The agenda for November 23rd’s meeting of the Metrolinx Board is now online and it includes several reports of interest. Here I will deal with GO transit performance and capacity issues. In a future article, I will turn to Presto (and the proposed TTC implementation which is also on the TTC’s agenda for November 23), the Air Rail Link, and planning issues at “Mobility Hubs”.
Ridership on the GO rail system for September 2011 is 9.6% above 2010 levels, and for the bus network, the increase is 4.7%. This strong growth in ridership will strain GO’s ability to add service fast enough either by running more trains or, where possible, making existing trains longer. A report later in the agenda addresses the problem of capacity in the rail corridor and station at Union, and shows how the era of simply adding a new train here and there to keep voters in the 905 happy is just about over.
GO’s target for seat availability on peak period trains is 80% (1 in five passengers, on average, would stand). However, they are only achieving 64% (1 in three passengers stands). Although they are looking for more time slots in which to operate trains, this is limited by track layout, conflicts between GO and other services and limitations at Union Station.
Union Station 2031 and Related Studies
Those of us who attended workshops on either the Georgetown corridor plans or the Electrification Study have heard of parallel work on network capacity, but this is the first any report has surfaced publicly. The information here is troubling, but also long overdue as part of the debate on GO’s future.
At long last, Metrolinx is discussing the problems of actually implementing the brave new world we were shown in The Big Move rather than just drawing lines on maps. While some have agonized over “congestion” and road capacity, we have ignored the vanishing ability to offer alternatives by transit. Not only do we have funding problems, we also have physical limits to what existing transit infrastructure can deliver.
Union Station has two limits on its capacity: the number of trains that can operate through the corridor and the number of passengers that can be handled by the network of platforms, stairways and concourses. Downtown population is expected to grow 80% from 2006 to 2031 (from 71k to 130k). While this may contribute some of the growth in commute trips to the core, the employment growth is expected to be 25% on a much larger base (315k to 400k). Peak period transit demand to the core will grow by over 50% (156k to 236k).
All of these factors will place the transit network downtown at or beyond capacity by 2031. Indeed, Union Station is expected to reach capacity after the current 10-year GO expansion plan completes. Although pedestrian capacity in the concourse is more than doubling with the “dig down” and the new west GO concourse, the capacity of the tracks will not increase proportionately.
Union Station Demands and Opportunities
Passenger traffic at Union is expected to double or triple relative to 2006 by 2031 (the actual figure depends on the capacity to run trains that we build). This is a staggering growth in demand, and it will not be solved with band-aid fixes to the transit system. A few more trains here and there will simply vanish in the tidal wave of new customers. Moreover, the historic pattern of treating GO and TTC as separate networks simply must end — they are one network, they must be planned and funded as one network.
There are only two major classes of change that can be made: either we must stuff more trains and passengers through existing facilities, or we must add new capacity via alternate routes.
Six groups of options were considered:
- Better TTC/GO integration notably at Eglinton/Weston, Bloor/Dundas and Danforth/Main.
- Satellite GO stations at Liberty Village, Bathurst and Cherry Streets.
- Service on the North Toronto Subdivision (connecting to the subway at Summerhill or Dupont Station).
- A Downtown Relief Line from the Danforth Subway to Dundas West via Queen or King through downtown.
- A GO tunnel via Queen linking services in the Weston corridor to the Richmond Hill and Lake Shore East corridors.
- A GO tunnel parallel to the Lake Shore route.
The first three of these were rejected as they do not provide sufficient relief at Union, although they may be worthwhile in their own right for other goals.
Two variations of option 4 were studied in more detail:
- A: A TTC DRL line from Dundas West to Danforth
- B: A hybrid configuration with a DRL from Exhibition Place to Danforth, and a new GO station in the current Bathurst North Yard for the Georgetown and Barrie services.
For option 5, services from the Milton and Georgetown corridors would use a tunnel under Queen, and would be through-routed with the Richmond Hill and/or Lake Shore East corridors with subway connections at Osgoode and Queen Stations.
