Metrolinx New Stations Report: The Details

In a previous post, I reviewed the updated evaluation of proposed new stations on the GO/RER/SmartTrack network. In brief, the situation for some locations is not as dire as in mid-2016 because Metrolinx has changed some of the operating rules and plans for it services. Whether the newly proposed services can actually be operated remains to be seen and is, as usual, a subject for further study.

This article is a station-by-station review of the primary issues at each proposed new stop. The stations are ordered here by corridor for ease of reference by geographical grouping, whereas in the Metrolinx report they are in alphabetical sequence.

(There will doubtless be a small industry in pushing for reviews of stops that are not in the Metrolinx list. That is not the purpose of this piece which reviews the updated evaluations as presented by Metrolinx.)

My apologies in advance for a long, text-only read. There were no illustrations beyond general maps in the Metrolinx report, and so there are none here either.

There is a series of planned public meetings about SmartTrack stations, and it is possible that these will include more details of current designs. If so, I will update this article to include them.

Lithuanian House
1573 Bloor Street West
6:30 pm – 8:30 pm
Presentation begins at 7:00 pm

Scarborough Civic Centre,
Council Chamber
150 Borough Drive
6:30 pm – 8:30 pm
Presentation begins at 7:00 pm

Queen Alexandra Middle School,
Small Gym
181 Broadview Avenue
6:15 pm – 9:00 pm
Presentation begins at 7:30 pm

A total of 17 stations are reviewed in this study. Of these, 5 were not recommended in the initial report in 2016. Of these, only Park Lawn has been resurrected to go forward for reasons discussed later. One of the 12 in the approved list, Mulock, has negative benefits and might fall off of the table if Metrolinx cannot find a way to make a better case for it.

General issues that are either not addressed by or not detailed in the report include:

  • There are no detailed design drawings of the stations, only very general location maps.
  • Details of the service plan(s) used to model demand. There are some specific references with respect to express and local operations at certain stations, but not for existing stations or the network as a whole. This affects demand modelling.
  • Modelled demand at all stations, not just the new ones, and of the cumulative on-train loads. This is important to ascertain whether the planned service can actually support the projected demand.
  • Details of the boardings and alightings at stations. Combined values are shown, and the descriptive text indicates which is the predominant flow, but not the proportions.
  • Differentiation of new riders attracted to GO service by the station as opposed to existing riders diverted from nearby stations (i.e. net new ridership).
  • The degree to which, if at all, performance improvements through electrification (whether by conventional power of hydrogen fuel cells) will offset the time penalty associated with new stations.
  • Additional infrastructure required for express and local operations to co-exist on each corridor. Some of this is mentioned, but not in a comprehensive way.
  • Details of train operations including use of express and local tracks, and track assignment on corridors with multiple services. Any requirement for individual services to cross each other affects capacity along the route and at Union Station.
  • Details of the implications for freight operations both with respect to existing spur lines and to clearance issues with new structures.
  • The anticipated volume and operational interference of freight operations on GO’s passenger service.

For the original station designs which, in some cases, have now been modified, please refer to the Metrolinx mid-2016 reports. Go to the Metrolinx New Stations page, scroll down to and open the Initial Business Cases bullet.

A consistent problem through all of these studies is the reliance on the imputed value of time savings to travellers. This is not “real money” in the sense that it can be recouped to pay for the transit investment, but a social benefit that transit confers. There is nothing wrong with this outlook, but readers are cautioned that when Metrolinx speaks of benefits exceeding costs, this does not mean that profits will roll in the doors at stations. Moreover, the model is very sensitive to the imputed effect of delays caused by new stations.

In their attempt to address the negative effect of adding stations to the corridors on riders making long trips, Metrolinx has changed their service design to include express and local trains. This fixes one problem, but adds others in terms of the resulting frequency at local stations, and the capacity of local trains to handle the projected demand.

All demand numbers cited here are for the 2031 projection which assumes the current fare structure with GO/TTC co-fares, but no “regional integration” beyond what is already in place:

The PDBC analysis assumes:

  • introduction of Presto on all TTC services across the City of Toronto;
  • the current discounted double fare agreement between the City of Toronto and Metrolinx – a $1.50 discount is applied when an adult Presto user’s journey includes both a TTC and GO segment;
  • the planned TTC 2-hour transfer to make the TTC more aligned with 905 transfer policy, planned for implementation in August 2018; and
  • progress by all transit agencies on addressing removal of fare barriers and improved service integration.

As a starting point, the base fare structure as of December 2017 is assumed for the PDBC analysis. [p 12]

Mayor Tory has trumpeted this report as showing a strong support for his SmartTrack project with 60-year benefits of $4.59 billion greatly exceeding the capital costs of $1.195 billion (2022$). However almost all of the benefit comes from two stations – East Harbour and King-Liberty.

East Harbour provides 55% of the demand and 84% of the imputed benefits from the six SmartTrack stations. King-Liberty adds a further 16% of the demand and 9% of the benefits. These stations stand on their own as worthwhile additions completely separate from SmartTrack.

Stouffville SmartTrack Corridor

Three stations were proposed for the Stouffville Corridor within the City of Toronto, of which two survived the cut in 2016 into more detailed analysis. They exist because of their role in the SmartTrack and Scarborough Subway network.

The modelled service pattern for this corridor is:

All-stop peak direction outer service between Lincolnville and Unionville stations; trains will also stop at Kennedy and East Harbour stations, otherwise, express to Union Station. “Inner” services will stop at all stations between Unionville and Union Station. [p 6]

Correction: References to service every 20 minutes by the local trains has been corrected to every 15 minutes based on the service plan shown on the RER website for this line.

Finch-Kennedy Station

Three significant changes have occurred in the design of this station:

  • The proposed platforms are moved further south to allow a road underpass which would include the link between bus service on Finch and the GO station above.
  • The parking and passenger drop off area has been eliminated.
  • This is a local station with trains every 20 15 minutes unless Metrolinx comes up with a scheme for improved local service in the corridor. It will not be served by express trains to or from stations north of Unionville (the north end of local service).

