Many background presentations informed the Expert Panel’s review of options for the Sheppard corridor. This article is the first of two summarizing and commenting on this information.
There are six groups of documents:
- Professor Eric Miller’s comments
- Metrolinx presentations and reports
- TTC presentations and reports
- Toronto Transit Infrsatructure Ltd. (TTIL) presentations and reports
- City of Toronto presentations and reports
- Third Party reports
TTIL is the TTC subsidiary through which Dr. Chong’s pro-subway work reported. Given the amount of material, I will deal with reports from TTIL, the City and Third Parties in the fourth and final article in this series.
Professor Miller begins by observing that the technology choice for Sheppard does not fix forever and for all possible corridors the choice between subway and LRT. In the context of this study, that is true because there is already a planned LRT network to which the Sheppard route would be added just as there is already an existing subway which could be extended. The panel’s decision would have been much more complex without the earlier vote by Council confirming LRT for Eglinton, Finch and the SRT conversion.
Demand projections by both the TTC and by Metrolinx do not support the investment required to sustain a subway extension. These projections rest on current land use plus reasonable projections of future growth in the GTA. It is worth noting that growth is non-uniform and some areas actually lose population and/or jobs rather than growing. This is demonstrated in maps later in the presentation.
We hear a lot about travel speed from subway advocates, but this misses a basic fact. There are many components to a trip including:
- Access time to and from a line
- Access time to and from a station
- Wait time
- In vehicle time
For a subway, access time to the line may be greater than for LRT because the stops are further apart. This depends on implementation details for specific proposals. It could also be affected by LRT and subway proposals taking different routes.
For a subway (or an underground LRT), access time at the station is a function of station design. Some of Toronto’s subway stations have a fairly short path from the surface entrance (or bus/streetcar loop) down to the platform while others are notoriously roundabout. This component can particularly affect those with mobility challenges because escalators/elevators may not be placed in ideal locations or may be out of service.
Wait time is a function of the level of service (headway). In Toronto, people love subways because there is a policy headway of about 5 minutes on most routes whether the demand justifies it or not. This is not the case in many other cities where headways well over 10 minutes during off-peak periods may be encountered. If the TTC had a maximum policy headway for key surface routes, it is unlikely that this would be as short as 5 minutes (10 was proposed in the “Transit City Bus Plan”). Riders in the York University area will benefit from the subway-based headway standard once the Spadina extension opens, but they would never have enjoyed this level of service on any form of surface transit unless the demand justified it.
Passengers typically “feel” access time as more onerous than time spend riding because it is less productive. A rider is not really “on the way” until they board a vehicle.
In vehicle time is a function of acceleration and speed. During the AM peak, the SRT operates on a higher performance setting than at other times so that trains can make slightly faster round trips and squeeze more capacity out of the available fleet. This contributes to the SRT’s ranking as the fastest of the rapid transit lines. The TTC could operate the subway at a higher performance rate, but chooses not to for a variety of technical, maintenance and institutional reasons. An LRT line will not operate at the same speed as a subway unless it has comparable characteristics — more widely spaced stations and unimpeded running between them — and the most likely sources of delay will be intersection design (including signals) and stop dwell time.
Toronto streetcars mostly use pay-as-you-enter fare collection and this forces all loading to take place through one set of doors (unless the operator uses the rear doors with or without the assistance of a TTC loader) . Moreover, the streetcars are high-floor cars and the need for passengers to climb into the cars adds to boarding time. Subways run with level platforms and all-door loading, an arrangement that will apply eventually to the streetcar system and to all of the new LRT lines as these will use low-floor cars with platforms matched to floor height.
Miller concludes this section with a basic observation:
… people use transit when it is accessible (within easy walking distance), frequent and reliable, and takes them where and when they need to go in reasonable time.
The section on connectivity draws on information from the travel database maintained by UofT on behalf of many planning and operating agencies in the GTA. The overwhelming majority of travel in Scarborough is not oriented westward across Victoria Park, but is local either within Scarborough or headed to points north and east. Of those who do travel to the west, the majority are headed downtown, not to Yonge-Eglinton or Yonge-Sheppard. This is easy to understand given the relative numbers of jobs downtown compared to other locations. Places like Yonge-Eglinton will always be nodes surrounded by residential neighbourhoods unlike downtown with block after block of office towers and a local toleration for high density development. The LRT proposal, by providing a spine across Scarborough, is much better suited to the actual demand patterns and the need to link with north-south routes than the subway proposal.
To illustrate the concept of an integrated network, Miller includes the diagrammatic map of Munich’s transit rail system. This map does not include the tram or bus services. Although services in Munich may be provided by multiple agencies, they are operated as one network with an integrated fare system.
