The subway juggernaut continues on its way with plans for an extension of the Spadina line northwest to York University. The Environmental Assessment completed recently, and all of the documents are available here. The discussion below is based on information contained in Appendix M regarding travel demand.

I say juggernaut, but really Toronto’s relationship to subway plans is more like a drug addict. We can’t afford them, the lines we want to built don’t do very much for the system overall, but we always want just one more, and we are willing to steal money from any other worthwhile transit project to pay for it.

For decades, planners have told us that subways need high density, concentrated development, but that’s not what we built. Now we have a city and region that need a web of transit services, not a few lines here and there. What’s the current focus? Subways.

Well, let’s have a look at the York University subway extension and see who will use it according to the TTC’s own projections.

The Demand Model

An important thing to remember about demand models is that they are built for road networks. These networks have very different behaviour from transit systems. The first problem is that they treat the metropolitan area as a collection of zones. These zones are much larger than the catchment area of individual stations. Assignment of trips to individual stations sometimes must be done with manual adjustments to correct for the model’s behaviour. When the model is run for existing conditions, it tends to get the numbers right overall, but the actual allocation to each station is off.

As it turns out, almost all of the riding on this line is going either to York University or to the terminal at Steeles, and these are treated as one location for some of the discussion. Allocation of trips to other stations is intriguing, but not all that important.

The EA contains alternative land use scenarios, but their impact on ridership is minimal because most of the riding originates at the end of the line as drive-ins or bus transfers.  Additional development along the line itself will not necessarily generate transit demand for the subway corridor.

A larger problem is the use of a fairly long screenline — Steeles Avenue from Jane Street to the Richmond Hill GO corridor — as a reference point for people crossing into the 416. This wide screenline was used so that changes in GO Rail service from Bradford and Richmond Hill could be included in the model.

One thing to watch carefully here is that a lot of growth occurs on routes other than the York University subway (including the Yonge subway and GO Rail), and we have to put the new subway’s function in the context of what it adds to the overall network. The modelled network includes planned peak-period GO Rail improvements to 2021, but notably this does not include all-day service on any of the north-south lines. The subway system remains the only all-day rail service into the core.

The model deals with the AM Peak period, a three-hour long window. This is standard practice for network planning because the morning peak is the more concentrated (work and school trips tend to overlap in the morning).  Therefore, the AM peak represents the highest expected demand. When we look at numbers from the model, it is important to remember that they are for a three-hour period, and the peak hour will probably be somewhere between 40 to 60 percent of these values.

Finally, the report is silent on whether the model is capacity constrained. In other words, the model may not take into account the ability of the network to accommodate the forecast changes in riding or auto travel. This was a big problem with earlier work for the Sheppard Subway that ignored the impact of modelled ridership on the existing subway system.

Existing Demands

The “before” model uses 2001 data because large-scale travel studies are not done every year and these are the most recent data. The TTC’s Station Usage Counts show a total of 20,445 boardings on the Spadina Subway between Downsview and Eglinton West Station, of which 4,530 are at Downsview. The current network, clearly, is not attracting hordes of riders from the 905.

At the cordon just north of Steeles Avenue, the peak hour transit modal share is about one third with 7,100 trips out of a 22,100 total crossing by transit. Remember that this is between Jane and the Richmond Hill GO line. Most of those transit trips are not headed for the Spadina Subway.

Meanwhile in a separate table showing trips at a line just south of Steeles, the peak period (3 hours) transit total is 18,000 out of a total of 54,100. Of the transit total, 5,200 are on GO Rail.

York University itself has about 5,300 employees and 50,000 students. The transit modal share for employee trips is 10 percent and for students is 40 percent today. This means that today we are getting about 530 employee trips and 20,000 student trips by transit.

An important aspect of the student travel is that it is not concentrated in the peak the same way that regular work trips are. The all-day transit usage by students is higher relative to the peak than for regular commuters. This is factored into the model of future ridership later.

Year 2021 Demand Forecast

By 2021, with the York University extension in place along with various GO rail and bus improvements, the figures for trips just south of Steeles are intriguing. Total peak period transit usage is up from 18,000 to 52,000 while auto trips are up from 36,100 to only 44,300. The transit modal share is now over 50 percent. However, the transit increases come in a variety of places:

• Rapid Transit – 4800
• GO Rail – 4800
• Surface Transit – 24400

The Rapid Transit figure is the Spadina Subway south of Steeles West Station. Much of the surface transit increase will be on lines feeding into the Yonge Subway. The report is silent on whether the Yonge Subway can handle the added demand.