The grade separation will affect neighbouring properties, roads and a stormwater facility . The new structure would protect for a future express track.

There is a strange remark that “Coordination is required with electrification works due to substandard clearance under the bridge”. This does not make sense if the road is going under the railway, and the text may be left over from a previous version of the design.

The AM peak demand of 1,100 is expected to be predominantly from boarding passengers who will arrive by transit, walking or cycling. All day demand is forecast at 4,200.

The total benefits (imputed value of travel time reductions, etc.) are $16 million measured over 60 years. In 2016, the capital cost was estimated at $108.9 million, and this value has gone up according to the 2018 report.

It is difficult to understand why this station survives other than its role as part of SmartTrack.


Ellesmere station would occupy the space now used by the SRT’s station of the same name, and it would be a local station. The projected capital cost in 2016 was $23.2 million, and the valeu is now higher. Track construction for the Stouffville Corridor is already underway in this area and would have to be modified to accommodate a station. No new GO station can be built here while the SRT is still operating.

The link with the bus service on Ellesmere is poor, and this is a very lightly used station on the SRT. In spite of the limited usefulness of the station, the projected demand is similar to that at Finch-Kennedy Station (1,200 AM peak; 4,600 all day).

This station was not in the original list, and its status has not changed.

Lawrence-Kennedy Station

Lawrence-Kennedy station is in the same general location as the existing Lawrence East SRT station, and this has been the source of some controversy between, broadly speaking, the transit advocate community and various politicians and professionals. The basic problem is that the two structures cannot co-exist because there simply isn’t enough room under the Lawrence East overpass for the SRT station and bus loop, two GO tracks and a GO platform.

Since the original design in 2016, there have been several changes:

  • Provision for co-existence with the operation of the SRT. That this must be done confirms that the original design could not exist with the SRT still in operation, and this condition should have been flagged and acknowledged in 2016.
  • Because the SRT isolates the railway line, there would be vertical links to bus stops on the overpass above the station. Once the SRT structure is removed, the GO station would be accessed from the existing bus loop.
  • The GO station will have two side platforms, and the easterly one will require an underpass just as is the case today with the SRT station.

There are many problems with building a GO station while the SRT remains in operation, but the political imperative is that this somehow be accomplished as the ability to use the SRT right up to the last moment was an essential part of the sales pitch for the Scarborough Subway. At a summary level, the story is that the City and Metrolinx are working on a “fix”, but there are many issues. The following points are quoted from the Metrolinx report :

  • Prior to removal of Line 3, access to west platform would be from the Lawrence bridge overpass and the east station platform, but not accessible from the area of the current Line 3 station. The station facilities and tracks will block direct access to the east until Line 3 is decommissioned.
  • Exemptions will be required from Transport Canada and GO Design Requirements Manual clearance requirements due to the narrow space between the Line 3 station and the existing Lawrence bridge piers.
  • Constructability and construction staging issues could be addressed by not commissioning the second track until after substantial completion of station construction, however, this would impose various operational challenges.
  • In order to accommodate ongoing Line 3 operation, tangent track cannot be achieved without property and business impacts. The proposed ‘S’ curve to the platforms, may create uneven gaps between the train and platform, and reduce end-to-end visibility, requiring mitigation to improve safety.
  • Tracks work must address turnouts to industrial spur lines approximately 370m, 425m, and 490m north of Lawrence Avenue.
  • A crash wall at the Lawrence bridge piers is under construction through the Stouffville corridor expansion contract (double tracking). The crash wall will fall within the proposed east platform of the Lawrence-Kennedy station blocking the main access route to the platform in addition and impeding pedestrian circulation. Other derailment mitigation measures should be considered. As trains will be travelling at reduced speeds the platform could be designed to act as a guide/barrier and/or track guard rails installed eliminating of the need for a crash wall.
  • Coordination is required with electrification works due to substandard clearance under Lawrence bridge. [Appendix I, pp 33-34]

These are not trivial issues, and it is clear that the Lawrence-Kennedy station will be a pinch point on the corridor, and may not be useable, until after the SRT station is demolished. This implies a transition period where no station may exist at this location, but that issue will have to be sorted out in detailed planning and staging for this site.

The projected AM peak demand here is 2,400 passengers, with an all-day demand of 9,200. Demand is primarily oriented to downtown, but some riders would travel north into Markham. This is a local station.

The user benefits over a 60 year span are estimated at $69 million. The cost of the station was estimated at $23.2 million in 2016, but this has since risen and is greater than the projected benefits.

This station exists as an alternative to the now-eliminated Lawrence East station on the Scarborough Subway, but service here is currently planned to be only every 20 15 minutes (the local trains) unless Metrolinx can find a way to fit more service into the corridor.

Access to Scarborough Town Centre

Passengers wishing to travel to STC from the GO train would most likely alight at the existing Agincourt station (Sheppard Avenue), although this would not be an express stop. They would take an express bus from Agincourt east and south to their destination. This arrangement means that STC-bound riders from north of Unionville would have to transfer to a local train that stopped at Agincourt. The same challenge faces riders bound for bus services on Steeles Avenue connecting at the existing Milliken Station.

Lake Shore East Corridor

Whites Road Station

This station would be located between the existing Pickering and Rouge Hill stations. It would be a local station with considerably less service than would be available nearby. Projected demand is 1,200 for the AM peak, and 3,500 all day.

The station would generate $73 million in benefits over 60 years, but its cost (unspecified) would exceed its benefits.

This was one of the five stations omitted from the 2016 list, and its status has not changed.

Gerrard-Carlaw Station

Correction: The original reference to peak service every 20 minutes has been changed to every 15 minutes to match the service plan shown for the Stouffville corridor on the RER website.