In a series of population density maps, Miller shows how different Scarborough (and for that matter all of the suburbs) is from downtown. This is not just a question of density around a few stations, but along entire corridors. Population gains and losses are interesting because they show that this is not a uniform pattern across the city. There are small pockets of population loss between 2006-2011 (the demolition of parts of Regent Park shows up particularly strongly), and a major area of growth lies in northeastern Scarborough, the last undeveloped suburban land in the City. This area is completely ignored by the subway proposal.
Metrolinx Presentations & Reports
This presentation begins with a statement of the”five principles” imposed by Queen’s Park on any plan that Toronto might come up with.
- Projects must support regional planning goals. By implication, vanity projects do not qualify, but I suspect this depends on who is primping in the mirror.
- The total cost must remain within the Provincial commitment of $8.4b (2010), be spread out to match (unspecified) cash flow requirements to 2020, and produce assets that the Province can own and depreciate (this is an accounting dodge to make the related Provincial debt vanish off of the books).
- Any penalties from changes in previous plans are at the City’s cost. As we are now back to the point we left off at in late 2010, this point is now moot.
- Any cost due to further delay will be charged to the City.
- “Any plan should minimize adverse impacts on traffic to the extent reasonably
possible”. That could mean just about anything.
Metrolinx goes on to review the Sheppard line’s design in the regional plan, The Big Move, and notes that it was intended to operate largely on the surface.
Next we come to the Benefits Case Analysis of options for the Sheppard corridor of which a version with a continuous Finch-Don Mills-Sheppard line ranked highest. This happens because there is a high demand on existing bus service on Finch west of Don Mills that would be subsumed into a Finch East LRT, but the proposal ignores the difficulty of LRT construction on the narrow, low-density residential portion of Finch east from Yonge Street. This scheme never got off the ground, but remains in the plans as a future extension.
The original 5-in-10 plan is summarized both as to cost and schedule.
- York Viva BRT: $1.4b 2009-2019
- Sheppard East LRT: $1.0b 2009-2014
- Finch LRT: $0.94b 2015-2019
- Scarborough RT: $1.8b 2015-2020
- Eglinton-Crosstown LRT: $5.0b 2010-2020
The Toronto projects here total $8.74b. Net of the Federal commitment of $333m to Sheppard East, this gives a Provincial total of $8.4b.
Work is already underway on the underpass at Agincourt Station, and some site preparation as been completed at the proposed Conlins Road yard.
The study tour report goes into some detail about the differences between the Madrid and Toronto environments for planning, design and construction of a rapid transit system. Ongoing, large-scale commitments allow for efficiencies of scale in construction, even without taking account of factors such as favourable soil conditions. Those who argue that Toronto can’t build anything without wasting a fortune would do well to consider the scale of transit commitment needed to achieve a comparable level of cost effectiveness.
One cannot help noticing that Madrid had a powerful, popular Mayor who pushed through a large-scale transit program with massive funding support from senior governments. When the same thing happened in Toronto, this was decried as an example of high-handedness. One clear difference is that the transit program in all its detail was an integral part of the Mayoral campaign platform in Madrid, and there was no doubt of the “mandate” to proceed with its construction.
(Additional information comparing Madrid and Toronto experiences appears in one of the TTC reports below.)
This presentation was prepared for the earlier Council meeting where the Rob Ford “MOU” with Queen’s Park was dismissed by Council and replaced with an endorsement of the original 5-in-10 plan except for the Sheppard corridor. This meeting also created the Expert Panel.
The presentation contrasts the original agreement Council had approved with Metrolinx to the MOU version of March 2011 and concludes that the 5-in-10 version is the better plan.
Projected passenger volumes through Kennedy Station in 2031 are shown with and without a through-routed service from the SRT to Eglinton. Where in past (and largely unpublished) reports, Metrolinx had trumpeted the higher volume possible on Eglinton as a reason to put that line entirely underground, the TTC points out that the riding is simply shifted from the BD subway which is then underutilized.
A series of maps shows the gradually disappearing Transit City LRT network as parts are hacked off thanks to funding constraints and political deal-making, and the presentation concludes with a table we have seen elsewhere showing the additional network coverage possible if the $2b that would put Eglinton underground goes instead to the Sheppard and Finch LRT lines.
begins with a set of maps building up from the TTC rapid transit network (the subject of so much current debate) to the full regional network, and puts the Sheppard question in a broader context — what exactly does it do for the region?
An overview of transit rider demographics follows together with maps showing targets for service improvements and accessibility under the Miller regime.