For option 6, there are also two sub-options:
- A: A Lake Shore tunnel that dodges north to Queen to connect with Osgoode and Queen Stations. (This has obvious problems with finding a clear alignment to and from the rail corridor.)
- B: A Lake Shore tunnel that would serve a new Union Station near Yonge Street.
Of this list, options 4B and 6B were deemed the best for a variety of reasons, and these will be studied in more detail. A full DRL (option 4A) does not make the cut in GO’s analysis because it does not address the Union Station problem. Indeed, GO’s main interest in the DRL is that it would provide a distributor/collector service between a new Bathurst station and the downtown core.
Transit integration is essential to making this scheme work, and part of this will be true fare integration between the TTC and GO.
The growing downtown population will contribute a substantial counter-peak flow away from the core. Projections of station capacity (stairs, waiting areas, platforms) must include provision for this traffic.
In brief, if we don’t build new GO capacity both in the central corridor and in our ability to handle passengers, Toronto will not have the capacity to handle the projected growth in core-oriented commuting.
Moreover, any scheme that involves putting GO underground demands electrification, and the foot-dragging that we have seen at GO on this issue must end. Electrification is also a factor in line capacity and speed improvements even for trains running on the surface.
Union Station and Rail Corridor Capacity
This study considered the requirements to serve three future scenarios:
- Completion of current GO expansion plans including the Airport Link (2015)
- Serving the base case scenario of the electrification study
- Serving the full Big Move plan (2031)
The 2015 peak hour capacity provides only slightly more than we have today: 4 Airport Link trains plus 2 more VIA trains. Although trains can be added within the peak period, the peak hour is completely spoken for. This is a major constraint on the growth of peak period GO service.
The Electrification Study base case requires 17 more peak hour trains, and a further 38 (approximately) are implied by The Big Move at full implementation. This is well beyond current capacity.
One option considered was a dig-down to provide four additional tracks and associated platforms. However, this is a very challenging project both in siting the tunnel portals and in providing access down to the platforms. This leads us back to the study of alternative routes discussed above.
The Downtown Relief Line
The TTC’s study of a DRL is now underway and is to be completed in early 2012. The scale of the problem is daunting when one considers the future demand into the core area as well as the possible additions to the network that will pump more demand onto TTC facilities. Although other changes such as resignalling and station capacity improvements are mentioned, there is no sense of when these will be exhausted or which of them provide added capacity at reasonable cost. Until we see the TTC’s study (or at least see a public meeting where an interim version might be reported), it’s impossible to know whether this will do more than support the TTC’s anti-DRT bias.
The maps used by Metrolinx all show the DRL ending at Danforth, and this is extremely short-sighted because it does not provide an intercept for traffic on the new Eglinton line.
Funding
None of this will happen without substantial funding, and the timing of requirements does not sit well with current Provincial plans that push much of the “big dollar” projects into the latter part of this decade (and beyond). Moreover, this is all infrastructure “downtown” although it will exist to serve commuters to and from outlying areas. Politically, the GTAH needs to accept that central Toronto will continue to grow. That growth, plus pressure on transit from the increased cost of auto commuting, demands big spending downtown.
GO’s role as a truly regional carrier will require full fare integration with the TTC, and that has a cost. We cannot demand that people pay double fares within the 416. However, a lower combined fare will also encourage within-416 demand on GO making its capacity problems even more critical.
You are focusing entirely on the outer end of the route. The inner end of the trip can, and at core stations more often than not, have feeder connections playing only a minor role. King dumps close to 10K onto its southbound platform in the AM peak hour. I believe Dundas drops around 7K in the AM peak hour on its southbound platform. This volume obviously isn’t going predominantly to the 504 or 505, respectively, because it is impossible for the streetcars to carry that volume. The DRL is going to be in this same general area, through the core, where many end destinations are within walking distance (and PATH links can even make some destinations not normally considered walking distance to be, well, walking distance after all). The O-D pair diversity in the core is especially important since that is where a lot of the market to be served will be concentrated. That requires a tight station spacing. South of College, almost all stations are less than 600m apart (not 600-800) on both Yonge and University lines.