This station would be located at Gerrard Street between Pape and Carlaw Avenues. This is also the proposed site for the Gerrard station on the Relief Line, and provision would be included in any GO station design for a link to the future subway station. The area around the station is now primarily low rise residential, but there is the potential for substantial growth as this is an “Avenue” in the Official Plan, and development applications are already in the works to intensify land use in the neighbourhood.

Since the original design, some changes have occurred:

  • The station has been shifted to better fit with other future uses including the subway station.
  • An onsite bus/streetcar loop and drop-off area have been removed to reduce property requirements.

The projected demand is 3,500 in the AM peak and 13,500 all day. This is modelled without the Relief Line in place. Comments in the report suggest that this station would be served only by the local service on the Stoufffille/SmartTrack line, not by express trains nor by the Lake Shore corridor trains. The level of demand for a 20 15 minute service is hard to believe, although the station is thought to attract riders who now might alight further east (for example, transfers to 506 Carlton would be simpler here than at Main/Danforth station, and 505 Dundas would be extended east to this station in the original plan).

The capital cost was estimated at $251.7 million in 2016, but the value is now lower. There are $138 million in projected 60-year benefits, but this is less than the revised cost of the station.

Gerrard will be a difficult station to build, and a strong argument can be made that any structures needed for the subway stop be built at the same time, if at all possible, to avoid future disruption. Indeed, one might argue that this location would be better served with the Relief Line stop rather than as part of GO/SmartTrack, but that configuration was not what Metrolinx studied.

East Harbour Station

East Harbour Station (formerly known as “Downtown East”) is the star attraction among all of the new stations for the simple reason that it sits in the middle of a very large proposed development site on the former Unilever lands at Lake Shore East and the Don River. All trains will stop here.

Since 2016, the design has been modified:

  • Side platforms have been added for both the Stouffville and Lake Shore East trains.
  • The platform has been shifted between the Don Yard and Unilever sites.
  • Provision for the Broadview Avenue extension has been added.

The station description show that both side of the Don River are to be served.

The area is an emerging transit node within a developing and intensifying area. There are planned connections to the Broadview streetcar extension and future Relief Line. The station platforms and path connections are planned to act as a bridge between the west and east side of the Don River, with entrance structures anchoring each side. Cycling facilities adjacent to the station would service two distinct, but complementary, functions by providing local connections to and from the station and by forming part of the larger cycling network. A multi-use path connection from the west Don River entrance across to the main entrance can be incorporated to connect to the Don Trail. [p 9]

Projected demand is 17,700 in the AM peak, and 68,000 all day. The majority of the AM peak riders are destined for the station as opposed to beginning their trips from surrounding residential areas or connecting transit services. The model does not include the Relief Line, and so it is possible the demand shown here includes riders who would otherwise arrive directly on the TTC rather than on a GO train.

The capital cost was projected to lie between $160 and $290 million in 2016 depending on the design, and it is unclear whether any of the earlier designs is a good comparison for what is now proposed. The cost is now higher (although relative to which 2016 number is unclear), but the benefits (overwhelmingly travel time savings because this area is now very hard to reach) are almost $4 billion.

The station is on a tricky curving site, and many other construction projects will take place around it including the East Harbour redevelopment, the flood protection land form and the rebuilding of the Gardiner-DVP ramps, among other conflicting works.

Weston/Kitchener SmartTrack Corridor

The service plan modelled for this corridor is not specified in the report, although there is reference to express service in the corridor limiting the effect of a new station on trips by riders destined for Union. This implies that King-Liberty, despite its high demand, would only be served by local trains.

King-Liberty Station

This station is located north of King Street. Changes since the original June 2016 design include:

  • Stopping service has been reassigned to tracks 1 and 4 from the original plan of 1 and 2, and the station now uses side platforms. This leaves tracks 2 and 3 free for express services.
  • The station is shifted north to handle the extra width of separate platforms and track geometry changes.
  • Link overhead to the “King Highline” multi-use trail.

This is a high density area with new tall residential buildings stretching south from Queen Street to the Lake Shore West corridor.

The station would be used by 5,100 riders in the AM peak, and 19,600 on an all-day basis. However, AM peak demand at King-Liberty is expected to be primarily from passengers alighting to access jobs in Liberty Village, and most of these riders would be existing GO Transit users.

The projected cost in 2016 was $30.8 million, and the value has increased since then by an unknown amount. The projected benefits, mainly in travel time savings (without the station, GO riders would have to travel to Liberty Village on local transit from Union or Bloor stations), would be $426 million over 60 years.

St. Clair-Old Weston Station

This station is located at St. Clair where the rail corridor crosses St. Clair just east of Keele Street and Old Weston Road. The existing bridge, underpass and surrounding streets are part of a redesign process now underway to improve capacity of the street including the pinch point where the 512 St. Clair streetcar occupies the central lanes of the four-lane underpass.

Projected demand is 2,300 in the AM peak, and 8,900 all day. The report claims that riders would be attracted to this stop from as far east on St. Clair as Dufferin Street, although this is hard to believe except for counterpeak travel (e.g. AM peak trips from St. Clair West to destinations on the Kitchener corridor). It is unclear how much of the corridor service would stop here.

In 2016, the projected cost of this station was $27.4 million and this has since risen. This cost is now shown as less more than the 60-year benefits of $89 million indicating a substantial jump in the capital cost.

[In the original version of this post, the word “more” above was “less” which was not as intended.]

Construction at the station will affect adjacent property and the CP rail line which lies immediately east of the GO Transit corridor.

There is an odd comment in the description of the station’s effects:

The Kitchener corridor includes express services, which serves to limit the delays experienced by upstream riders at the station. However, upstream impacts remain higher than station user benefits in all but the Fare Integration scenario. [Appendix I, p 45]

If there are upstream impacts, this implies that express trains will stop at St. Clair, although the effect could be mitigated by other changes in the service plan. If the effect of this stop is higher than the user benefits, then this site should have a negative, not a positive, “benefit”. If a “Fare Integration scenario” is needed to overcome this, the question becomes what scenario does the positive benefit for this station actually represent. Either the station has a benefit, or it does not, but the conditions under which this occurs should be clear.

