handled differently on buses, streetcars and subways.   Differences in types of equipment, life cycles of these equipment and operating environments will also contribute to the differences in calculating RCCO and MKBD between modes. [Email of January 16, 2017]

Continue reading →