I agree that feeders are the dominant factor at the outer end of the line, but the core must be very local to serve the fine-grained-but-high-concentration demands that makes the existing core-bound subways successful, or a victim of its own success in Yonge’s case.
Steve: There is a difference between an inner stop like King or Dundas in the middle of a very dense commercial area and a stop in low-density residential like Chester. King is a major destination in its own right. Chester is not.
Steve said: “Either you want a line to have feeders implying that stops should be at transfer points, or you are prepared to spent about $100m a pop (plus operating costs) on stations with very low demand.”
My argument is about the choice of subway route: that it is OK to break from the street grid and divert to another street / intersection to optimize the number and convenience of riders coming from feeder routes.
Once the route is selected, we may or may not build those additional mid-block stations for walk-in traffic only; I don’t hold a strong opinion either way. If those stations are built, they do not interfere with what I consider the main function of a subway line: carrying large numbers of riders from one part of the city to another. (They slow the line down a bit, but that’s unlikely to cause any ridership loss.)
Is it set in stone that a transition point from a Queen-DRL northwards must occur via the Weston Galt rail corridor though? And wouldn’t digging a deep enough excavation, like how subways are constructed in Montreal, not avoid all building foundations and utilities?
Steve: The further down you go, the harder it is to build the tunnel, and the longer the grade back up the the surface, wherever that is. If this is a GO train, not a subway (that’s what we are talking about here), there are constraints for gradient, and also requirements related to powering the trains.
As for the various fantasy connections to the BD subway via Ronces or Parkside, well, I won’t comment on the havoc these would create to build through the affected areas.
You are mixing up two separate proposals. One is the GO subway under Queen, and the other is a true DRL. I was commenting on the GO proposal which, by definition, runs on the Weston corridor.
There’s a couple of good justifications NOT to use the rail corridor.
- Too much redundant duplication of service
- Parkdale gets no improved service. An alignment straight across Queen could result in four stations (Dovercourt, Dufferin, Jameson, Roncesvalles)
- Queen/Roncesvalles is a natural terminus point for the streetcar services westward as well a goof interchange point for GO train service. 504 King streetcars (which would still exist if the alignment’s Queen) would transfer here. The 504 could even takeover the 501′s role east of Carlaw.
- Going up Parkside/Keele could be cheaper in the long run. Think about it: no need to stop between Bloor and Queen/Roncesvalles (though a future infill station could exist south of Howard Park though it’d be extremely underused) amking the alignment very rapid. Also at Bloor, because the existing subway’s elevated this leaves the space underneath street level vacant for a new station; and transfer points could be built directly into Keele Station’s existing platform area at the westernmost end as well as into the bus terminal building.
- Future northern extension to Mount Dennis could seamlessly merge into the Weston Galt north of St Clair and result in a Junction Station locale that would be central to the entire community (Annette south exit, Dundas north exit).
- Using the Weston Galt is far more technically complicated given all the other uses of that ROW and would result in an inconvenient transfer point at Sterling/Dundas. The Bloor transfer point would also be prohibitively expensive.
Bain is not justified, nor Dundas East. However, a bare minumum of 6 fundamental stations ought to exist between Pape-Danforth and the Yonge Subway (Jarvis, Parliament, River, Broadview, Carlaw, Gerrard Square). I think it would be a mistake to underestimate the ridership potential of a River Street Stn because it likely would be the gateway stop to the West Don Lands which will extend upwards to Queen St. To the north, BRT routes like 144 Don Valley Express could originate from River Stn and travel north along that street providing local service as well to Regent Park residents.
There are also other potential stops like John, Jarvis and Strachan which won’t necessarily have connecting surface routes but are none of the less in oppurtune locations to serve a lot of customers.
Steve: I mentioned Bain and River because they were in a list proposed in an earlier comment by someone who wanted closely spaced stops whether they had any feeder traffic or not.
The problem with this entire discussion is that the actual proposal we are talking about shifts around with each comment.
I’ve grown tired of bringing up this point, so I’ll make it brief?