The merit of this station is dubious, and it exists only because it is part of the SmartTrack scheme.

Breslau Station

This station would be located east of Kitchener-Waterloo adjacent to new development in the Township of Woolich south of Highway 7. About half of the adjacent land is an environmental protection area and would not be developed.

The projected demand is 1,100 riders in the AM peak and 3,100 all day. The primary function of the station would be as a park-and-ride site for travellers from the KW area with “very few walk or transit access trips” to the station.

The benefit value of $286 million is driven by the reduction in travel time for commuters who would use the station, and this exceeds the capital cost, although that value is not known.

This section of the line is now single track, and the station design must take into account the future addition of high speed service that will not stop here.

Lake Shore West Corridor

The service plan modelled for this corridor is:

Alternating trains with bi-directional 15 minutes service on the corridor with stops at Mimico and Park Lawn stations. Mimico and Park Lawn stations would therefore receive 30 minutes service inbound and outbound all day. [p 5]

Park Lawn and Mimico Stations

This station would be located east of Park Lawn Road and north of the “Mr. Christie’s” site in southeastern Etobicoke. This is an area of intense high-rise residential development where, originally, little thought was given to the transit requirements of the new population. Now there is pressure for both a GO Transit station and for improved TTC service by way of the proposed Waterfront West LRT line to Union Station.

A station at Park Lawn was rejected in earlier studies because it would be very close to the existing Mimico station, and because available space would not accommodate the 12-car trains operated on the Lake Shore corridor. The close station spacing would affect travel times for riders originating to the west.

Metrolinx has resolved the station size problem by considering a shorter, 8-car platform where only part of trains stopping here would open their doors. The travel delay problem has been “fixed” by subdividing service now planned to stop at Mimico into a split Mimico/Park Lawn operation with trains every 30 minutes at each station.

The original capital cost estimated for Park Lawn station was $189.4 million, but that was for a larger station. The 60-year benefits of $156 million are now shown as exceeding the projected cost (not published).

Projected demand is 2,600 riders in the AM peak, and 10,000 all day primarily from the residential community around the station. The peak demand number raises the question of available train capacity given how few trains would stop her during the peak, and their accumulated load on the trip inbound to Toronto. If there are five trains over two hours, this would mean an average of about 500 riders would attempt to board each train here, about one quarter of the trains’ peak capacity.

One criticism of the original review for this site was that it did not take into account the population growth at Humber Bay. It is unclear whether an updated population base was included in the demand model, but the high density around the station is now acknowledged. Also of interest is the future of the Christie’s lands south of the rail corridor and the Ontario Food Terminal to the north. The City of Toronto wants these kept as commercial/industrial lands for job growth, but it is not clear whether Christie’s will instead turn into a forest of even more condos. As for the Food Terminal, the obvious question is where it would go if it relocated, but the property is huge and there would be even more pressure for condos because that’s what developers want to build.

The area could become a “West Harbour” by analogy to the development planned at the Don River, but this would require a major change in thinking about transportation and the transit/auto balance in southern Etobicoke.

Barrie Corridor

The service plan modelled for this corridor is:

Outer service stopping at all stations between Allandale Waterfront and Aurora; trains will also stop at Downsview Park and Spadina stations, otherwise, express to Union Station. Inner services will serve all stations between Union Station and Aurora. [p 6]

According to the RER website, “local” service is planned to operate every 15 minutes.

Mulock Station

The station would be located south of Newmarket in a generally low density area including the flood plain of the East Holland river.

There are no major developments or intensification plans approved that would suggest an increase to the existing density. This density does not meet Metrolinx’s Mobility Hub Guidelines suggested minimum density for regional or express rail, and is below the target identified in the Growth Plan for a priority transit corridor. [p 36]

Projected demand is 1,500 during the AM peak, and 4,200 all day. The projected benefits are negative due to the delay for riders originating further north, and the station’s principal function appears to be as a park-and-ride service.

The total includes new and existing GO riders, who would now use this station rather than another such as Aurora and Newmarket stations. A new rider at the station would have previously used their automobile or taken transit to reach their destination. Approximately 75% of riders are expected to access the station via drive-and-park, and via some walk-in and transit access users from surrounding areas located between Leslie Street and Yonge Street. [p 36]

CEO Phil Verster noted at the media briefing that Metrolinx is trying to find ways to improve the performance of this station, but it is fairly clear that it exists more for political reasons than as a professional choice.

Kirby Station

This is the infamous station that was originally rejected by Metrolinx staff, but then resurrected thanks to pressure from then-Minister Del Duca.

The surrounding area presently is largely made of farmland and natural open space, with the exception of two residential subdivisions and a few commercial and industrial establishments. [p 28]

Its primary function is expected to be as a park-and-ride location, and the design has been modified since 2016 to expand the amount of parking space available.

Ridership forecasts suggest that Kirby station would attract approximately 10,600 daily riders by 2031. The total includes new and existing riders who formerly used Maple or King City stations. Ridership is driven by a combination of projected growth in the vicinity of the station and in the rest of the City of Vaughan. The station’s proximity to Highway 400 (via Teston Road exit) serves to extend the station’s catchment area to the north. The majority of trips forecasted to use Kirby station in the AM peak period are boardings, i.e. travelling from their residence to their destination (primarily to Toronto). Of the total riders at the new station, the model indicates that approximately two thirds would be drive-and-park users, while the remainder would walk or use transit to access the station. [p 29]

This station has moved from a no-hoper to one where the projected benefits exceed the costs and sets up the obvious conundrum that either the estimates are entirely fictional thanks to political pressure, or the original station review was miles off of the mark throwing into question the quality of all reviews. The projected AM peak usage is 3,800 which is quite respectable considering this station was not to be included in the original plans. However, it is a “local” station on the Barrie corridor, and it is unclear how this number of users will be accommodate on the service that actually stops here.