Next we have a short lecture in planning followed by maps of various older plans. Much of what is on these maps is not new, and there are only limited places for new rapid transit lines. “Intermediate Capacity Transit” in the 1975 refers to the Skytrain technology known here on the SRT. Very little of that proposed network was actually built.
By the time we reach Network 2011 (1986), much has falled off of the map leaving only the Sheppard subway, the DRL and Eglinton West. Note that the Downsview and STC extensions of a Sheppard line are priority 4 behind the DRL and part of Eglinton.
One important point to remember is that the idea of a Sheppard connection to Downsview goes back to a scheme to loop the Yonge and Spadina subways and, thereby, avoid terminal congestion problems that limit headways on this route. However, what was once low-density North York quickly filled with new development, and by 1990 the “loop” had been pushed out to Steeles Avenue at which point the “loop” ceased to have credibility.
An historic footnote: Many years ago I asked the TTC why the Spadina extension’s EA did not consider LRT as an alternative as a jumping off point for a northern LRT network. The TTC used the “loop” scheme as a demonstration of why LRT could not work — it would interrupt the loop — even though this scheme was impractical from the day it was proposed. The TTC’s hands are not clean in the suppression of debate about LRT much as they would have us believe “modern” LRT is a recent, widely accepted transit mode.
Next we see a rather sad set of maps showing the minimal growth of rapid transit in Toronto and the large areas still remove from the subway network.
A graph on page 31 shows the transit modal share and density at all stations in the network. This is arranged from lowest to highest density. This chart continues the old TTC myth that density = demand when so many of their stations, especially in the suburbs, depend on bus feeders for their customers. Kennedy has a low transit share for trips originating near the station, but we would hardly call this stop a failure. Conversely Dupont has a higher modal share even though its density sits right next to Kennedy and its usage (a variable not included in the chart) is unquestionably much lower. This is a nice chart, but it only tells part of the story.
Next we turn to the policy shifts of the Miller era and the new Official Plan. Although there were designated centres at North York and Scarborough, these were not the great successes planners in the 1980s had hoped for. Sadly, it was a bit early to write about the emperor’s new clothes, and planning continued on the myth that somehow these centres would become major hubs within the city.
The chart on page 37 claims that Light Rail Technology was “unproven” when the earlier plans were done. I am not sure how I and many others managed to ride on LRT systems elsewhere in the world (including Canada’s first in Edmonton) so long ago if this technology was “unproven”. This is either an outright lie, or a very creative retelling of history just as the TTC takes credit for “saving” its streetcar system, a feat achieved by citizen activism and Council direction, not by enlightened policy at the TTC.
By 2003, the TTC was looking at surface transit improvements, mainly bus-based, although including the St. Clair line from Yonge to Lansdowne. Almost none of this proposal was implemented.
Then came Transit City’s LRT and Bus plans. The latter never got off of the ground due to budget wrangling between the TTC and City (this in the Miller/Giambrone era), but the idea was to guarantee 10-minute service on a grid of bus routes. There are some odd exceptions notably all of the future LRT corridors even though they might not see LRT service for a decade. The streetcars were assumed to already be at 10-minutes or better although there are a few exceptions. The plan is a good idea, but poorly drafted and it needs revision if it surface again in a more enlightened Commission.
In a table of daily ridership figures, we see that several bus routes compete with the streetcar network for supremacy. This is not the way such stats should be presented because it assumes that demand characteristics are uniform across all routes. This is not true. One simple example — the length of a route will affect how many people it carries presuming similar spatial and temporary patterns of demand. One simply cannot compare numbers for the short 510 Spadina with the much longer 504 King or the combined 32/34 Eglinton West/East services spanning much of the city.
Page 45 shows a breakdown of riding on the Sheppard subway, the bus service on Sheppard East and the 190 Rocket to STC. Note the comparatively high ratio of all day demand on the subway to peak hour demand, although neither number is impressive, as against figures on the bus routes. A great deal of the bus demand does not show up in the peak hour, peak point counts because it does not fit the core-oriented commuter pattern underlying this type of analysis. Clearly, large numbers of riders use the Sheppard bus, but they don’t do so at times and places caught by core-oriented planning.
This is a fundamental issue with the Transit City network — it considers a variety of demands, not just the most obvious one flowing into a subway terminal. Whether Rob Ford likes it or not, “Transit City” is very much alive.
The next section traces the evolution of the Sheppard proposals in the context of the last decades evolving plans. Ridership for a Sheppard LRT is lower than for the proposed subway, but this does not address the underlying questions of which land use and demand pattern will actually evolve over the decades. Much of the new “subway” riding comes from proposed intensification around STC, and the subway itself does not serve many of the riders who might otherwise be on the LRT line.