By the time we get south of Finch Avenue, the situation is more dramatic. Total peak period transit usage is up from 33,400 to 61,700 while auto trips are up from 24,600 to 39,100. The transit modal share is now over 60 percent. The distribution of transit increases is:

• Rapid Transit – 26600
• GO Rail – 4900
• Surface Transit – minus 3400

The drop in surface transit is due to the diversion of trips that were on buses in 2001 into the extended Spadina Subway.

Projected link volumes on the Spadina Subway are interesting:

• Steeles to Finch: 13150
• Finch to Sheppard: 15350
• Sheppard to Downsview: 17100

What this shows is that the overwhelming majority of inbound riders will originate at the end of the line from the large carpark and from transit feeder routes.

York University Demand

The forecast assumes that there will be a modest improvement in the modal split for employee trips from 10 to 20 percent, and that for student trips the share will rise from 40 to 55 percent. One vital point here is that some of this increase will arrive on the much-improved feeder bus network from the north, not on the subway. AM peak demand into the university is projected to be 2300 southbound (from the bus network, possibly with a one-stop hop from Steeles West down to York U station) and 5500 northbound. These numbers include both employees and students.

Depending on the assumptions we make about the concentration of demand within the three-hour peak, the peak hour demand would range between 2000 and 3000 for university-bound trips on the subway. This is the demand of a respectable streetcar line or a heavy bus line.

Although York University has always been a strong advocate of the subway extension, the very nature of the demand their campus creates works against their position.

• About 30 percent of their students live outside of Toronto.
• Of the roughly 35,000 who live in the 416, roughly one third will drive to the campus.
• For the 24,000 transit users, the trips to the campus are more dispersed through the day than for conventional commuting.
• The TTC estimates that about 20 percent of the student trips occur during the AM peak, and this brings us down to around 5000, roughly in line with the forecast volume northbound into the campus.

Putting Things in Context

The forecast total ridership on the York University extension in 2021 is about 100,000 per day, or 30-million annually. This is more than double the current riding on the Sheppard line, and so at least we seem headed in the right direction. We would be building a line where there are more riders.

That said, the demand is still unspectacular by subway standards. A large proportion of the riding is commute trips bound for the central area, and the university-bound traffic is both divided by direction and spread out in time.

Looking at the GTA and various plans to increase transit use in the 905, I am bothered that we always look at options for extending the existing 416 subway system outward rather than looking at a 905 transit network building inward. This is particularly important in the plans for a major regional centre in Vaughan that would involve a northward extension of the Spadina line. Does it make sense to continue building subway structure to serve less and less concentrated demand?

Parts of the 905 have discussed LRT (private right-of-way streetcars) options notably in Mississauga and as a future upgrade for the Viva bus network. The flavour of the decade at the Ministry of Transportation seems to be BRT (bus rapid transit) because initial implementation costs are cheaper (as little as a few bus shelters) and it fits in with road expansion plans. Building a BRT network means building or expanding roads. Building LRT means taking space away from cars. The logic is quite simple especially if you believe that transit will never really be an alternative.

What would a network look like if we stopped building subways and considered LRT instead? We would have a north-south spine from somewhere in Vaughan through York University and down to the Spadina subway, although the alignment would probably be different to take advantage of surface rights-of-way. This would connect with an east-west spine in Vaughan itself as well as an east-west line in Toronto, possibly along Finch Avenue.

What would this cost? How soon could we build it? What sort of riding would it attract? Nobody knows, because nobody wants to examine anything other than the subway option. When I asked this of the TTC, the response was, in short, that an Environmental Assessment had already been done and approved for the Spadina extension over 15 years ago. The current study was only an update, and alternate technologies like LRT had already been discarded.

Never mind that the context for such a decision was very different. The Spadina line was then part of the proposed Yonge-Steeles-Spadina loop. The TTC and the GTA had not yet discovered that LRT existed as a viable transit mode. Much of the land now occupied by houses was still open fields. Suburban gridlock was only starting to be a problem.

We cannot afford to be trapped into building plans that are decades old without examining the alternatives. These alternatives must embrace variations in network structure that are possible with different technologies. The Spadina Extension EA has done a thorough job of looking at one option, and I fear that’s the only option we will ever see.