Innisfil Station

Innisfil station would be located south of Barrie, and it would serve primarily park-and-ride customers. The projected demand is 1,000 in the AM peak, and 2,800 all day.

The report claims that the benefits at $214 million will exceed the cost of the station. These benefits arise mainly from reduced auto user mileage as the station would be closer to their homes than other nearby stations they use currently.

Service would be provided every 30 minutes by all trains as this is north of the turnback location for local service in the corridor.

Highway 7-Concord Station

This station would be at Highway 7 and would compete with the Vaughan Metropolitan Centre subway station for riders. Projected demand is 1,900 in the AM peak, and 5,500 all day.

Those accessing the GO station by transit would have lower travel times, but the overall travel time effect is projected to be minus $102 million. This station was not included in the 2016 plan and its status has not changed.

The model indicates that users walking or using transit to board or alight at the station experience the highest travel time savings, ranging between 10 to 20 minutes or more. Travel time savings benefits are highest for alighting riders that are destined to areas along the Highway 7 corridor in Vaughan.

Drive-and-park users would save relatively little time because the auto travel time between Highway 7-Concord and alternate upstream and downstream stations is small. The new stations along the TYSSE also provide an attractive alternate option for riders that have destinations in downtown Toronto or other areas that are accessible via the TTC subway network.

The station is directly downstream of King City, Maple, and Rutherford stations, which currently attract a significant number of boarding riders and are expected to continue to do so in the future. Although Express service plays a significant role in reducing delay impacts to these upstream riders that originate from Aurora station and points north, the station would continue to delay the riders that board at these northern stations. Refinements to the express service concept for the Barrie corridor could limit upstream impacts at Highway 7–Concord. [Appendix II, pp 7-8]

These comments suggest that the benefits of express service in limiting upstream effects are not uniform and/or that the service plan is still too fluid to nail down the effects. This also provides wiggle room should the political imperative to build this station overwhelm its currently projected negative impact on corridor performance.

St. Clair West Station

A station at St. Clair and Caledonia, the former site of the CN St. Clair station, was examined and rejected in 2016. Projected demand is 1,600 in the AM peak, and 6,200 all day. Where these people would actually come from is difficult to imagine considering that there will be stations on the Barrie line at Eglinton (Caledonia) and Bloor (Lansdowne) connecting with rapid transit services, and the land around the St. Clair West station (not to be confused with the subway station of the same name) is low rise residential, parkland and a cemetery. The station would also be near the proposed SmartTrack station at Keele and St. Clair.

In the 2016 evaluation, the capital cost was $27.4 million, and the estimate is now higher. The 2018 evaluation includes the usual penalties due to delays of upstream riders, and there is no discussion of express service although this would, according to the corridor’s service plan, be a local station.

The status of this station has not changed.

Bloor-Lansdowne Station

The idea for a Bloor-Lansdowne station originated in the community debates over the effect of the Davenport grade separation roughly 1 km north of Bloor Street. If the neighbourhood had to accept the rail bridge, it wanted a station at Bloor in return. One can certainly argue that the ability to transfer between the Barrie corridor and the Bloor Subway is very attractive for riders whose destinations lie along that east-west axis rather than in the core.

A few problems make this link difficult:

  • Because the grade up to the bridge must start well south of the actual crossing over the CP North Toronto sub, a station at Bloor must lie south of the street rather than spanning across to serve both sides.
  • The subway lies north of Bloor, and the west end of Lansdowne station is almost at Lansdowne Avenue, well east of the tracks.
  • An enclosed connection between GO and the subway here is unlikely until (and if) the property northwest of Bloor and Lansdowne is redeveloped.

The distance from the north end of the GO station platform to the subway’s street entrance would be about 300m.

Projected demand at this station is 2,200 during the AM peak, and 8,500 all day. AM peak usage would be primarily by riders transferring off of the GO Barrie service to the subway. This type of trip would be encouraged if and when a sheltered connection between the GO and subway stations is built. However, this is a local station in the service plan and would not provide a frequent connection to the subway service.

The imputed 60-year benefit value of this station is trivial at $11 million, while its projected cost in 2016 was $32.3 million, and is now higher.

The model indicates that transferring users at Bloor-Lansdowne connect to a range of destinations, particularly west to Etobicoke and Mississauga via the TTC subway Line 2, and to the north along the Barrie corridor in York Region. The Barrie corridor’s Downsview Park station can also be used as a direct connection to the subway system. In terms of travel time savings, the primary consideration is the amount of time that the new station saves relative to a previous transfer at Downsview Park, which would require an additional transfer to the Bloor-Danforth line at Spadina or St. George station. Local area users destined for downtown Toronto would save less time. [Appendix I, p 4]

This is a clear situation where an obvious link in the larger network makes sense, especially if more trains stopped here.

Spadina-Front Station

This station would be located in part on land now occupied by the Bathurst North yard where GO trains lay over during midday. With the increase in off-peak service, less storage will be needed for peak-only trains.

The origin of this station lies in a scheme to offload demand at Union Station by terminating some trains from the Weston/Kitchener corridor as well as the Barrie service at a satellite station west of Union. Distribution of riders to the core area would be handled by the Relief Line which was expected to have a station near Spadina and Front. However, with the shift of the RL north to Queen Street, this arrangement is highly unlikely, and a Spadina-Front station is now seen as serving downtown by foot traffic, not with a rapid transit link.

Moreover, the station as now proposed would be a through station only for the Barrie service which would continue east to Union Station. The service plan includes Spadina-Front as an express station and so all trains would stop here.

The design has evolved because it now only needs to serve Barrie trains, and one of the two station tracks can be an existing Barrie corridor track. This reduces the amount of the Bathurst North yard that would be removed to make way for the new station.

Projected demand is 10,200 in the AM peak, and 39,300 all day. This is primarily a destination station for commuters who would otherwise alight at Union and double back westward. The travel time benefits are very high with an overall 60-year benefit of $1.3 billion compared to the $25.1 million value, since increased, in the 2016 report.