The Sheppard Subway, even at its peak point, barely reaches 8k/hour just east of Yonge Street by 2031. This lies within LRT’s capability, but I am not sure that a retrofit would be cost-effective (I will turn to that question in a separate article).
Page 68 compares the 510 Spadina car with the proposed Sheppard LRT. It’s worth noting that the TTC claims the Spadina car has signal priority when, in fact, this exists at only a few locations on the route, and the “priority” can do as much to slow transit as it helps with the service. It would actually make more sense in this context for the TTC so say there is “no” or “little” priority to reinforce known problems with running speeds.
This presentation begins with a series of photos of LRT in major cities. Although many of the shots are attractive, they do not appear to have been selected to illustrate comparable conditions to what would be built on Sheppard. However, the photos do illustrate that LRT is alive and well in many “world class” cities.
Next comes a discussion of subway construction in Madrid and specifically of the MetroSur. A chart plots results from a US Federal Transportation Administration study of construction. The points are arrange from lowest to highest cost/km. Many, notably the MetroSur, fall below the average while others lie above it including projects in Hong Kong and London (the Jubilee Line extension). TTC projects at an average of $290m/km lie slightly above the international average of $275m/km (2010). The most expensive of the TTC’s projects is the Yonge Subway extension from Eglinton to Finch that came in at over $350m/km in 2010$. The “Finch LRT” is included in this chart at about $300m/km, but I must assume that this would be for a subway alternative as the estimated cost of the LRT itself is only $1b.
The TTC lists several cost drivers affecting underground construction costs including:
- the frequency/km and length of stations
- the number of major interchange stations
- the rate of advance of tunneling depending on conditions
- the effect of a high water table especially as it may affect settlement of properties above the tunnel work
- the complexity of underground utilities
- international competition for tunneling expertise and construction crews
The TTC presents a chart of construction times for projects and claims to be faster than the international average. However, the values do not appear to be scaled to the length of the projects, and a few outliers from Washington DC skew the average above that of TTC. This chart is not a meaningful presentation of the information because it tells us nothing about the characteristics of each project nor the source of major delays (or unusually fast construction) that might affect each project.
This is not to say that the TTC is bad at what it does, and peer reviews through the American Public Transit Association (APTA) have validated TTC processes. One important point the TTC does not mention is that a project’s timing can be affected by non-technical factors some of which apply to the Transit City network:
- To minimize disruption over a wide area, construction schedules may be extended. This applies to the Eglinton line where Metrolinx does not want to have every station under construction more or less simultaneously with the resulting massive disruption across the corridor.
- To manage financing requirements, projects may be artificially stretched out so that spending does not peak in certain years.
- Projects may be ready to build, but not have committed funding.
Next, the TTC turns to the MetroSur in Madrid and contrasts it with the Sheppard subway, the most recent large, completed project in Toronto.
Aside from the advantage of being a larger project with economies of scale, the MetroSur was part of an ongoing construction program pursued by the Mayor of Madrid over many years. The administrative environment including aspects of design (no “EA” process or site permit requirements), ease of property acquisition (the city has strata ownership of land 10 metres below grade), and a simpler project change management environment all contributed to lower costs.
Construction of the MetroSur took place in a very different corridor from Sheppard Avenue:
- A great deal of the corridor was greenfield with little or no utilities or traffic disruption to worry about.
- 30% of the line was cut-and-cover as compared to almost 100% tunnel for Sheppard between stations.
- Soil conditions allowed the tunnel boring to progress in Madrid at over twice the rate on Sheppard.
- Greenfield construction meant that surface settlement above the tunnel work in Madrid could be accepted.
- Fire code for Madrid is less restrictive than in North America allowing the use of a single tunnel without escape shafts between stations.
- Stations on the MetroSur are 130m long versus 200m for Sheppard.
- Track construction in the tunnel in Madrid does not include the vibration isolation used on all newer Toronto projects beginning with the Spadina subway.
When these factors are taken into account, the cost of the Sheppard subway is still 17% higher than the MetroSur, but not as wildly different as portrayed in other reports, notably those by Dr. Chong and TTIL.
The TTC concludes with a review of changes in conditions since the “Network 2011” plan of 1986. This covers much of the same ground as in other reports, but wraps up with a major change in the TTC’s outlook — subways do not always generate development contrary to the standard wisdom of past decades. Any new line is proposed in a complex environment of the GTA development industry and competes with many other locations as an attractive market for new buildings.