Does Metrolinx have *any* documents discussing the possibility of an eventual transition to something closer to the sort of EMU rolling stock found on busier commuter railways more or less anywhere, or does all the planning still hinge on the assumption that Toronto may as well be stuck running locomotive-pulled bilevels for the rest of eternity?
It strikes me that if we had plans a decade ago on how to eventually switch to something more manageable in size than the current 10-to-12-car double-decker behemoths as demand grows and becomes more fine-grained, then trying to figure out where to cram a GO tunnel through downtown would be much less of a dilemma.
Steve: GO’s analysis runs roughly like this: We can only fit “n” trains through Union Station per hour. Therefore, we need to fit as many people on them as possible. Therefore, we must use double-deckers. No, we don’t want to electrify the double-deckers, and given half a chance we wouldn’t electrify anything. Now they are discovering the corner they have painted themselves into.
Steve,
I would be interested in your thoughts on re-purposing the ARL from a premium fee business class service to one that delivers the downtown west relief line service (and more) in the near term. Seems like this requires both a change of mindset at Metrolinx and TTC. Peter Smith for one wants to ensure the ARL business case is framed for business people, not locals. I doubt he can envision how DRL stops at Bathurst, Queen/Dufferin, Bloor, Dupont, Eglinton, Lawrence and Rexdale could ease congestion on Yonge/University/Bloor Subway lines or the core-west transit network. Or how it could take people more directly to the 50000 jobs in the Pearson Eco Business district, and lighten road traffic in the northwest. Gary Webster might have to give priority to adding new integrated stations and picking up some of the cost. And then there is that ridiculous fare issue that gets used as a block for why it cannot be a locally integrated public transit service. As the taxpayer is for providing big bucks for dedicated tracks and frequent service, we should be demanding a better service model. And electric from day 1! Besides providing better stop-start performance for the added stations and making for healthier trackside neighbourhoods, electrification allows for running underground from Bathurst to Queen/Yonge and up the eastern branch of the DRL. This seems at least as worthwhile as running underground across Sheppard.
Steve: The first problem with your premise is the idea that we actually need a “DRL West”, and that it should be at a priority that would be built in the same timeframe as the ARL. The capacity problem into downtown is greater from the east. This is not to say that a northwest route isn’t useful in its own right, but that merging the two ideas is an arrangement of convenience mainly in support of the ongoing fight against the diesel ARL. Any “DRL” needs to have trains with sufficient capacity to have an effect on the larger network. The ARL and its infrastructure do not support this type of operation.
The fare issue and the artificial separation between planning for the TTC and GO networks has forced planning down a path where inside-416 traffic isn’t part of GO’s mandate, and has been ignored until quite recently.
The whole discussion of airport access treats the airport as if it is the centre of the universe rather than one of many destinations, and looks only at downtown-bound travel even though we know this is less than 20% of the total. You speak of the jobs in the airport region, but the people travelling to these jobs originate all over the GTA. If we are going to serve these with transit, we need a network of lines converging on the airport and good local circulation and stop spacing. Neither of these will be provided by a modified ARL. Oddly enough, the Eglinton and Finch LRT lines (moreso Eglinton) would have improved access to the airport from a wide variety of origins across the city, not just downtown. Another issue here is the location of the link from Eglinton to the airport and the relationship with the Mississauga busway. Finally, people bound for the airport should not be forced into an artificial transfer from, say, the Eglinton line to the ARL for the last leg of their journey.
Toronto has been very badly served by much of its transportation planning, and the ARL is a particularly good example. That said, I don’t think that conflating it with a DRL west is an appropriate fix.
This report is really funny for one thing.
The employment growth is expected to be 25% on a much larger base (315k to 400k). Peak period transit demand to the core will grow by over 50% (156k to 236).
Toronto’s downtown employment was over 418,000 in 1989. It dropped, and now it is supposed to be sitting around 400,000 right now. so we basically carried more people over 20 years ago into downtown, and yet this report is saying that something will have to be done to meet demand for 400k.
Someone does not have their numbers right.
Steve: The difference lies in where the people are coming from. All the same, your question raises a good point about how the base of reference may have changed, and how the modelling was done.