In one of the bizarre side-effects of the Metrolinx methodology, this station is expected to discourage some upstream users by adding to travel time, and this shows up as a negative value in the benefit analysis due to increased auto trips. That negative is overwhelmed by the imputed benefit of shorter travel times for riders who no longer have to go all the way in to Union station.

There are several design challenges, mainly for clearance and the effect on yard capacity, at Spadina-Front, as well as for workable designs for passenger access from the surrounding streets to the platforms which would not necessarily be at the north edge of the rail corridor.

19 thoughts on “Metrolinx New Stations Report: The Details

  1. That Barrie-Bloor station remains a bad design. There must be some way to move it north, albeit at a higher cost. It’s crazy not to have all trains stop there too, the network benefits are too large to ignore.

    I would hate to see this station built, and then the connection left to rot for decades, as at Dufferin West.


  2. Would it not be better served to come up with a design plan to have the Spadina-Front station serve all lines running west of Union if not at least Smart track? It would off load some of the crowd going to Union during peak and also serve events at the Roger centre, if frequency could be 10 minutes or better.


  3. I stand to be corrected but the schedule I was provided states that the Toronto Stations (SmartTrack) will be put to a vote at April’s Council Meeting and the TPAP is scheduled for the summer. I don’t follow the politics too closely, but it looks to me like Tory is ramming through building SmartTrack without the steps necessary for large projects. Does anyone follow this stuff?

    Steve: The “Stage Gate 3” vote cannot occur until updated capital costs are presented to Council. Whether they have the appetite to proceed with some or all of “SmartTrack” remains to be seen, especially as Metrolinx has reneged on the service levels people were expecting.


  4. Hey Steve I know that you said in your analysis that the local stops will have 20 min service but the report states the following:

    All-stop peak direction outer service between Lincolnville and Unionville stations; trains will also stop at Kennedy and East Harbour stations, otherwise, express to Union Station. “Inner” services will stop at all stations between Unionville and Union Station. [p 6]

    Now the current GO RER plans for Stouffville corridor during the peak is 4 trains between Unionville and Union and 3 trains between Lincolnville and Union. Therefore, I got the impression that those three trains will run express and the other 4 trains would offer local service. Now I don’t know whether it’s even possible to have a 20 min express and 15 min local operating at the same time but it does seem to be that the local stops would still get 4 trains an hour. Also the report doesn’t say anything about how the service during other periods would change and the original GO RER plans had 4 trains/hr between Unionville and Union and 1 train/hr from Mount Joy to Union. So it would seem that the local stops would still get 4 trains/hr but it prob wouldn’t be a perfect 15 min service.

    Steve: Thanks for catching that. Yes, the local service will be every 15, although it will be reliably 15 we don’t know. However, at a SmartTrack station presentation I just attended, the claim was that service will be every 15 minutes off peak, and better, possibly less than every 10 minutes during the peak. This is based on the original scheme where all trains made all stops.

    I will correct/clarify things in the article.


  5. If you would indulge me, I would like to talk about hockey. Don’t worry, I am going somewhere with this.

    For those unfamiliar with the topic of sports statistics, hockey has been thrust through a revolution in the past few years. A very quick and far reaching revolution. It’s akin to Moneyball in baseball. A problem though has been that speed of change. Hockey has caught up in three years what baseball statisticians did in thirty.

    It’s caused a problem I think may be shared with these Metrolinx business cases. People take garbage data, run it through a sound statistical evaluation, and wind up with stats that have a gloss of intelligence, but on second reading make no common sense.

    A famous example is one of the ‘black box’ models developed by a group of University sports management degree students to identify ‘hidden gems’ in the NHL; players who are undervalued for various reasons, and would make good additions to teams at low cost. It produced promising results; players on the list many had long stated were great players, only stuck in menial and limited roles who could be better deployed on different teams, with different coaches, with great success. It had all kinds of obscure data analyses to support this. It all sounded very good.

    Except there was one problem.

    This model also showed that Sidney Crosby wasn’t good at hockey. It showed he wasn’t close to being in the top 30 players in the NHL.

    On it’s face this is nonsense, yet the authors of the analysis defended it to the end because they had used sound statistical principals to manipulate the input data. They couldn’t accept that the inputs themselves had to be flawed because they deflected all criticism back to the analysis, not the input data.

    This brings me to Kirby GO station. I have read through the IBC to which you linked and I am seeing the same thing. There’s lots of extremely fine grained stats like “travel time delay impacting upstream riders will total over 91,000 minutes per day.” Sounds impressive. But, what? On it’s face this seems implausible. “188 net riders” per day will abandon GO for the delay stopping here 15 years in the future?

    This is like saying Crosby is not a good hockey player because a selection of data manipulated with sound statistical analysis says so. It’s “evidence based” analysis the same way a ham sandwich from Subway is “artisan”.

    GO riders are not changing their commutes because of a few minutes here or there. I’ve been one of them. I have taken more than ten thousands GO train trips. Where do they get the idea people decide to use GO because of this? GO itself has altered the duration of trips shorter and longer by five or ten minutes dozens of times, even without new stations added. They have even added stations with zero passenger loss before. Clarkson was unilaterally added on express trips from Hamilton with zero comment from anyone about what a horror it was to stop there an extra three minutes.

    Maybe I am wrong. I never use this line. Maybe one more station could cause a net loss of “188” riders in the year 2031. I’d love to have a deep dive into the data collected to perform these analyses, but, I guess Metrolinx doesn’t allow us to see the raw details, sample sizes, and variance; or the rationale behind the raw subjective numbers. Until I get that, this smells like garbage in, garbage out, and so I am not going to complain about a minister overriding data that appears to be so shoddy.

    Steve: I fully agree that Metrolinx’ analysis has very serious problems in its purported linkage between small changes in travel time and ridership. I suspect that this is not a linear function, and at small deltas, a change in travel time is imperceptible to the rider versus the convenience of being on the train. A big change is quite another matter, but that’s not what is on the table, especially when Metrolinx previously claimed that electrification would offset the penalty of having more stations. A further problem is that those small delays are translated to a “cost” for riders even though they stay on the trains, and faster trips become a “benefit”. No real money changes hands.

    Garbage in, garbage out does not begin to describe the situation. Consultants get rich on this sort of thing.


  6. Steve wrote about Bloor-Lansdowne: “The distance from the north end of the GO station platform to the subway’s street entrance would be about 300m.”

    Interestingly enough, this is about the same distance separating the Leslie subway station from GO Oriole.

    300 metres is also approximately the proposed distance for the “High-Order Pedestrian Connection” that Metrolinx is proposing to build to connect the Hamilton GO station with the new LRT.

    So there seems to be an expectation that people using transit should expect walking connections up to 300 metres. Yet if the same expectation of walking is put upon car drivers in the context of the King Street pilot removing on-street car parking, suddenly the sky is falling and every business is going bankrupt. There is nothing quite like the fine smell of bovine effluent.


  7. Steve, can you see any logic to justify the proposed Park Lawn station vs. a stop at Humber Loop, which seems to be the glaringly obvious alternative? Park Lawn links to nothing but developers’ dreams and is too close to Mimico. Humber Loop, by contrast, is 2,100m from Mimico, well within Metrolinx norms. Direct links to Queen/Lakeshore streetcars; 80 & 66 buses and potential for express shuttle to Jane Stn. Track there is arrow-straight, with ample room on the north side to “bulge” the tracks in order to add platforms, plus even a narrow strip (about 10m) on the Gardiner side. The TTC-owned Humber Loop site (a triangle with about 2ha / 5 acres) has ample room for station, multi-story parking, etc. — and there’s an even larger low-density property next door. It appears it would even be easy to add QEW on/off ramps allowing access to Humber Loop parking. Plus, a 150m pedestrian bridge over the QEW would put Humber Park and all the existing high-rises within 500-700m of the station. And Humber Loop is closer (w/in 500-700m) of all the existing high-rises. So what gives? Is this corruption at work? Greetings from your fan in Japan, John Harris.

    Steve: I can see several reasons for not using Humber Loop. First off, the right-of-way is quite constrained and there is little if any room for platforms. Remember that the rail line is on an embankment at this point and it would have to be widened to fit in anything new. Second, the loop is not, as you claim closer to the high rises which now stretch from the river all the way west to Louisa Street which is two car stops west of Park Lawn. There is over 1 km worth of condo strip west of Humber and less than half a km to the east.

    There is most definitely not room for a parking structure at Humber Loop, let alone ramps from the QEW, and some of this would run smack into a hydro line and towers that runs through Humber Loop.


  8. They should run their net benefit model and Mimico station. If even little or no capital costs, I suspect it will have a negative benefit and should be closed in favour of Park Lawn. It is poorly connected to transit and is in a low density area. Just because it was a good location in the 1800s doesn’t mean it’s a good location now.

    Steve: A more general problem with their modelling is that they never apply it to the existing network to get comparative numbers, and a great deal of the “benefit” is derived from the notional value of saved time over a very long period.

    Also, much depends on the conditions assumed around the site. A lot of condo building is in progress or proposed near Mimico and this will change its trip generation numbers. Closing the station would mean that current and potential users would have to travel to another station. I am not sure the numbers would come out as you expect.


  9. Kevin Love said: So there seems to be an expectation that people using transit should expect walking connections up to 300 metres. Yet if the same expectation of walking is put upon car drivers in the context of the King Street pilot removing on-street car parking, suddenly the sky is falling and every business is going bankrupt. There is nothing quite like the fine smell of bovine effluent.

    And yet, the 50-100 metre walk from the subway at Kennedy up to the RT is derided as being unfair and unacceptable to transit riders in Scarborough. Indeed, it really is a bunch of bovine effluent.

    Steve: There is a whole field full of the stuff with hardly a safe patch to stand on.


  10. There is a local south Etobicoke group pushing for Humber loop to become a ‘Transit Hub’ with a GO station there.

    I’ve looked at overhead views, and also wandered around the loop, and no, there isn’t room. Not without a wholesale reconstruction of the whole area, including probably the Gardiner/Lake Shore bridges.

    Humber Loop is constrained by the Humber valley to the east, the rail lines and Gardiner and Lake Shore bridges to the south, the elevated eastbound lanes of The Queensway to the north, and gradients to the west, where the big box centre would have to be expropriated. And north of The Queensway is a sewage treatment plant that isn’t going anywhere. It’s no place for a transit hub.

    Of course this same group is pushing for an LRT along The Queensway, based on the excellent planning criteria of “the road is wide enough” (well, except where it isn’t, I guess), and naturally no LRT on Lake Shore, lest it require the demolition of half of Mimico and New Toronto.


  11. While shooting down the Scarborough subway, Steve had suggested that SmartTrack would provide better transit for Scarborough than the subway but now, Steve has dismissed every single Scarborough SmartTrack station. It is clear that according to Steve, nothing should be built in Scarborough so that all that Scarborough money can be spent in Downtown. I have complete confidence in Mayor Tory’s intentions to build both the Scarborough subway way and SmartTrack and both are needed. If you want to cancel the Scarborough subway, then please run for the mayor’s office. I will be supporting John Tory.

    Steve: I’m letting this comment through because it’s not unduly incendiary or insulting, unlike a lot of stuff you leave here.

    First off, I did NOT ever claim that SmartTrack would provide better transit for Scarborough. In fact I consider StartTrack to be a scam that results in Toronto paying both in capital and operating costs for ST, while we don’t invest in an LRT network in Scarborough which is what is really needed.

    What is also needed is the Relief Line whose purpose is to take riders who now originate in Scarborough and eastern North York downtown without going through Bloor-Yonge station.

    But it suits you to think that I’m all about downtown, and that’s your right. You’re dead wrong, but that has never stopped a Ford or Tory supporter from spouting their usual fables about my position.


  12. “Metrolinx is currently preparing an Initial Business Case (IBC) for level boarding across the GO rail network. The dwell-time savings of level boarding are significant. The run-in and dwell time at stations built for level boarding is expected to be reduced by at least 30 seconds. Riders can board and alight more easily and efficiently. Dwell times will be reduced as it will not be necessary to deploy the accessibility ramp, although use of the accessibility ramps may need to continue during a transition period at stations not equipped for level boarding, until they are modified. All new stations are anticipated to be built for level boarding from the outset. This reduces the impact on upstream riders, taking the incremental trip time impact from 2 minutes down to about 1.5 minutes. On the TTC subway for example, which has always had level boarding, station dwell times can be 15 seconds or less.”

    15 seconds here, 15 seconds there, another 15 seconds over there, and so on. If there is a reduction in dwell-time at all the GO train station, especially the Lakeshore West train stations, then adding the Park Lawn GO Station could see no increase in the time trains need to run from Hamilton to Union Station. With dwell time of 1.5 minutes, instead of 2 minutes, only 3 stations need to have level boarding. With additional stations with level boarding, there should be an overall improvement in overall times.

    Steve: Not to mention the benefits of electrification. Metrolinx seems to mention the “penalty” of extra stops when it suits them to paint a proposal in a bad light.


  13. If the developer of the old Christie’s property wants a GO station, let them pay for it 100%. Condo buyers will benefit while in turn paying for it Alternatively, set up a shuttle bus to/from Royal York GO.

    There is a multi million dollar rebuild of Royal York station to be amortized.

    Steve: The TTC has looked at a shuttle bus to Mimico station, and the travel time is not trivial.


  14. Ed… Your logic doesn’t hold up. Start with “west, where the big box centre would have to be expropriated.” What owner of a 9-acre, low-density plaza not jump at the chance to integrate his property with a project that’d give his new high-rise condo/commercial complex direct, climate-controlled access to a station under 10 mins from Union? Add to that air rights over the 4.5 acre TTC property and you have a huge site. Any rail operator here in Tokyo would laugh at your contention that there’s not enough room: 750+ meters straight track from Humber Bridge to expwy underpass with ample room to the north side, and a strip to the south just wide enough for pillars to support a structure over the tracks. Yes, significant construction would be involved. But this is exactly the place for a massive parking structure, just before the expwy narrows and tolling should begin. Add in a Pickering-style pedestrian bridge and the existing Lakeshore condos are almost all w/in 500m of the station.

    As y’day’s Star noted: “Roughly 11,400 people lived in the Humber Bay Shores census tract in 2016, about 6,000 more than lived there in 2011. Despite being only about 10 km away from Toronto’s downtown, it takes at least one hour to make the trip by public transit.”

    The proposed Park Lawn stn, meanwhile, is 400m from Lakeshore, 600m from Queensway and over 1000m from most of the existing condos. Very key question: Who benefits from the political pressure in favor of Park Lawn? And how is that interest linked to the politician who applied it? Maybe Ben at The Star should be looking into this?


  15. “What owner of a 9-acre, low-density plaza not jump at the chance to integrate his property with a project that’d give his new high-rise condo/commercial complex direct, climate-controlled access to a station under 10 mins from Union?”

    Who is the owner? I think people don’t realise a lot – actually almost all – of these kind of plazas are owned by real estate investment funds which operate with a goal to have steady long-term income streams, not to develop or redevelop their sites for capital returns. Most don’t even have the free cash from investors to redevelop sites, they are supposed to return income and if that plaza is returning sufficient income to meet the fund’s objectives they are content to do nothing.

    And why would they do anything anyway? They know from experience that if they wait long enough, the government will step in and pay for it all, and then they can sell the land and reap the capital return benefit.

    Steve: There is a much more fundamental issue here. A stop this close to downtown is no place to be attempting to create a major parking node on top of the population already in this location. There is the small problem of capacity on trains that will originate further west and have no room for riders. Indeed, at the service level GO proposes, I wonder whether the projected demand just from the residents will fit on the trains.

    If we are trying to intercept motorists and get them out of there cars, this should occur much further west.


  16. John, you say you’re a “fan in Japan”.

    I live in south Etobicoke, I have been to Humber loop too many times to count, and have had a good chance to look around — especially those times when the streetcars have inexplicably not come on time. I have looked at detailed aerial maps with GIS measuring tools. I have a good feel for the topographic difficulties around Humber loop. (For example, it’s impossible to enter the bus loop when travelling westbound, and it’s impossible to depart eastbound. The TTC actually owns a little loop at the corner of High St. and The Queensway to accomplish turnarounds).

    So I’m going to go with my observations over your opinions.


  17. Mike wrote: “..all that Scarborough money can be spent in Downtown.”

    Mike appears to be unaware of who is paying municipal property taxes in Toronto. For example, in the year 2015, Ward 28 Toronto Centre – Rosedale paid $366 million in property tax, while Ward 43 Scarborough East paid $31 million.

    Scarborough receives a very large net inflow of funds from downtown, not the other way around. See here for a ward-by-ward breakdown of property tax payments.


  18. Steve: The TTC has looked at a shuttle bus to Mimico station, and the travel time is not trivial.

    In this regard I was suggesting the developer or some some other third party setup shuttle bus with mini-vans/bus like an Uber Pool setup. Not TTC large, slow buses with a high paid driver on full days pay. Needs to be flexible and quick to change with low cost.

    Steve: I am always amused when a service with considerable potential usage is simply palmed off with a minivan solution.


  19. They should just remove Humber Loop completetly. It is near nothing, just an empty streetcar change. No shopping, no living. The world just bypasses the TTC Loop. Streetcars should just travel right thru, not even a flag stop. But why not reopen Sunnyside railroad station, it’s near a hospital, access to major streetcar lines, near the lake and people live there.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s